IMO other than the Microsoft IP issue, I think the biggest thing that has shifted since this acquisition was first in the works is Claude Code has absolutely exploded. Forking an IDE and all the expense that comes with that feels like a waste of effort, considering the number of free/open source CLI agentic tools that are out there.
Let's review the current state of things:
- Terminal CLI agents are several orders of magnitude less $$$ to develop than forking an entire IDE.
- CC is dead simple to onboard (use whatever IDE you're using now, with a simple extension for some UX improvements).
- Anthropic is free to aggressively undercut their own API margins (and middlemen like Cursor) in exchange for more predictable subscription revenue + training data access.
What does Cursor/Windsurf offer over VS Code + CC?
- Tab completion model (Cursor's remaining moat)
- Some UI niceties like "add selection to chat", and etc.
Personally I think this is a harbinger of where things are going. Cursor was fastest to $900M ARR and IMO will be fastest back down again.
Agreed on everything. Just to add, not only anthropic is offering CC at like a 500% loss, they restricted sonnet/opus 4 access to windsurf, and jacked up their enterprise deal to Cursor. The increase in price was so big that it forced cursor to make that disastrous downgrade to their plans.
I think only way Cursor and other UX wrappers still win is if on device models or at least open source models catch up in the next 2 years. Then i can see a big push for UX if models are truly a commodity. But as long as claude is much better then yes they hold all the cards. (And don't have a bigger company to have a civil war with like openai)
The way I am doing the math with my Max subscription and assuming DeepSeek API prices, it is still x5 times cheaper. So either DeepSeek is losing money (unlikely) or Anthropic is losing lots of money (more likely). Grok kinda confirms my suspicions. Assuming DeepSeek prices, I've probably spent north of $100 of Grok compute. I didn't pay Grok or Twitter a single cent. $100 is a lot of loss for a single user.
Claude API pricing has significant margin baked in. I think it's safe to assume that anthropic is getting 80% margin on their api and they are selling claude code for less than that.
To me, claude usually feels like a bumbling idiot. But in extremely rare cases it feels like a sentient super intelligence. I facetiously assumed that in those cases it ran on the correct RNG seed.
I'm also curious about this. Claude Code feels very expensive to me, but at the same time I don't have much perspective (nothing to compare it to, really, other than Codex or other agent editors I guess. And CC is way better so likely worth the extra money anyway)
Pretty easy to hit $100 an hour using Opus on API credits. The model providers are heavily subsidized, the datacenters appear to be too. If you look at the Coreweave stuff and the private datacenters it starts looking like the telecom bubble. Even Meta is looking to finance datacenter expansion - https://www.reuters.com/business/meta-seeks-29-billion-priva...
The reason they are talking about building new nuclear power plants in the US isn't just for a few training runs, its for inference. At scale the AI tools are going to be extremely expensive.
Also note China produces twice as much electricity as the United States. Software development and agent demand is going to be competitive across industries. You may think, oh I can just use a few hours of this a day and I got a week of work done (happens to me some days), but you are going to end up needing to match what your competitors are doing - not what you got comfortable with. This is the recurring trap of new technology (no capitalism required.)
There is a danger to independent developers becoming reliant on models. $100-$200 is a customer acquisition cost giveaway. The state of the art models probably will end up costing hourly what a human developer costs. There is also the speed and batching part. How willing is the developer to, for example, get 50% off but maybe wait twice as long for the output. Hopefully the good dev models end up only costing $1000-$2000 a month in a year. At least that will be more accessible.
Somewhere in the future these good models will run on device and just cost the price of your hardware. Will it be the AGI models? We will find out.
I wonder how this comment will age, will look back at it in 5 or 10 years.
Your excellent comments make me grateful that I am retired and just work part time on my own research and learning. I believe you when you say professional developers will need large inference compute budgets.
Probably because I am an old man, but I don’t personally vibe with full time AI assistant use, rather I will use the best models available for brief periods on specific problems.
Ironically, when I do use the best models available to me it is almost always to work on making weaker and smaller models running on Ollama more effective for my interests.
BTW, I have used neural network tech in production since 1985, and I am thrilled by the rate of progress, but worry about such externalities as energy use, environmental factors, and hurting the job market for many young people.
I've been around for a while (not quite retirement age) and this time is the closest to the new feeling I had using the internet and web in the early days. There are simultaneously infinite possibilities but also great uncertainty what pathways will be taken and how things will end up.
There are a lot of parts in the near term to dislike here, especially the consequences for privacy, adtech, energy use. I do have concerns that the greatest pitfalls in the short terms are being ignored while other uncertainties are being exaggerated. (I've been warning on deep learning model use for recommendation engines for years, and only a sliver of people seem to have picked up on that one, for example.)
On the other hand, if good enough models can run locally, humans can end up with a lot more autonomy and choice with their software and operating systems than they have today. The most powerful models might run on supercomputers and just be solving the really big science problems. There is a lot of fantastic software out there that does not improve by throwing infinite resources at it.
Another consideration is while the big tech firms are spending (what will likely approach) hundreds of billions of dollars in a race to "AGI", what matters to those same companies even more than winning is making sure that the winner isn't a winner takes all. In that case, hopefully the outcome looks more like open source.
The SOTA models will always run in data centers, because they have 5x or more VRAM and 10-100x the compute allowance. Plus, they can make good use of scaling w/ batch inference which is a huge power savings, and which a single developer machine doesn’t make full use of.
Yes I do, it’s just for new comers who are used to cursor where, without careful prompting you just lock yourself out of premium requests, that’s not immediately a given that CC is more dangerous and does not work the same way at all.
Of course it requires careful planning but the trap is easy to fall into.
In my experience, it's more about the tool's local indexing and aggressive automatic upload and model usage limitations to avoid them (and you) overpaying.
People are recreating this with local toolchains now.
This is around what what Cursor was costing me with Claude 4 Opus before I switched to Claude Code. Sonnet works fine for some things, but for some projects it spews unusable garbage, unless the specification is so detailed that it's almost the implementation already.
This is where something like Perplexity's "memory" feature is really great. It treats other threads similarly to web resources.
I would love to understand better just how Perplexity is able to integrate up-to-date sources like other theads (and presumably recent web searches, but I haven't verified this, they could be just from the latest model) into it's query responses. It feels seamless.
Have you been human before? competition for resources and status is an instinctive trait.
It rears its head regardless of what sociopolitical environment you place us in.
You’re either competing to offer better products or services to customers…or you’re competing for your position in the breadline or politburo via black markets.
Even in the Soviet Union there were multiple design bureaus competing for designs of things like aircraft. Tupolec, Ilyushin, Sukhoi, Mikoyan-Gyurevich (MiG), Yakolev, Mil. There were quite a lot. Several (not all, they had their specialisations) provided designs when a requirement was raised. Not too different from the US yet not capitalist.
Not really, it's possible with any market economy, even a hypothetical socialist one (that is, one where all market actors are worker-owned co-ops).
And, since there is no global super-state, the world economy is a market economy, so even if every state were a state-owned planned economy, North Korea style, still there would exist this type of competition between states.
Consider also that VC funds often have pension funds as their limited partners. Workers have a claim to their pension, and thus a claim to the startup returns that the VC invests in.
So yeah it basically comes down to your definition of "worker-owned". What fraction of worker ownership is necessary? Do C-level execs count as workers? Can it be "worker-owned" if the "workers" are people working elsewhere?
Beyond the "worker-owned" terminology, why is this distinction supposed to matter exactly? Supposing there was an SV startup that was relatively generous with equity compensation, so over 50% of equity is owned by non-C-level employees. What would you expect to change, if anything, if that threshold was passed?
> Supposing there was an SV startup that was relatively generous with equity compensation, so over 50% of equity is owned by non-C-level employees. What would you expect to change, if anything, if that threshold was passed?
If the workers are majority owners, then they can, for example, fire a CEO that is leading the company in the wrong direction, or trying to cut their salaries, or anything like that.
>If the workers are majority owners, then they can, for example, fire a CEO that is leading the company in the wrong direction, or trying to cut their salaries, or anything like that.
Why wouldn't the board fire said CEO?
The most common reason to cut salaries is if the company is in dire financial straits regardless. Co-ops are more likely to cut salary and less likely to do layoffs.
Because the board doesn't understand the business at the level that employees do. Or because the board has different goals for the business than employees do. Or because the board is filled with friends of the CEO who let them do whatever.
Also, lots of companies reduce salaries or headcount if they feel they can get away with it. They don't need to be in dire financial straights, it's enough to have a few quarters of no or low growth and to want to show a positive change.
How specifically would you expect a typical SV corp's policies to change if employee equity passes from 49% to 51%?
Remember, if employees own 49%, if they can persuade just 2% of the other shareholders that a change will be positive for the business, they can make that change. So minority vs majority is not as significant as it may seem.
Can you give me an idea of how much interaction would be $50-$100 per day? Like are you pretty constantly in a back and forth with CC? And if you wouldn’t mind, any chance you can give me an idea of productivity gains pre/post LLM?
Yes, a lot of usage, I’d guess top 10% among my peers. I do 6-10hrs of constant iterating across mid-size codebases of 750k tokens. CC is set to use Opus by default, which further drives up costs.
Estimating productivity gains is a flame war I don’t want to start, but as a signal: if the CC Max plan goes up 10x in price, I’m still keeping my subscription.
I maintain top-tier subscription to every frontier service (~$1k/mo) and throughout the week spend multiple hours with each of Cursor, Amp, Augment, Windsurf, Codex CLI, Gemini CLI, but keep on defaulting to Claude Code.
I am curious what kind of development you’re doing and where your projects fall on the fast iteration<->correctness curve (no judgment). I’ve used CC Pro for a few weeks now and I will keep it, it’s fantastically useful for some things, but it has wasted more of my time than it saved when I’ve experimented with giving it harder tasks.
It's interesting to work with a number of people using various models and interaction modes in slightly different capacities. I can see where the huge productivity gains are and can feel them, but the same is true for the opposite. I'm pretty sure I lost a full day or more trying to track down a build error because it was relatively trivial fpr someone to ask CC or something to refactor a ton of files, which it seems to have done a bit too eagerly. On the other hand, that refactor would have been super tedious, so maybe worth it?
Mostly to save money (I am retired) I mostly use Gemini APIs. I used to also use good open weight models on groq.com, but life is simpler just using Gemini.
Ultimately, my not using the best tools for my personal research projects has zero effect on the world but I am still very curious what elite developers with the best tools can accomplish, and what capability I am ‘leaving on the table.’
I’m a founder/CTO of an enterprise SaaS, and I code everything from data modeling, to algos, backend integrations, frontend architecture, UI widgets, etc. All in TypeScript, which is perfectly suited to LLMs because we can fit the types and repo map into context without loading all code.
As to “why”: I’ve been coding for 25 years, and LLMs is the first technology that has a non-linear impact on my output. It’s simultaneously moronic and jaw-dropping. I’m good at what I do (eg, merged fixes into Node) and Claude/o3 regularly finds material edge cases in my code that I was confident in. Then they add a test case (as per our style), write a fix, and update docs/examples within two minutes.
I love coding and the art&craft of software development. I’ve written millions of lines of revenue generating code, and made millions doing it. If someone forced me to stop using LLMs in my production process, I’d quit on the spot.
Why not self host: open source models are a generation behind SOTA. R1 is just not in the same league as the pro commercial models.
> If someone forced me to stop using LLMs in my production process, I’d quit on the spot.
Yup 100% agree. I’d rather try to convince them of the benefits than go back to what feels like an unnecessarily inefficient process of writing all code by hand again.
And I’ve got 25+ years of solid coding experience. Never going back.
> data modeling, to algos, backend integrations, frontend architecture, UI widgets, etc. All in TypeScript, which is perfectly suited to LLMs because we can fit the types and repo map into context without loading all code.
Which frameworks & libraries have you found work well in this (agentic) context? I feel much of the js lib. landscape does not do enough to enforce an easily-understood project structure that would "constrain" the architecture and force modularity. (I might have this bias from my many years of work with Rails that is highly opinionated in this regard).
When you say generation behind, can you give a sense of what that means in functionality per your current use? Slower/lower quality, it would take more iterations to get what you want?
Context rot. My use case is iterating over a large codebase which quickly grows context. All LLMs degrade with larger context sizes, well below their published limits, but pro models degrade the least. R1 gets confused relatively quickly, despite their published numbers.
I think Fiction LiveBench captures some of those differences via a standardized benchmark that spreads interconnected facts through an increasingly large context to see how models can continue connecting the dots (similar to how in codebases you often have related ideas spread across many files)
i've been using llm-based tools like copilot and claude pro (though not cc with opus), and while they can be helpful – e.g. for doc lookups, repetitive stuff, or quick reminders – i rarely get value beyond that. i've honestly never had a model surface a bug or edge case i wouldn’t have spotted myself.
i've tried agent-style workflows in copilot and windsurf (on claude 3.5 and 4), and honestly, they often just get stuck or build themselves into a corner. they don’t seem to reason across structure or long-term architecture in any meaningful way. it might look helpful at first, but what comes out tends to be fragile and usually something i’d refactor immediately.
sure, the model writes fast – but that speed doesn't translate into actual productivity for me unless it’s something dead simple. and if i’m spending a lot of time generating boilerplate, i usually take that as a design smell, not a task i want to automate harder.
so i’m honestly wondering: is cc max really that much better? are those productivity claims based on something fundamentally different? or is it more about tool enthusiasm + selective wins?
> I’ve written millions of lines of revenue generating code
This is a wild claim.
Approx 250 working days in a year. 25 years coding. Just one million lines would be phenom output, at 160 lines per day forever. Now you are claiming multiple millions? Come on.
It's impossible as an IC on a team, or working where a concept of "tickets" exists. It's unavoidable as a solo founder, whether you're building enterprise systems or expanding your vision. Some details -
1. Before wife&kids, every weekend I would learn a library or a concept by recreating it from scratch. Re-implementing jQuery, fetch API via XHR, Promises, barebones React, a basic web router, express + common middlewares, etc. Usually, at least 1,000 lines of code every weekend. That's 1M+ over 25 years.
2. My last product is currently 400k LOCs, 95% built by me over three years. I didn't one-shot it, so assuming 2-3x ongoing refactors, that's more than 1M LOCs written.
3. In my current product repo, GitHub says for the last 6 months I'm +120k,-80k. I code less than I used to, but even at this rate, it's safely 100k-250k per year (times 20 years).
4. Even in open source, there are examples like esbuild, which is a side project from one person (cofounder and architect of Figma). esbuild is currently at ~150k LOCs, and GitHub says his contributions were +600k,-400k.
5. LOCs are not the same. 10k lines of algorithms can take a month, but 10K of React widgets is like a week of work (on a greenfield project where you know exactly what you're building). These days, when a frontend developer says their most extensive UI codebase was 100k LOCs in an interview, I assume they haven't built a big UI thing.
So yes, if the reference point is "how many sprint tickets is that", it seems impossible. If the reference point is "a creative outlet that aligns with startup-level rewards", I think my statement of "millions of lines" is conservative.
Granted, not all of it was revenue-generating - much was experimental, exploratory, or just for fun. My overarching point was that I build software products for (great) living, as opposed to a marketer who stumbled into Claude Code and now evangelizes it as some huge unlock.
No, it’s not. At all. At the overwhelming majority of companies I’ve worked for or heard of, even 400-500 lines fully shipped in a week, slightly less than your figure here, would be top quartile of output - but further, it isn’t necessarily the point. Writing lines of code is a pretty small part of the job at companies with more than about 5-6 engineers on staff, past that it’s a lot more design and architecture and LEGO-brick-fitting - or just politicking and policying. Heck, I know folks who wish they could ship 400 lines of code a month, but are held back by the bureaucracies of their companies.
Now extrapolate. That’s maybe 50k a year assuming some PTO.
10 years would make 500k and you just cross a million at 20.
So that would have to be 20 years straight of that style of working and you’re still not into plural millions until 40 years.
If someone actually produced multiple millions of lines in 25 years, it would have to be a side effect of some extremely verbose language where trivial changes take up many lines (maybe Java).
Unless you're getting paid for your commute, you're just giving your employer free productivity. I would recommend doing literally anything else with that time. Read a book, maybe.
If you can't do your job in your 8 hours then you're either not good enough or the requirements are too much and the company should change processes and hire.
Right, I'm not saying anyone should actually be in the office 40 hours a week that sounds terrible. And even with all the RTO of the last couple years that doesn't seem to be expected many places.
Personally I use dev containers on a server and I have written some template containers for quickly setting up new containers that has claude code and some scripts for easily connecting to the right container etc. Makes it possible to work on mobile,but lots of room for improvement in the workflow still.
The project is just a web backend. I give Claude Code grunt work tasks. Things like "make X operation also return Y data" or "create Z new model + CRUD operations". Also asking it to implement well-known patterns like denouncing or caching for an existing operation works well.
My app builds and runs fine on Termux, so my CLAUDE.md says to always run unit tests after making changes. So I punch in a request, close my phone for a bit, then check back later and review the diff. Usually takes one or two follow-up asks to get right, but since it always builds and passes tests, I never get complete garbage back.
There are some tasks that I never give it. Most of that is just intuition. Anything I need to understand deeply or care about the implementation of I do myself. And the app was originally hand-built by me, which I think is important - I would not trust CC to design the entire thing from scratch. It's much easier to review changes when you understand the overall architecture deeply.
you can easily reach 50$ per day.
by force switching model to opus
/model opus
it will continue to use opus eventhough there is a warning about approaching limit.
i found opus is significantly more capable in coding than sonnet, especcially for the task that is poorly defined, thinking mode can fulfill alot of missing detail and you just need to edit a little before let it code.
Claude Code with a Claude subscription is the cheap version for current SOTA.
"Agentic" workflows burn through tokens like there's no tomorrow, and the new Opus model is so expensive per-token that the Max plan pays itself back in one or two days of moderate usage. When people reports their Claude Code sessions costing $100+ per day, I read that as the API price equivalent - it makes no sense to actually "pay as you go" with Claude right now.
This is arguably the cheapest option available on the market right now in terms of results per dollar, but only if you can afford the subscription itself. There's also time/value component here: on Max x5, it's quite easy to hit the usage limits of Opus (fortunately the limit is per 5 hours or so); Max x20 is only twice the price of Max x5 but gives you 4x more Opus; better model = less time spent fighting with and cleaning up after the AI. It's expensive to be poor, unfortunately.
>less time spent fighting with and cleaning up after the AI.
I've yet to use anything but copilot in vscode, which is 1/2 the time helpful, and 1/2 wasting my time. For me it's almost break-even, if I don't count the frustration it causes.
I've been reading all these AI-related comment sections and none of it is convincing me there is really anything better out there. AI seems like break-even at best, but usually it's just "fighting with and cleaning up after the AI", and I'm really not interested in doing any of that. I was a lot happier when I wasn't constantly being shown bad code that I need to read and decide about, when I'm perfectly capable of writing the code myself without the hasle of AI getting in my way.
AI burnout is probably already a thing, and I'm close to that point already. I do not have hope that it will get much better than it is, as the core of the tech is essentially just a guessing game.
I tend to agree except for one recent experience: I built a quick prototype of an application whose backend I had written twice before and finally wanted to do right. But the existing infrastructure for it had bit-rotted, and I am definitely not a UI person. Every time I dive into html+js I have to spend hours updating my years-out-of-date knowledge of how to do things.
So I vibe coded it. I was extremely specific about how the back end should operate and pretty vague about the UI, and basically everything worked.
But there were a few things about this one: first, it was just a prototype. I wanted to kick around some ideas quickly, and I didn't care at all about code quality. Second, I already knew exactly how to do the hard parts in the back end, so part of the prompt input was the architecture and mechanism that I wanted.
But it spat out that html app way way faster than I could have.
Claude Code pro is ~$20USD/ month and is nearly enough for someone like me who can’t use it at work and is just playing around with it after work. I’m loving it.
cursor on a $20/month plan (if you burn thru the free credits) or gemini-cli (free) are 2 great ways to try out this kinda stuff for a hobbyist. you can throw in v0 too, $5/month free credits. susana’s free tier can give you a db as well.
Zed is fantastic. Just dipping my toes in agentic AI, but I was able to fix a failing test I spent maybe 15 minutes trying to untangle in a couple minutes with Zed. (It did proceed to break other tests in that file though, but I quickly reverted that.)
It is also BYOA or you can buy a subscription from Zed themselves and help them out. I currently use it with my free Copilot+ subscription (GitHub hands it out to pretty much any free/open source dev).
You can tell Claude Code to use opus using /model and then it doesn't fall back to Sonnet btw. I am on the $100 plan and I hit rate-limits every now and then, but not enough to warrant using Sonnet instead of Opus.
This is what I don’t get about the cost being reported by Claude code. At work I use it against our AWS Bedrock instance, and most sessions will say 15/20 dollars and I’ll have multiple agents running. So I can easily spend 60 bucks a day in reported cost. Our AWS Bedrock bill is only a small fraction of that? Why would you over charge on direct usage of your API?
Seems like the survival strategy for cursor would be to develop their own frontier coding model. Maybe they can leverage the data from their still somewhat significant lead in the space to make a solid effort.
I don’t think that’s a viable strategy. It is very very hard and not many people can do it. Just look at how much Meta is paying to poach the few people in the world capable of training a next gen frontier model.
The basic concept plus a lot of money spent on compute and training data gets you pretraining. After that to get a really good model there’s a lot more fine-tuning / RL steps that companies are pretty secretive about. That is where the “smart decisions” and knowledge gained by training previous generations of sota models comes in.
We’d probably see more companies training their own models if it was cheaper, for sure. Maybe some of them would do very well. But even having a lot of money to throw at this doesn’t guarantee success, e.g. Meta’s Llama 4 was a big disappointment.
That said, it’s not impossible to catch up to close to state-of-the-art, as Deepseek showed.
I’d also add that no one predicted the emergent properties of LLMs as they followed the scaling laws hypothesis. GPT showed all kinds of emergent stuff like reasoning/sentiment analysis when we went up an order of magnitude on the number of parameters. We don’t don’t actually know what would emerge if we trained a quadrillion param model. SOTA will always be mysterious until we reach those limits, so, no, companies like Cursor will never be on the frontier. It takes too much money and requires seeking out things we haven’t ever seen before.
There are plenty of people theoretically capable of doing this, I secretly believe some of the most talented people in this space are randos posting on /r/LocalLlama.
But the truth is to have experience building models at this scale requires working at a high level job at a major FAANG/LLM provider. Building what Meta needs is not something you can do in your basement.
The reality is the set of people who really understand this stuff and have experience working on it at scale is very, very small. And the people in this space are already paid very well.
It's a staggeringly bad deal. It's a hugely expensive task where unless you are the literal best in the world, you would never even see any usage. And even for those who are BOTH best and well known they have to be willing to lose billions on repeat with no end in sight.
It's very very rare to have winner takes all to such an extreme degree as code llm models
I don't think it's literally "winner takes all" - I regularly cycle between Gemini, DeepSeek and Claude for coding tasks. I'm sure any GPT model would be fine too, and I could even fall back to Qwen in a pinch (exactly what I did when I was in China recently with no ability to access foreign servers).
Claude does have a slight edge in quality (which is why it's my default) but infrastructure/cost/speed are all relevant too. Different providers may focus on one at the expense of the others.
One interesting scenario where we could end up is using large hosted models for planning/logic, and handing off to local models for execution.
I'd recommend reading some of the papers on what it takes to actually train a proper foundation model, such as the Llama 3 Herd of Models paper. It is a deeply sophisticated process.
Coding startups also try to fine-tune OSS models to their own ends. But this is also very difficult, and usually just done as a cost optimization, not as a way to get better functionality.
You need a person that can hit the ground running. Compute for LLM is extremely capital intensive and you’re always racing against time. Missing performance targets can mean life or death of the company.
As an actual user of Windsurf model, I don't think "tried" is fair. I sometimes use it. It's not as smart as Gemini but it iterates quicker and is very well aligned with their tool calls
- Forking VSCode is very easy; you can do it in 1 hour.
- Anthropic doesn't use the inputs for training.
- Cursor doesn't have $900M ARR. That was the raise. Their ARR is ~$500m [1].
- Claude Code already support the niceties, including "add selection to chat", accessing IDE's realtime warnings and errors (built-in tool 'ideDiagnostics'), and using IDE's native diff viewer for reviewing the edits.
The cost of vscode fork is that microsoft has restricted extension marketplace for forks. You have to maintain separate one, that is the real dealbreaker
Their base is $20/mth. That would equal 3.75M people paying a sub to Cursor.
If literally everyone is on their $200/mth plan, then that would be 375K paid users.
There’s 50M VS Code + VS users (May 2025). [1] 7% of all VS Code users having switched to Cursor does not match my personal circle of developers. 0.7% . . . Maybe? But, that would be if everyone using Cursor were paying $200/month.
Seems impossibly high, especially given the number of other AI subscription options as well.
The $20/month cursor sub is heavily limited though, for basic casual usage that's fine but you VERY soon run into its limits when working at any speed.
I also just prefer CC's UX. I've tried to make myself use Copilot and Roo and I just couldn't. The extra mental overhead and UI context-switching took me out of the flow. And tab completion has never felt valuable to me.
But the chat UX is so simple it doesn't take up any extra brain-cycles. It's easier to alt-tab to and from; it feels like slacking a coworker. I can have one or more terminal windows open with agents I'm managing, and still monitor/intervene in my editor as they work. Fits much nicer with my brain, and accelerates my flow instead of disrupting it
There's something starkly different for me about not having to think about exactly what context to feed to the tool, which text to highlight or tabs to open, which predefined agent to select, which IDE button to press
Just formulate my concepts and intent and then express those in words. If I need to be more precise in my words then I will be, but I stay in a concepts + words headspace. That's very important for conserving my own mental context window
Claude Code is just proving that coding agents can be successful. The interface isn’t magic, it just fits the model and integrates with a system in all the right ways. The Anthropic team for that product is very small comparatively (their most prolific contributor is Claude), and I think it’s more of a technology proof than a core competency - it’s a great API $ business lever, but there’s no reason for them to try and win the “agentic coding UI” market. Unless Generative AI flops everywhere else, these markets will continue to emerge and need focus. The Windsurf kerfuffle is further proof that OpenAI doesn’t see the market as must-win for a frontier model shop.
And so I’d say this isn’t a harbinger of the death of Cursor, instead proof that there’s a future in the market they were just recently winning.
I was being hyperbolic saying their ARR will go to zero. That's obviously not the case, but the point is that CC has revealed their real product was not "agentic coding UI", it was "insanely cheap tokens". I have no doubt they will continue to see success, but their future right now looks closer to being a competitor to free/open tools like cline/roo code, as well as the CLI entrants, not a standalone $500M ARR juggarnaut. They have no horse in the race in the token market, they're a middleman.
They either need to create their own model and compete on cost, or hope that token costs come down dramatically so as to be too cheap to meter.
Digging in here more... why would you say it isn't in Anthropic's interest to win the "agentic coding UI" market?
My mental model is that these foundation model companies will need to invest in and win in a significant number of the app layer markets in order to realize enough revenue to drive returns. And if coding / agentic coding is one of the top X use cases for tokens at the app layer, seems logical that they'd want to be a winner in this market.
Is your view that these companies will be content to win at the model layer and be agnostic as to the app layer?
My intuition is that their fundamental business is executing on the models, and any other products are secondary and exist to drive revenue that they can use to compete against Google/OpenAI/Meta as well as to ensure - and demonstrate - that their models are performant in these new markets. Claude needs to be great at coding, but Anthropic doesn’t need to own Coding. Claude Code is growing their core business, just like a Claude Robotics or a Claude Scheduling might, but they cant focus on robotics or scheduling because that takes them away from the core business of models. A strategic relationship with Cursor might have been enough to accomplish this, but it wasn’t - maybe Cursor couldn’t execute fast enough, or didn’t align on priorities, or whatever. I’ve watched a bunch of interviews with the CC team and I very much get the impression that it was more “holy shit, this works great” than a product strategy.
You may be right about “they need to invest in and win” in order to have __enough__ revenue to outcompete the nation-state sized competition, but this stuff is moving way to fast for anyone know.
Cursor see it coming - it's why they're moving to the web and mobile[0]
The bigger issue is the advantage Anthropic, Google and OpenAI have in developing and deploying their own models. It wasn't that long ago that Cursor was reading 50 lines of code at a time to save on token costs. Anthropic just came out and yolo'd the context window because they could afford to, and it blew everything else away.
Cursor could release a cli tomorrow but it wouldn't help them compete when Anthropic and Google can always be multiples cheaper
> Anthropic just came out and yolo'd the context window because they could afford to
I don’t think this is true at all. The reason CC is so good is that they’re very deliberate about what goes in the context. CC often spends ages reading 5 LOC snippets, but afterwards it only has relevant stuff in context.
Heard a lot of this context bs parroted all over HN, don't buy it. If simply increasing context size can solve problem, Gemini would be the best model for everything.
I think this is an interesting and cool direction for Cursor to be going in and I don't doubt something like this is the future. But I have my doubts whether it will save them in the short/medium term:
- AI is not good enough yet to abandon the traditional IDE experierence if you're doing anything non-trivial. Hard finding use cases for this right now.
- There's no moat here. There are already a dozen "Claude Code UI" OSS projects with similar basic functionality.
Strictly speaking about large, complex, sprawling codebases, I don't think you can beat the experience that an IDE + coding agent brings with a terminal-based coding agent.
Auto-regressive nature of these things mean that errors accumulate, and IDEs are well placed to give that observability to the human, than a coding agent. I can course correct more easily in an IDE with clear diffs, coding navigation, than following a terminal timeline.
You can view and navigate the diffs made by the terminal agent in your IDE in realtime, just like Cursor, as well as commit, revert, etc. That’s really all the “integration” you need.
Some excellent points. On “add selection to chat”, I just want to add that the Claude Code VS code extension automatically passes the current selection to the model. :)
I am genuinely curious if any Cursor or Windsurf users who have also tried Claude Code could speak to why they prefer the IDE-fork tools? I’ve only ever used Claude Code myself - what am I missing?
Cursor's tab completion model is legitimately fantastic and for many people is worth the entire $20 subscription. Lint fixes or syntax-level refactors are guessed and executed instantly with TAB with close to 100% accuracy. This is their final moat IMO, if Copilot manages to bring their tab completion up to near parity, very little reason to use Cursor.
Idk. When you're doing something it really gets it's super nice, but it's also off a lot of times and it's IMO super distracting when it constantly pop up. No way to explicitly request it instead - other than toggling, which seems to also turn off context/edit tracking, because after toggling on it does not suggest anything until you make some edits.
While Zed's model is not as good the UI is so much better IMO.
Just to offer a different perspective, I use Cursor at work and, coming from emacs (which I still use) with copilot completions only when I request them with a shortcut, Cursor’s behavior drives me crazy.
Which Emacs Package do you use for CoPilot, i tried using Copilot.el a long while ago, but had problems with it. Is there something new or does copilot.el fulfill your needs?
I only get suggestions if I use that key (the prefix looks huge but I have a Keyboardio Model 100 and I have that bound to the Any key, so I intentionally picked a crazy-long prefix hoping to avoid collision with other shortcuts) and that's the way I like to use these tools (which is why the behaviour in Cursor drives me crazy, though I admit I haven't spend time looking at its configuration, maybe it's something that can be turned off).
I haven't used Cursor or Claude much, how different is it from Copilot? I bounce between desktop ChatGPT (which can update VS Code) and copilot. Is there an impression that those have fallen behind?
IME, one of execution. Copilot is like having your cousin who works at Bestbuy try and help you code - it knows what a computer is, and speaks english, but is pretty bad at both
The story I've heard is that Cursor is making all their money on context management and prompting, to help smooth over the gap between "you know what I meant" and getting the underlying model to "know what you meant"
I haven't had as much experience with Claude or Claude Code to speak to those, but my colleagues speak of them highly
It's quite interesting how little the Cursor power users use tab. Majority of the posts are some insane number of agent edits and close to (or exactly) 0 tabs.
At my company we have an enterprise subscription and we're also all allowed to see the analytics for the entire company. Last I checked, I was literally the number one user of Tab and middle of the pack for agent.
It's interesting when I see videos or reddit posts about cursor and people getting rate limited and being super angry. In my experience tab is the number one feature, and I feel like most people using agent are probably overusing it tasks that would honestly take less time to do myself or using models way smarter than they need to be for the task at hand.
I use cursor strictly for agent edits and do anything else in a proper IDE meaning in a Jetbrains product that I run in a separate window.
Many of my co-workers do the same. VC Code is vastly inferior when it comes to editing and actual IDE feature so it is a non-starter when you do programming yourself.
I once tried AI tab-complete on Zed and it was all right but breaks my flow. Either the AI does the editing or I do it but mixing both annoys me.
I'd like to ask the opposite question: why do people prefer command line tools? I tried both and I prefer working in IDE. The main reason is that I don't trust the LLMs too much and I like to see and potentially quickly edit the changes they make. With an IDE, I can iterate much faster than with the command line tool.
I haven't tried Claude Code VS Code extension. Did anyone replaced Cursor with this setup?
I replaced. My opinion: Cursor sucks as an IDE.
Cursor may have a average to above average quality in IDE assistance - but the IDE seems to get in the way. It's entire performance is based on the real-time performance and latency from their servers and sometimes it is way too slow. The TAB autocomplete that was working for you in the last 30 minutes suddenly doesn't work randomly, or just experiences severe delays that it stops making sense.
Besides that, the IDE seems poorly designed - some navigation options are confusing and it makes way too many intrusive changes (ex: automatically finishing strings).
I've since gone back to VS Code - with Cline (with OpenRouter and super cheap Qwen Coder models, Windsurf FREE, Claude Code with $20 per month) and I get great mileage from all of them.
You're looking at (coloured) diffs in your shell is all when it comes to coding. It's pretty easy to setup MCP and have claude be the director. Like I have zen MCP running with an OpenRouter API key, and will ask claude to consult with pro (gemini) or o3, or both to come up with an architecture review / plan.
I honestly don't know how great that is, because it just reiterates what I was planning anyways, and I can't tell if it's just glazing, or it's just drawing the same general conclusions. Seriously though, it does a decent job, and you can discuss / ruminate over approaches.
I assume you can do all the same things in an editor. I'm just comfortable with a shell is all, and as a hardcore Vi user, I don't really want to use Visual Studio.
I also use vim heavily and I've found that I'm really enjoying Cursor + VS Code Vim extension. The cursor tab completion works very nicely in conjunction with vim navigate mode.
heh, including "for diffing" is selling short when our new job as software developers now seems to be reviewing code, of which looking at a diff is only one tiny part. That goes infinitely more for dynamically typed languages, where there is no compiler to catch dumb typos. If I have to actually, no kidding, review code then I want all the introspections, find references, go to declaration, et al for catching the intern trying to cheat me
I can roll back to different checkpoints with Cursor easily. Maybe CC has it but the fact that I haven’t found it after using it daily is an example of Cursor having a better UX for me.
I like using Claude Code through Roo Code (vscode extension). I find it easier to work with text using a mouse, vscode diff viewer etc. I guess if you're very good at vim shortcuts etc you can use that in Claude Code instead of selecting text with a mouse. Claude Code has a vscode extension too so I feel that using Claude Code through vscode just adds a better UI.
Using cline for a bit made me realize cursor was doomed. Everything is just a gpt/anthropic wrapper of fancy prompts.
I can do most of what I want with cline, and I've gone back from large changes to just small changes and been moving much quicker. Large refactors/changes start to deviate from what you actually want to accomplish unless you have written a dissertation, and even then they fail.
I agree with all you’ve said but with regards to writing a dissertation for larger changes : have you tried letting it first right a plan for you as markdown (just keep this file uncommitted) and then let it build a checklist of things to do?
I find just referencing this file over and over works wonders and it respects items that were already checked off really well.
I can get a lot done really fast this way in small enough chunks so i know every bit of code and how it works (tweaking manually of course where needed).
But I can blow through some tickets way faster than before this way.
IIRC problem is that VS Code does not allow extensions to create custom UI in the panels areas except for WebViews(?). It makes for not a great experience. Plus Cursor does a lot with background indexing to make their tab completion model really good - more than would be possible with the extensions APIs available.
Not if you want custom UI. There are a lot of things you can do in extension land (continue, cline, roocode, kilocode, etc. are good examples) but there are some things you can't.
One thing I would have thought would be really cool to try is to integrate it at the LSP level, and use all that good stuff, but apparently people trying (I think there was a company from .il trying) either went closed or didn't release anything note worthy...
When the Copilot extension needs a new VS Code feature it gets added, but it isn't available to third party extensions until months later... Err, years later... well, whenever Microsoft feels like it.
So an extension will never be able to compete with Copilot.
I've really struggled with using the extensions - their UI/UX is worse, they're much more limited in what they can do and they're much more unstable (in IntelliJ at least).
Then again writing mostly kotlin I cannot get along with the VS Code forks as they're just not that great outside of typescript projects in my experience.
I tend to prompt in cursor/windsurf and refactor in IntelliJ which is okay but a bit of a pain.
I use Augment extensively and find it superior to cursor in every way - and operates as an extension. It has a really handy task planning interface and meta prompt refinement feature and the costs are remarkably low. The quality of output implantation is higher IMO and I don’t have to do a lot of model selection and don’t get Max model bill explosions. If there’s something Cursor provided that Augment doesn’t via extension it was not functionally useful enough to notice.
I think Augment has been flying under the radar for many people, and really reserve better marketing.
I've been using Augment for over a year with IntelliJ, and never understood why my colleagues were all raving about Cursor and Windsurf. I gave Cursor a real try, but it wasn't any better, and the value proposition of having to adopt a dedicated IDE wasn't attractive to me.
A plugin to leverage your existing tools makes a lot more sense than an IDE. Or at least until/if AI agents get so smart that you don't need most of the IDE's functionality, which might change what kinds of tooling are needed when you're in the passenger seat rather than the driver's seat.
One competitor to Claude Code that I don't hear much about is Jetbrains Junie. From my experience, the code it generates is as good as CC, and if you've purchased a Jetbrains license you probably have some amount of free Junie every month.
I'll fill in some context. I think the value of Cursor as an IDE is probably somewhat ephemeral. It's mostly combating Microsoft's ambitions to keep other players busy and box them out of the market. Agents gain a lot of value from model context protocol and there's an amazingly short list of clients that fully support the protocol, but the VSCode Chat window is one of them: https://modelcontextprotocol.io/clients
I actually do prefer the view that having the agent built into an IDE brings me but I'll be damned if I'm forced to use CoPilot/OpenAI. Second to that, the agent does have access to a lot more contextual tools by being built into the editor like focused linting errors and test failures. Of course that demands your development environment is setup correctly and could be replicatable with Claude Code to some extent.
I never got the valuation. I (and many others) have built open source agent plugins that are pretty much just as good, in our free time (check out magenta nvim btw, I think it turned out neat!)
> with a simple extension for some UX improvements
What are the UX improvements?
I was using the Pycharm plugin and didn’t notice any actual integration.
I had problems with pycharm’s terminal—not least of which was default 5k line scroll back which while easy to change was worst part of CC for me at first.
I finally jumped to using iterm and then using pycharm separately to do code review, visual git workflows, some run config etc.
But the actual value of Pycharm—-and I’ve been a real booster of that IDE has shrank due to CC and moving out of the built in terminal is a threat to usage of the product for me.
If the plugin offered some big value I might stick with it but I’m not sure what they could even do.
#1 improvement for VS Code users is giving the agent MCP tools to get diagnostics from the editor LSPs. Saves a tremendous amount of time having the agent run and rerun linting commands.
Does anyone have a comparison between this and OpenAI Codex? I find OpenAI's thing really good actually (vastly better workflow that Windsurf). Maybe I am missing out however.
Codex CLI is very bad, it often struggles to even find the file and goes on a rampage inside the home directory trying to find the file and commenting on random folders. Using o3/o4-mini in Aider is decent though.
> What does Cursor/Windsurf offer over VS Code + CC?
Cursor's @Docs is still unparalleled and no MCP server for documentation fetching even comes close. That is the only reason why I still use Cursor, sometimes I have esoteric packages that must be used in my code and other IDEs will simply hallucinate due to not having such a robust docs feature, if any, which is useless to me, and I believe Claude Code also falls into that bucket.
> Cursor's @Docs is still unparalleled and no MCP server for documentation
I strongly disagree. It will put the wrong doc snippets into context 99% of the time. If the docs are slightly long then forget it, it’ll be even worse.
What packages do you use it for? I honestly never had that issue, it's very good in my use cases to find some specific function to call or to figure out some specific syntax.
just curious because I'm inexperienced with all the latest tools here
> - Tab completion model (Cursor's remaining moat)
What is that? I have Gemini Code Assist installed in VSCode and I'm getting tab completion. (yes, LLM based tab completion)
Which, as an aside I find useful when it works but also often extremely confusing to read. Like say in C++ I type
int myVar = 123
The editor might show
int myVar = 123;
And it's nearly impossible to tell that I didn't enter that `;` so I move on to the next line instead of pressing tab only to find the `;` wasn't really there. That's also probably an easy example. Literally it feels like 1 of 6 lines I type I can't tell what is actually in the file and what is being suggested. Any tips? Maybe I just need to set some special background color for text being suggested.
and PS: that tiny example is not an example of a great tab completion. A better one is when I start editing 1 of 10 similar lines, I edit the first one, it sees the pattern and auto does the other 9. Can also do the "type a comment and it fills in the code" thing. Just trying to be clear I'm getting LLM tab completion and not using Cursor
It gets even worse when all three of IntelliSense, AI completion, and the human are all vying for control of the input. This can be very frustrating at times.
I use Windsurf so I remain in the driver's seat. Using AI coding tools too much feels like brain rot where I can't think sharply anymore. Having auto complete guess my next edit as I'm typing is great because I still retain all the control over the code base. There's never any blocks of code that I can't be bothered to look at, because I wrote everything still.
I often use the same setup. Qwen 2.5 coder is very good on its own, but my Emacs setup doesn’t also use web search when that would be appropriate. I have separately been experimenting with the Perplexity Sonar APIs that combine models and search, but I don’t have that integrated with my Emacs and Qwen setup - and that automatic integration would be very difficult to do well! If I could ‘automatically’ use a local Qwen, or other model, and fall back to using a paid service like Perplexity or Gemini grounding APIs just when needed that would be fine indeed.
I am thinking about a new setup as I write this: in Emacs, I explicitly choose a local Ollama model or a paid API like Gemini or OpenAI, so I should just make calling Perplexity Sonar APIs another manual choice. (Currently I only use Perplexity from Python scripts.)
If I owned a company, I would frequently evaluate privacy and security aspects of using commercial APIs. Using Ollama solves that.
I agree it has a good chance of catching up, but the difference in quality is pretty noticeable today. I'd much rather stick with vscode, because I hate all the subtle ways Cursor changes the UI; like taking over the keyboard shortcut for clearing the scrollback in the terminal. But I find it's pretty hard to use Copilot's tab completion after using Cursor for a while.
Windsurf big claim to fame was that you could run their model in airgap and they said they did not train on GPL code. This was an option available for Enterprise customers until they took it away recently to prevent self hosting
I think CC is just far more useful; I use it for literally everything and without MCP (except puppeteer sometimes) as it just writes python/bash scripts to do that far better than all those hacked together MCP garbage bins. It controls my computer & writes code. It made me better as well as now I actually write code, including GUI/web apps, that's are always fully scriptable. It helps me, but it definitely helps CC; it can just interrogate/test everything I make without puppeteer (or other web browser control, which is always brittle as hell).
CC would explode even further if they had official Team/Enterprise plan (likely in the work, Claude Code Waffle flag), and worked on Windows without WSL (supposedly pretty easy to fix, they just didn't bother). Cursor learnt the % of Windows user was really high when they started looking, even before they really supported it.
They're likely artificially holding it back either because its a loss leader they want to use a very specific way, or because they're planning the next big boom/launch (maybe with a new model to build hype?).
Cursor's multi-file tab completion and multi-file diff experience are worth $20 easily IMO.
I truly do not understand people's affinity for a CLI interface for coding agents. Scriptability I understand, but surely we could agree that CC with Cursor's UX would be superior to CC's terminal alone, right? That's why CC is pushing IDE integration -- they're just not there yet.
I can't stand the UX, or VS Code's UX in general. I vastly prefer having CC open in a terminal alongside neovim. CC is fully capable of opening diffs in neovim or otherwise completely controlling neovim by talking to its socket.
Fair enough. I guess a better way to put it is: for people who like Cursor's UX, but prefer Claude Code's performance as an agent, the combination of both would be the true killer app. Having to choose between these feels like a temporary gap in the evolution of these tools, and I'm ready for us to get past it.
I sympathize with your feeling that there is a "gap", but I'm fairly certain that both my ideal workflow and your ideal workflow are unlikely to be anything more than evolutionary dead ends, like early automobiles that inherited the shape of horse-drawn carriages.
I don't know where the evolution of coding agents will take us in the next couple of years, but I would not be shocked if it looks more like a GitHub issue/PR tracker than a code+chat interface with autocomplete, etc. I'm already noticing the I'm starting to rely on tux + multiple CC instances with independent work trees instead of babysitting each proposed change.
I strongly agree with you.
I’m more of a CLI guy, and Claude Code just works. Most good projects have a CLI anyway (gcloud, GitHub CLI, Vercel, etc.). I prefer CLI vs MCP’s.
I’m on the $200 plan, and it’s absolutely worth it (never thought I’d say this for a CLI app).
I don’t see how there will be any money to be made in this industry once these models are quantized and all local. It’s going to be one of the most painful bubble deflations we have ever seen and the biggest success of open source in our lifetimes.
The forked IDE thing I don't understand either, but...
During the evaluation at a previous job, we found that windsurf is waaaay better than anything else. They were expensive (to train on our source code directly) but the solution the offered was outperforming others.
A lot of engineers underestimate the learning curve required to jump from IDE to terminal. Multiple generations of engineers were raised on IDEs. It's really hard to break that mental model.
We are working to resolve this. It's still in preview but expect to see healthy Gemini 2.5 Pro allocations for subscription customers in the near future.
Wait a minute, have you often run out of the gemini cli free daily quota? Their free quota is very generous because they are trying to get market/mind share.
OK, thanks, I understand now what is happening. I use gemini-cli for one specific task at a time and sometimes just one 5 to 10 minute session a day. If I use long work sessions then I will add my API key and pay.
Claude Code is totally different paradigm. You don't edit your files directly so there is no tab autocomplete. It's a chat session.
There are IDE integrations where you can run it in a terminal session while perusing the files through your IDE, but it's not powering any autocomplete there AFAIK.
Yes or running claude code in the cursor/vscode terminal and watching the files change and then reviewing in IDE. I often like to be able to see an entire file when reviewing a diff, rather than just the lines that changed. Plus it's nice to have go-to-definition when reviewing.
Yes, it shows you the file diff. But generally, the workflow is that you git commit a checkpoint, then let it make all the changes it wants freely, then in your IDE, review what has changed since previous commit, iterate the prompts/make your own adjustments to the code, and when you like it, git commit.
Depending on what I'm doing with it I have 3 modes:
Trivial/easy stuff - let it make a PR at the end and review in GitHub. It rarely gets this stuff wrong IME or does anything stupid.
Moderately complex stuff - let it code away, review/test it in my IDE and make any changes myself and tell claude what I've changed (and get it to do a quick review of my code)
Complex stuff - watch it like a hawk as it is thinking and interrupt it constantly asking questions/telling it what to do, then review in my IDE.
Apparently they are, which is crazy to me. Zed agent mode shows modified hunks and you can accept/reject them individually. I can't imagine doing it all through the CLI, it seems extremely primitive.
As far as I can tell, terminal agents are inferior to hosted agents in sandboxed/imaged environments when it comes to concurrent execution and far inferior to assisted ide in terms of UX so what exactly is the point?. The "UI niceties" is the whole point of using cursor and somehow, everyone else sucks at it.
Done. Now you have a SOTA agentic AI with pretty forgiving usage limits up and running immediately. This is why it's capturing developer mindshare. The simplicity of getting up and going with it is a selling point.
Plus it’s straightforward to make Claude Code run agents in parallel/background just like Codex and Cursor, in local sandboxes: https://github.com/dagger/container-use
You’re missing the point tho. The point of the cli agent is that it’s a building block to put this thing everywhere. Look at CCs GitHub plugin, it’s great
CC on github just looks like Codex. I see your point, but it seems like all the big players basically have a CLI agent and most of them think that its just an implementation detail so they dont expose it.
It's another Character.ai situation [0]. Unfortunate for any employees who aren't founders or researchers, as they don't get any payout or a nice new job from this exit structure. In fact they lose their whole time invested at the company.
What a harsh time to work for an AI startup as a rank and file employee! I wonder how the founders justify going along with it inside their mind.
Edit: Thank you @jonny_eh for the clarification. I can't imagine it feels awesome being a leftover but at least you vested out. "Take the money and leave" is still a bit raw when the founders and researchers are now getting the initial payout + generous Google RSU's.
2.5x in 10 months. With returns like that - if I were the full time chef of the investors spare private jet I would be updating my CV and looking for a new gig.
Honestly depends on when they got in. Seed investors? They're probably fine with their preferences. Series B and beyond? That's where it gets messy. What round you thinking?
You've reminded me of when I first watched Idiocracy in 2006. At the time, I delighted in the comedic, sophomoreish, and seemingly ridiculous take on a possible trajectory of humanity. But now much of it is actually coming to pass. It's sad.
P.s. As a sidenote, apparently I love all of Mike Judge's productions, which also includes Office Space, and Beavis and Butthead.
Character.ai reached out to me for an opportunity, but they've already been carved up.
I think it's great that the rank and file got some of their equity cash-out (based on the other comment), but I imagine it isn't an attractive prospect as a start-up to join at this point.
I just ignored the recruiter. I can't imagine their would be a second liquidity event.
> Unfortunate for any employees who aren't founders or researchers, as they don't get any payout or a nice new job from this exit structure. In fact they lose their whole time invested at the company.
Windsurf’s value didn’t go to $0 overnight. The company will continue and their equity is likely still worth a decent amount wherever the company ends up.
Obviously a disappointing outcome for the people who thought life changing money was right around the corner, but they didn’t lose everything.
Just like with Character I'm assuming the employees get something. Whatever nonsense "licensing" fee Google is paying to not cause an antitrust investigation should be paid out straight to employees
The general character of capital markets is to pay as little as possible. Otherwise you lose out to those who are more ruthless. It is plausible that Windsurf employees really are getting very little value for their work. We need to see details of the deal.
Isn't there not some contractual agreement between the VCs and the founders? (I understand that a non-compete might not apply [in CA], but taking VC money is a little different that simply getting hired).
Were I a Windsurf investor, I'd be pissed right now and calling my lawyer.
the founder is on a vesting schedule set with the vc. walking away forfeits his ownership in the company (not sure of the specifics of this weird deal, but this is true in 99% of situations) which returns his ownership to the VCs either directly or functionally.
the only reason he'd walk away is because he thinks other opportunities are higher EV. if he believes this, a) the investors investment is likely worth virtually 0 anyway and b) if it's not, removing a leader who doesn't want to be there probably increases P(success) for the company and further increases the value of the investment.
founder departure isn't good for the narrative, but it's a symptom of an investment going bad, not often a cause.
Presumably the founder(s) is/are getting a better deal by walking away in this case. If they've been through a few round of funding, they may have been diluted to the point when this sort of exit is better (for them).
I went from Emacs to VS Code, then to Cursor, next to Claude Code, which is so good that I feel like I am having half a dozen junior devs at my fingertips, 24/7.
Since Claude Code is cli based, I reviewed my cli toolset: Migrated from iTerm2 to Ghostty and Tmux, from Cursor to NeoVim (my God is it good!).
Just had a 14h workday with this tooling. It’s so good that I complete the work of weeks and months within days! Absolutely beast.
At this point I am thinking IDEs do not reflect the changing reality of software development. They are designed for navigating project folders, writing / changing files. But I don’t review files that much anymore. I rather write prompts, watch Claude Code create a plan, implement it, even write meaningful commit messages.
Yes I can navigate the project with neovim, yes I can make commits in git and in lazygit, but my task is best spent in designing, planning, prompting, reviewing and testing.
I work at a company with over 700 employees. And there are tons of use cases where a simple CRUD app is sufficient. Or where glue code needs to be written / changed for legacy systems. Or where an OS system like Camunda is deployed and needs to be configured, workflows developed, etc
The reality of companies out there is much simpler than the challenges of a startup that needs to build systems that are state of the art, scale for millions of users, etc There are companies out there that make millions, in areas you‘ve never heard of, and their core business does not depend on software development best practices.
In our company we have an IT team with the median age of fifty, team members who never have developed software, just maintain systems, delegate hard work to expensive consultants.
Now in that setting someone coming from a startup background is like someone coming from the future. I feel like a wizard who can solve problems in days, instead of weeks or months waiting for a consultant to solve.
Fair enough. There are valid use cases for vibe coding scripts and simple CRUD apps, which current AI tools are fairly competent at producing.
The thing is that those don't typically take weeks and months to build with conventional tooling. And I find it hard to believe that all you're doing is this type of integration work. But I suppose there are companies that need such roles.
> There are companies out there that make millions, in areas you‘ve never heard of, and their core business does not depend on software development best practices.
That is true.
I do think that this cowboy coding approach is doing these companies a disservice, especially where tech is not their main product. It's only creating more operational risk that on-call and support staff have to deal with, and producing more technical debt that some poor soul will inevitably have to resolve one day. That is, it all appears to work until one edge case out of thousands brings down the entire system. Which could all be mitigated, if not avoided, by taking the time to understand the system and by following standard software development processes, even if it does take longer to implement.
What you describe isn't new. This approach has existed long before the current wave of AI tooling. But AI tools make the problem worse by making it easier to ship code quickly without following any software development practices that ensure the software is robust and reliable.
So, it's great that you're enjoying these tools. But I would suggest you adopt a more measured approach and work closely with those senior and junior engineers, instead of feeling like a wizard from the future.
> I do think that this cowboy coding approach is doing these companies a disservice, especially where tech is not their main product. It's only creating more operational risk
I absolutely agree with you! Alas, management in those companies (I have worked for some) does not understand the necessity of solid software engineering work. It’s way above their head. That’s why they have systems that constantly cause problems. And then expensive consultants are called in to fix symptoms, not solve root causes. Shocking to see that in lots of otherwise very successful companies.
Who's reviewing all the code you are churning out with ai? If everyone is used to maintaining not developing software it doesn't sound like they'd be best suited to have to review lots of complex pull requests.
It sounds like you are moving very fast and probably have people just clicking "approve".
Good luck for the future to who ever owns your company!
When I setup systems, I thoroughly document them, test them, develop them according to architectural best practices. My AI assisted code generation is lightyears ahead of what I see in companies I have worked for. The best they —the companies—-do is hire expensive consultants. Who sell them preconfigured system. And when you look into those systems you won’t believe your eyes either. Because you instantly realise that those devs do not know much about architectural patterns, aout systems design, about software development best practices. Yet they sell their systems as well, because they offer a niche product where they have only a handful competitors.
In that setting someone with solid software engineering background using AI to solve problems is like a wizard from the team‘s perspective.
When I worked for startups I was constantly panicking to miss the latest tech trends, and I feared that I would be not marketable in case I didn’t catch up. But in mature companies things work much slower. They work with decades old technology. In that setting not the latest tech counts but being able to solve problems, with whatever means you can.
I don’t review every single line that AI generates. I glance over the files to see if they meet my standards, prompt it to rewrite this or that portion, when necessary. Or change it myself.
Writing code is the most tedious part, not reviewing.
Personally I've had mixed experience when I let Sonnet 3.7 document my (and its) code and write commit messages, for some wip stuff it's alright but it soon gets out of hand and because it doesn't really have a direct view in my mind it ends up documenting what it sees instead of the intention behind it, which is totally fair but eh.
So yeah, mileage varies and agentic tools usually spit out more code and redundant comments than I'd like to review. I'm still waiting for a company to develop some sanity check for this somehow but snapshot testing and manual review aren't enough sadly.
Man, who sucked the joy out of your life. Just try the damn thing. I have the staunchest anti-hypsters in my org and even they are using these tools heavily now.
I build most of not all of my stuff for work, and I ain't sharing that.
It's no panacea, but is there something to be had there? Abso-fucking-lutley. All of this would have been complete scifi at the beginning of this decade.
I’m super pro AI. I’ve been using ChatGPT since the day it released. I use an agent coder at work semi-regularly to reasonable levels of success. Big fan.
But I am exceedingly tired of phrases like “complete the work of weeks and months within days”. If AI is making devs 5x to 10x faster then I’d like to see some actual results. Internet is full of hypesters that make bombastic claims of productivity but never actually shown anything they’ve made.
Not the OP but this is smth I've vibecoded using cursor: https://bestphoto.ai/ MRR ~$150. It basically started as a clone of my other site: https://aieasypic.com (MRR 2.5k, 5-8k/mo rev) since I was having trouble keeping code context in mind and claude was pretty bad at doing full features with the tech stack I used for that site(Django BE, NextJS FE) making adding new features a pain, so I completely switched to a stack that claude is very good at NextJS fullstack(trpc BE) and now it can basically one-shot a feature request.
Just putting this here because a lot of times AI coding seems to be dismissed as smth that can't do actual work ie generate revenue, while its more like making money as a solo dev is already pretty rare and if you're working in a corp. instead you're not going to just post your company name when asked for examples on what you're using AI for.
Those are exactly the kind of AI slop products I would expect to be vibe coded. You've created yet another wrapper around LLM APIs where the business model is charging a premium over existing services. Your revenue depends on the ignorance of customers to not realize they can get the same or better service for cheaper from companies that actually do the hard work. I bet SEO hacking is really important to you.
It's irrefutable that AI tools can be used to create software that generates revenue. What's more difficult is using them to create something that brings actual value into the world.
I'm not arguing in favor of technical complexity or ingenuity.
My point is that the perceived value of a service or product is directly related to its competitive advantage, product differentiation, and so on. When the service is made from the same cookie cutter template as all the others, the only value that can be extracted from it is by duping customers who don't know better.
There are entire industries flooded with cheap and poorly made crap from companies that change brand names every week. Code generation tools have now enabled such grifters to profit from software as well.
Eh, it is more like an extended/better UI. Plenty of people are willing to pay for just that.
There are lots of people that only use LLMs in whatever UI the model companies are providing. I have colleagues that will never venture outside the ChatGPT website, even though with some effort they could make their tooling richer by using the API and building some wrapper or UI for it.
Sure man, any product you don't like is just "another wrapper". I guess every website is just a wrapper over postgres or wordpress too. I run my own serverless GPU containers on runpod with a combination of comfy and my own fastapi servers using diffusers, not that it'd even matter if I just used some third party APIs. It originally even started as smth that was hacked together using 4x 4070ti supers in my basement that I then moved to runpod. Indiehacking is mostly marketing, nobody cares if you built some technically beautiful thing.
Also its easy to criticize from the sidelines but, do you have products that you made by yourself that are used by hundreds of thousands of people? I have 5 such sites, 2 of which I named above
Hey, don't blame me for the fact that your sites are indistinguishable from hundreds of others that offer the same service. Everything I said is logical to assume, since all these sites look the same.
Good on you for learning how AI tools work, but there's no way for anyone to tell whether your backend is self-managed or not, and practically it doesn't really matter. I reckon your users would get better results from proprietary models that expose an API than self-hosted open source ones, but then your revenue would probably be lower.
> Also its easy to criticize from the sidelines but, do you have products that you made by yourself that are used by hundreds of thousands of people? I have 5 such sites, 2 of which I named above
That's a lazy defense considering anyone is free to criticize anyone else's work, especially if they're familiar with the industry. Just like food and film critics don't need to be chefs and movie producers.
But I'll give you credit for actually building and launching something that generates revenue. I admit that that is more than I have managed with my personal projects.
AI is often used to pump out sites and apps that scam users, SEO spam, etc. So there is definitely a revenue stream that makes scammers and grifters excited for AI. These tools have increased the scope and reach of their scams, and provide a huge boost to their productivity.
That's partly why I'm curious about OP's work. Nobody who's using these tools while following best software engineering practices would claim that they're making them that much more productive. Reviewing the generated code and fixing issues counteracts whatever time is saved by generating code. But if they're vibe coding and don't even look at the code...
I review PRs/commits, not files. Given the right cage to lock the agent inside, and guardrails built around, and conventions and guidelines, and agentic flows so it can pull in what's needed.. the need to look at every line and file during implementation is significantly lessened. So then I review the final output (which is a unit of work/task wrapped in a PR).
yes vibecoding is addicting like that. but if you are not reviewing any code and simply vibing then
in my expreience you'll eventually get stuck in "its still not working" loops beause you have no other context or insight to provide it other than that. Then you have either accept what you have or throw the whole thing out and/or actually read the code . kind of rules out last option because code is now just too far gone with too many special cases hardcoded because AI sucks at abstraction or real software engineering.
At work we’re encouraged to use AI, so I do. For me the one thing that works well is using it to write one off scripts that do stuff and would be a chore to write.
Usually in 2-3 prompts I can get a python or shell script that reads some file list somewhere, reads some json/csv elsewhere. Combines it in various ways and spits out some output to be ingested by some other pipeline.
I just test this code if it works it’s good.
Never in my life would I put this in a critical system though. When I review these files they are full of tiny errors that would blow up in spectacular manner if the input was slightly off somewhere.
It’s good for what it is. But I’m honestly afraid of production code being vibe coded by these tools.
Ghostty is gpu accelerated. It’s super fast, and tmux in it is a joy to use. That combined with NeoVim gives me an increadibly smooth dev experience in the terminal, something I had never have with iTerm2 and emacs.
Haha :) I lived inside Emacs, used orgmode for everything, have written tons of Elisp, used org-roam as my second brain, used vanilla Emacs shortcuts instead of Evil (with a special keyboard settup using Karabiner Elements), did even my googling from Emacs, used emacs calc instead of my calculator, but in the end I spent more time tinkering my Emacs setup than doing real work. Emacs was a lifestyle. At some point I realized: Unix and the terminal are what Emacs try to be: It tries to be a one-stop shop offering you everything: Surfing the web, writing emails, word processor, calculator, planner, terminal. Unix and the terminal offer me all of that. Plus any scripting language. Why miss all the beautiful apps, just to be an Emacs zealot? The editor in emacs is just one usecase. Neovim does it just as well, if not better.
But relax, noone is taking your Emacs from you :) I still like it, but am not a disciple anymore ;)
Well you can't risk Claude quitting overnight. It forgets everything it did the day before and now you have to start over ... must ... finish ... tonight ... within ... context ... window.
Fortunately LLMs are stateless thus not affected by passage of time - your context stays exactly as it was while the tool maintaining it is running.
(Prompt caches are another thing; leaving it for the night and resuming the next day will cost you a little extra on resume, if you're using models via API pay-as-you-go billing.)
Cursor (and Garry Tan’s X post) has shown us that the VC money is propping up these companies astounding growth, the only way for them to become profitable is to increase the cost per a request, which means they need to innovate like crazy.
The moat is paper thin.
GitHub has open sourced copilot.
The open source community is working hard on their own projects.
No doubt Cursor is moving fast to create amazing innovations, but if the competition only focuses on thin wrappers they are not worth the billion dollar valuations.
I love watching this space as it is moving extremely fast.
There is no moat. If you’re a true believer that strong agents are around the corner, then all of these add on companies will be obsolete in a few years. The first company to strong agents can trivially rebuild Cursor or Windsurf.
If AGI is around the corner, I don't believe one single company will "own" that tech. It will be like it is today, where you have multiple models competing.
And after that, AGI will be open source.
In the end, ownership of data and compute will be the things that define the victors.
That would be true if the product was the goal. In my experience,
marketing > market > product
Even with AGI in hand, there will still be competition between offerings based on externalities, inertia, or battle-testedness, or authority. Maybe super-intelligence would change the calculus of that, but you'd still probably find opportunities beyond just letting your pool of agents vibe code it.
If you believe AGI is round the corner (I don't), then you face the dilemma of investing in a company that will be the one profitable survivor in a disintegrating world.
This would only be true if it was cheap to run and would return results quickly. If AGI only has compute to serve 1 customer per hour then their is an upper bound of market share it can take from other products.
We don’t eat intelligence: bikes, bottles, food, energy… have room for improvement but I hardly see how AGI would replace them.
Same for physical services like labors, miners and cooks, even taxi/bus drivers for +99% of the world. Automation immensely improve their efficiency and the Modern Times is the past for half of the globe, but AGI isn’t the main facilitator.
Replace all (most*) Silicon Valley -and cousins- similar "products" and services, perhaps yes !
You don’t see how robots would replace all these roles? I take it you never seen a Waymo or read about how agriculture needs a tiny fraction of humans in the US vs a couple of decades ago? And those aren’t even “AGI-level” solutions…
Waymo aren’t build, maintained and disposed by robots only. Even if they were we’ve still to see how that can be deployed (and make sense economically) to more than a couple percentiles of the richest people in the top cities. That has nothing in common with "replacing the humans".
Agriculture efficiency mainly comes from last century fertilizer. The modern robot trucks are also limited to some richest countries and farmers and have nothing to do with AGI.
Your vague dream of AGI solutionism missed the barriers to technologie deployment: raw ressources availability, politics, demography, energy… IA can have a small impact there but it certainly won’t replace ALL products. How would a computer replace a mining truck or a reactor turbine?
If I missed your point you’d be wonderful to clarify it, perhaps by detailing the first post "replacing all products":
- what do you consider a product?
- ALL as in Everything (obviously not I guess)?
Regarding "don’t you see how robots…" I clearly see the theory but also can’t see how it would happen in reality for the reasons I mentions in my precedent post: mainly access to ressources/energy to build/maintain those robots for a signifiant time scale and socio-politico barriers.
I’d be glad you reconsider explaining your point because I only see vague allusions, but if you want to stop the conversation here that’s all up to you.
I think the recent Grok release and considering xAI was relatively late to the game shows that the only moat to training giant models is how many GPUs you can buy. ChatGPT was earth-shattering and it took less than two years for multiple credible competitors to match or exceed them. Making these models profitable is proving extremely difficult in the face of so much competition and such unsustainable expectations being set. Google seems to be most likely to sustain themselves through this melee. Them and the Chinese companies.
Cursor just committed mass consumer fraud at worst, and at best pissed off all their best customers. I feel really sorry for those who invested at a 9bb valuation.
flip-flopping on pricing has led users to feel nickel-and-dimed
i like cursor fine, but check out the forum/subreddit to see people talking like addicts, pissed their fix is getting more expensive
i think this aggressive reaction is more pronounced for non-programmers who are making things for the first time. they tasted a new power and they don't want it taken away.
I agree with your take, but I still don't excuse anti consumer practices like that. It annoys me because this is a repeat problem in this space, where these companies don't take into account the market dynamics, or costs of their service. From the start I've been looking at these $20.00 subscriptions, and then my own personal api per token costs and been wondering how they aren't all bankrupt.
Look no further than founders in the sports betting space, like the fanduel founders. Borrow a bunch of money at huge valuations because of hype and ignore the fact, that despite it being exciting and popular, the margins are like <5%. Fanduel founders sold for 400 something million, walked away with nothing. Its now a multibillion dollar company when the new owners realized the product was marketing, not the vig. These AI companies are shifting towards their "marketing" eras.
I think any "value add" business that has a primary product built on top of another larger business' non-commodity service(s) runs the risk of having to re-do their pricing in ways that are outside their control.
This is nothing new. I'm not sure if it's "anti-consumer" as much as it's just a risky play from a brand and customer happiness viewpoint. Because your prices can be forced up by your supplier, and your customers will be mad at you, not at your supplier.
I do also think it is on consumers - in some part - to go into it with eyes open and do their research.
Thankfully a product like Cursor is a monthly sub and not a big up-front investment so if you don't like - or can't afford - the new pricing, you can just stop paying.
I think my primary concern are the customers who paid $220 for a yearly subscription, only for it to get nerfed and asked to pay $200.00 a month to get similar output.
What surprises me is just how much they've missed the mark.
I'm not an extreme user of Cursor. It has become an essential part of my workflow, but I also probably on the lower/medium section of users. I know that a lot of my friends were spending $XXX amounts/month on extra usage with them, while I've never gone beyond 50% included premium credits usage.
After their changes I'm getting hit with throttling multiple times a day, which likely means that the same thing happens to almost every Cursor user. So that means one or more of:
- They are jacking up the prices, to squeeze out more profit, so it looks good in the VC game
- They had to jack up the prices, so that they aren't running at a loss anymore (that would be a bad indicator regarding profitability for the whole field)
- They are really incompetent about simulating/estimating the impact of their pricing decisions, which also isn't a good future indicator for their customers
My guess is that it's your second scenario there (avoid running at a loss). In the start-up game scale/growth is the most important thing and profits really aren't that important. you want to show to later stage VCs that your idea has traction and there's a large addressable market.
Whilst profits aren't important you also can't burn all your current capital, so if the burn rate gets too high you have to put up prices, which seems likely to be what Cursor is doing.
If the people are still paying the increased prices, it's a success. Rugpulling literal addicts is a great business model. Remember that business profits are primary, not consumer opinion.
I think what everyone, including those programmers advocating coding agents, are forgetting is that if you can have a full time programmer for $200/m, then that becomes the new value of programming labour in the open market.
IOW, the market will slowly but surely drive the labour rate for programming down to the cost of the cheapest coding agent.
So, sure, boasting about a 10x speedo on boilerplate has good metrics, but let's not delude ourselves that programmers are going to be paid enough to afford the $200/m coding agent in the future.
I seriously cannot keep up. I fell a bit behind and now I feel I need a primer to know who owns/acquired/developed all these additional things surrounding the ai space
Until something else comes around for similar price and is much better.
Good thing for consumers who use AI coding tools is that there is no lock-in like in Photoshop or similar software where you hone your skills for years to use particular tool. Switching from Cursor to any other platform would literally take 10 minutes.
His reaction seems entirely appropriate. He could ignore you, but then you might keep replying to his posts and potentially spread incorrect but damaging information. He is losing close to zero by blocking you, but preventing a potential big loss. Why did you make that remark, if not to damage YC's reputation? This does not seem like the correct approach, if you wanted to improve their selection process.
The hilarious part is that I never interacted with him directly at all. I was just commenting something to a mutual on X, the thread blew up, he snooped it, and went on a blocking spree. It may have been different if I were directly accosting him in replies to his posts.
> Why did you make that remark, if not to damage YC's reputation?
Seems harsh and cultish to assume malice. He didnt say you parents have false credentials
I would say calling out people and institutions like that is important so as to keep them honest, and if they arent honest and are trying to grift/defraud people then they deserve the reputation loss
> He is losing close to zero by blocking you, but preventing a potential big loss.
Thats great for gary, but the rest of the world isnt there waiting to be optimized for his benefit. If people trust YC to incubate good talent, but feel its becoming a hub for grifters, then some accountability is in order. Institutions are beholden to their public stakeholders, even private institutions, because they still have people who are using and supporting them
This whole situation feels shockingly close to the Meta/Scale situation, where founders and specific employees were plucked out, and effectively gutted any future prospects for the company.
At least in the Scale case there seemed to be some form of payout to employees and equity holders, but this takes it a whole lot further by just throwing out all other employees.
There is supposed to be the concept that “all common stock is the same”. These fake-acquisitions completely undermine that.
Yep, if investors and early employees keep getting left out in the cold while execs get a soft landing at Big Tech, it's going to shake a lot of trust in the startup game
I don’t think anyone trusts any tech company much these days. It’s been a steep decline in the past 5 years, from arbitrary mandates to the constant talk about firing everyone and hiring an AI. Even as an investor, it’s hard to trust that the “honor system” that once existed is still in play.
So Google, Meta, and Microsoft will just hollow out the best AI startups of their talent instead of buying them - out of fear of monopoly lawsuits I'm assuming?
Nice plan I guess. Kind of obvious to spot though.
I actually don't know if there's much that can be done unless there's some non-competes in those employees' contracts which are usually not very enforceable outside of finance iirc.
The whole point of funding a company is for the company to build IP that makes the company valuable. Founders can't take investor money and then just go start another company -- that's specifically barred in most docs. There have been a lot of these weird "loopholes" lately that are completely against the spirit of company building.
I’m honestly just surprised that the CEO and co-founder decided to walk away from the company and leave behind all these employees he was leading. Especially considering many of them probably joined for lower pay, hoping for a big upside.
Yes, startups are always a bit of a gamble, but this feels like a captain abandoning ship while it’s still full of sailors (many of whom have families depending on them).
This really is a whole new level of getting screwed.
This is why advice is always to treat options for a non-public company as if they're near zero in value.
Because for most people, they will end up being worth exactly zero in value. Less if they went and exercised those options prior to a liquidity event that may never happen.
“Buying the startup” just means handing over megabucks to do-nothing investors. If Google isn’t buying any product or technology, why should investors get a talent fee?
Do nothing investors who enabled the company to reach this point? Employees who chose lower salaries in expectation of shares being worth something? Come on now.
Sure if you want to be negative about it and only look at the worst VCs. But the best VCs provide significant value outside capital and can be instrumental in a startups success or failure.
This is overly reductionist. The are plenty of laws that work well.
Any time I hear someone talk about more or less regulation, instead of talking about better or worse regulation, I suspect they are ideologists and trying to shift the narrative, or else they would be able to criticise based on actual merit.
There are many AI startups and we are just in the beginning of learning how to use them. There will be some stupid company like those you’ve listed that figures out a way to use AI that is far better than any other implementation, and Google, Meta, and Microsoft may go the way of Yahoo and AOL, but we’ll see
The “talent” is not very talented, trust me. These are the short term whims of very large, increasingly bloated organizations. A leaner startup that knows what it has will not sell so quickly. At least, the odds will soon be in favor of whoever first decides to take that bet.
I never knew anyone who used Windsurf. These AI acquisitions have been unbelievable(in a bad way). WIX acquired some garbage Lovable.dev clone for 80 million. I think many of us are waiting for this bubble to pop(economy will likely pop too)
It was barely better than Cursor and they got shafted by Anthropic because of the takeover announcement so nobody really used it anymore because let's face it - Claude Sonnet is just the best coding model. Design-wise the chat panel and autocomplete integration was a bit nicer than in Cursor but not by much. Subscription for Windsurf was/is also 5$ cheaper.
i don't think it was better than or comparable to cursor at all. except for the month prior to the OpenAI Acquisition news where some minor influencers on X were calling it better.
Everyone has a niche, Windsurf is the only large provider if you are a Jetbrains shop.
There are some alternatives like continue.dev or Jetbrains own AI offering but no Cursor or Claude Code ( Sonnet 3.7/4) you can get through Jetbrains plugin or others, but Anthropic does not provide support same with cursor.
I just tried it out. It does feel better than CC for some tasks .
I think what Junie does differently it has better context on what the IDE knows or sees , from what files are open or typedefs are loaded, I could achieve some simpler tasks with lesser prompts and less tokens and cycles than with CC or even windsurf.
Jetbrains has the best incentive to integrate deeply I imagine, it has always been question of execution, it seems a fairly robust system.
The biggest problem for me is that I can't link my Claude Pro (or Max) account to it. So now I'm paying for two LLM accounts (even if I used Junie for coding I use Claude for other things, research, etc., so wouldn't drop it).
Check out sweep. Completely unaffiliated, their only offering is the jetbrains plugin so it gets a lot more focus than windsurf. Only downside is that Claude code is still a better agent, but at least its tab complete is some of the best
The copilot agent stuff in IntelliJ works relatively well in my experience, they managed to implement a quite cursor-like “accept/reject” UI in a plugin, you know, forking IDEA. There are some areas like getting it to use git tools where cursor works more smoothly but you can coax Copilot into producing the same results. I’m just generally happier working in IntelliJ vs VSCode so I’ve tended to favour Copilot.
Never tried Windsurf in it’s recent form but we did evaluate it when it was still called Codeium and everyone liked Copilot better.
Windsurf was also used by enterprises because of their on-prem plan. They gutted that after OpenAI acquisition was announced and since then I am sure none of those enterprises that used it will switch to their cloud offering and look for other venues.
I used it for about six weeks in the spring, at which point I tried cursor and found out that it was far better for my use case (backend Go): better and faster autocomplete that can work across files, less buggy UI, and it actually follows your rules.
I’m not bothered by this news — it’d be pretty shortsighted to assume that any of these disposable AI assistants will continue to exist forever.
Base44 is absolutely not garbage. I’ve tried it and can say it’s as good or better of a vibe-builder than Lovable or Bolt. Have you benchmarked it against the competition or can you otherwise substantiate the “garbage” claim? FWIW I do know one amazing engineer using Windsurf
Agree in principle, but when evaluated against the competition and likely acquisition targets of Wix, it's certainly not garbage. I've seen it vibe code an entire app that was -- admittedly mostly working -- and deploy it with a prompt of 5 words, in about 2 minutes.
If by Documentary, you mean a new Silicon Valley sitcom, yes , all the ingredients are there: The AGI believers, the doomers, the "cure all diseases" people, the board drama, the money grabbers, the VC dance , the poaching, the lawsuits for copyrights ... there s a whole new universe of caricatures
Works out for Google and the C-suite. Horrible for the employees. These fake-acquisitions are effectively arbitrage against employees, who get left holding nothing. Should be illegal and regulated.
Not sure how the VCs get their cut. I'm guessing that Google can balance it out by participating in rounds for other startups in that VC's porfolio.
Cursor's Accept / Reject feature for each change it makes in each file is nice whereas I have to use a diff tool to review the changes in Claude Code.
Also, if I go down a prompt alley that's a dead end, Cursor has the Restore Checkpoint feature to get back to the original prompt and try a different path. With Claude Code, you had better have committed the code to git, otherwise you end up with a mess you didn't want.
My company pays for both, but I mostly use Cursor unless I know I am doing a new project or some proof of concept, which Claude Code might have an edge on with a more mature TODO list feature.
I got burned too many times from that Restore Checkpoint thing not working right, maybe it's been fixed by now but seems silly to rely on something thats not a literal tool built for the job (version control), not a good shortcut.
I was so surprised (or shocked) to hear that Windsurf was getting acquired for 3 billion dollars, I made an HN post asking about the truth of that news - https://news.ycombinator.com/item?id=43933825. HN's system didn't like my tone I guess and removed it, lol.
But in any case, I just can't see how AI code editors like Windsurf or Cursor, without any proprietary model, can be valued at billions. What's the underlying IP that justifies these valuations?
Similar thing to what we’ve witnessed with crypto coins.
It’s AI season and those with money invest in it, pump it and will exit post IPO.
Difference here is, that besides value that those products “hold”, it’s possible also to provide AI as a service, making Google / Microsoft etc interested.
> I just can't see how AI code editors like Windsurf or Cursor, without any proprietary model, can be valued at billions.
Windsurf used to be GPU rental company before they decided to make a text editor. I imagine they own a lot of silicon they are running their coding assistant on.
you meant Exafunction? Nothing of Exafunction was discussed in OpenAI or Google deals.
Windsurf did start as a rinky-dinky starup, just as a plugin (Codeium). And in its current state, it still has no moat.
The only reason big tech leaders are throwing money at this is - they need to show they are doing something to stay ahead. And right now they are going on a hiring frenzy as a show for their investors.
I have been there when shopping businesses were being bought for billions of dollars, and then they all got bundled up into a mess swapped under the rug silently. I don't think this is going to be any different.
It's really a simple story, they built infrastructure years ago, initially they rented it, so companies outsourced their computing tasks to them. But eventually came up with the idea what product could they put on top of their infrastructure to sell it for dollars rather than pennies. And they did.
They don't need a moat because they provide turnkey utility to a lot of very rich corporations. They are not in the business of selling the vs plug-in to you. They are in the business of selling developer experience powered by their own vast compute to a lot of very rich corporations.
They had millions in revenue before they hired their first sales rep. I don't think just a VS plug-in startup could do that.
Anyone can build a proprietary model in under a week; but the point is - Cursor and Windsurf don't have proprietary models that remove dependency on external models, and/or have capabilities that competitors don't have.
> Google will instead hire Windsurf CEO Varun Mohan, cofounder Douglas Chen, and some of Windsurf’s R&D employees and bring them onto the Google DeepMind team, [...] Google will not have any control over nor a stake in Windsurf, but it will take a non-exclusive license to some of Windsurf’s technology. [...] Google didn’t share how much it was paying to bring on the team. OpenAI was previously reported to be buying Windsurf for $3 billion.
Why not an acquisition?
How did Google get Windsurf and investors to agree to this maneuver that decapitated the leadership and key talent, without a big exit event for everyone?
My read of the article: "Here's x% of what OpenAI offered you, you waive legal challenges while we cherry-pick your people and license the tech in their heads, and you can keep the company, and everyone left behind can promote themselves to fill the vacancies."
I wonder whether the left-behind small-fry employees make out comparably to how they would with an IPO or big acquisition.
Or if the deal is more like: "We'll pay you double what you could make elsewhere for the 2 years we want you to Potemkin Village, or you can walk and get paid zero; we don't care, because enough people will stay."
I guess VCs can't force founders to stay (the only penalty for joining Google is loosing some/all of their Windsurf equity, but I'm sure they chose what's better for them), and employees didn't need to agree (they have no vote).
I'm not surprised. I started using Windsurf when it came out because I liked its UX better than Cursor's.
However, while Cursor and GH Copilot improved, Windsurf went in the opposite direction. On each update, I started to get more and more issues. The agent often tried to run shell commands, and it hung up, or I found minor UI bugs. One day, I decided to give GH Copilot another chance, and I was surprised by how it evolved, to the point that it worked better than Windsurf for my usage. I don’t know what happened internally at Windsurf, but I notice the degradation as a user. If my case indicates what happened to other users, maybe OpenAI saw declining subscriptions and canceled the deal.
It's unclear if OpenAI cancelled the deal, or Google poached them? Either way, this season of "OpenAI Drama" is wild. First Meta, now Google. Your turn Amazon / Microsoft.
Apparently OpenAI allowed the deal to expire; likely Google had already been in discussion with Windsurf as I'm sure they knew the deal was likely to die well before today.
> OpenAI’s deal to buy Windsurf is off, and Google will instead hire Windsurf CEO Varun Mohan, cofounder Douglas Chen, and some of Windsurf’s R&D employees and bring them onto the Google DeepMind team, Google and Windsurf announced Friday.
> Mohan and the Windsurf employees will focus on agentic coding efforts at Google DeepMind and work largely on Gemini. Google will not have any control over nor a stake in Windsurf, but it will take a non-exclusive license to some of Windsurf’s technology.
Sounds to me like they're "hiring" them like one "hires" a consultant?
I don't know anyone who heard or used Windsurf outside the Bay Area. Even Cursor feels very Bay Area bubbly (although that is the market to go after if you're in ai dev tools).
Cursor does add value but it's just a thin layer on top of VSCode so companies could just build that in-house and don't need to acquire. There's no moat there.
Cursor has custom tab and embedding models. And has a lot of distribution / paying users already.
Arguably they have the strongest product moat, and I wouldn’t be surprised if they beat OpenAI in a vertical coding model from that. Easy for them to have users generate evals and have model product feedback loop here.
The tab completion is fast and the best available right now but is still so garbage that I turn it off 99% of the time because the suggestions are mostly noise.
I have the opposite experience, it’s at least 90% correct. For example, if I start writing the name of a function that I just added in a different file, tab will suggest the function, then jump to the top of the file to import it. If I’m changing the way something is called in 5 places, if I change it in the first place, tab will jump to make the same change in the other places. It’s honestly pretty spot on.
Zed tab is a lot worse in comparison (partly because it’s slow)
> I have the opposite experience, it’s at least 90% correct. For example, if I start writing the name of a function that I just added in a different file, tab will suggest the function, then jump to the top of the file to import it
Intellisense has done this for ages. A decade at least.
I did not see this coming. Wow. The game of thrones in SV.
I wonder what happened with the OpenAI deal. Anyone have any guesses? My first guess is "Look at Claude Code, we can do this ourselves." But, I am likely thinking too simply.
edit: does this mean that Windsurf and its users will stop being iced-out by Anthropic? Or, is this the end of Windsurf?
I have been using Windsurf for few months. They even have their own AI model SWE-1 model. I really liked using Windsurf. They also have integrations with other IDEs ex: jetbrains, VS code, etc.
This week I have been using Claude Code and Windsurf side by side. I would make change with one, stash it, ask the other for similar change and then would diff it.
Overall Windsurf was pretty on a par with Claude code.
I'm a rank and file dev at a non-big tech company and I got a call from a Windsurf sales rep this week who I had connected with on LinkedIn the day before (I never gave them my number). They told me my company was in talks with Windsurf about a licensing deal but that they would give me a 30 day trial of an enterprise account for use on personal projects to let me try it in advance. I guess the idea for them is to build enthusiasm among devs in the company?
Is this a standard sales strategy for products like this? It seems pretty aggressive to me but I'm just an engineer so I wouldn't know.
Very standard yep. Sales folks are sort of trained /indoctrinated into telling white lies like that in order to get in the door. There are loads of examples of using fake momentum to close deals. If its a senior person it’s “My CEO asked me to personally reach out to you” or a fake email from the CEO forwarded by the rep. If one person at the company uses it, it’s “we’re negotiating a company wide license” or “we already have a group license with extra seats” or “one of your teammates sent us a list of priority teammates” yada yada.
Are these "acquihire & license" the new M&A...? I recall hearing that this was a "hack" to avoid DOJ and FTC scrutiny over acquisitions, but I have no clue how such deals are structured. Anyone care to chime in?
This deal always looked strange in the first place. The usage of Windsurf was significantly lower than Cursor and Copilot and somehow it was worth $3B.
Given the release of Claude Code, it was already over for them.
For the love of God, can we get a reboot of the Silicon Valley television show? Just on AI. Like when they wrapped it, they wrapped it on AI usage. So, it's got the perfect arc for a reboot that focuses perfectly on AI.
My favorite thing about that series was watching it with friends who weren't from the Bay Area. Often they'd be laughing at the sheer absurdity of a situation, and I'd get to point out that it was barely exaggerated from real life.
That's what my friends not from SF said. "This is insane, this would never happen"
Dude, I saw a lot crazier things happen on a monthly basis. And don't even get me started on the personal lives and partying that the show didn't display.
There’s a lot of talk about Claude Code in here, and I agree it’s a great agentic coding tool. One of the benefits of Cursor & Windsurf is/was the ease for smaller companies to setup Team accounts and have control over spend.
Claude Code I think misses this. You can get an enterprise account if you commit to over, what.. 70 seats annually?
If you’re an individual you can get Max 5x/20x ..
But for smaller companies, I don’t think they are addressing that space. Am I wrong? Are there any Agentic tools like Claude Code that can provide a fixed cost per user?
Kilo Code CEO here. We’d like to welcome ex-Windsurf users by offering you $100 in credits. :D We'd love to show how through open source there is a better way (better community, more transparent pricing, won't mess with your or the product). https://blog.kilocode.ai/p/windsurf-is-over-switch-to-open-s...
This certainly aligns with my own usage. I'm currently using OpenAI's own Codex 50:1 compared to Windsurf. For me, I'd rather take some time to create a good quality prompt and have it work away for a few minutes and create a material delta. It isn't always perfect, and I often have to make a few tweaks myself, but it is much nicer and waiting around and watching Windsurf bang around on a tiny part of the solution. Windsurf is still nice to use for quick UI iteration however.
There are probably some extremely unhappy VC's, angels and employees over at Windsurf right about now.
Windsurf's Funding Rounds
Windsurf has raised a total of $243M over 2 funding rounds. One was Series C round of $150M from Investors like General Catalyst and Kleiner Perkins and the other was Series B round of $65M from Investors like Kleiner Perkins and General Catalyst.
Here is the list of all funding rounds of Windsurf:
Round Details
Series C, $150M, Aug 29, 2024
Post money valuation
-
Revenue multiple
-
Investors
General Catalyst, Kleiner Perkins, Greenoaks
Round Details
Series B, $65M, Jan 30, 2024
Post money valuation
-
Revenue multiple
-
Investors
Kleiner Perkins, General Catalyst, Greenoaks
It's something you should never assume is true until the wire hits your account. I had a deal where I was going to make $15 million called off 36 hours before closing.
This my friends is how the next iteration of venture capital contract templates becomes even longer...
Otherwise, normally with the amount of capital raised by Windsurf, the founders must have signed some kind of non-compete for the event of a bad-leaver (which this obviously is). Guess covering these penalties was just part of Google's deal, hm?
Perhaps it as combination of how much founders were diluted and how much they are being offered upfront. We are hearing about $100M signing bonuses.
It is hard to say no when Google/Meta gives you say $100M upfront and hundreds more if not Billion+ in RSUs. After 3 rounds it is not unreasonable to have only 5-10%.
10% of a company worth a few billion burning a lot of cash, that needs to keep raising more rounds i.e more dilution, may have less value than RSUs from multi-trillion dollar publicly traded liquid tech company today.
It is also quite hard to raise $5-10+Billion in cash. There are only handful of startups which have ever done so
Very few funds/investors can afford to do so large rounds. This was SoftBank's thesis for most of last decade, compete by just outfunding competing products in a market.
The deal for the founders may not have been as good as what Google offered. They may only hold 10% after those rounds, a serious part of the acquisition price could go to liquidation preferences of the VCs and the deal is mostly in OpenAI stock instead of cash. Not that hard to imagine the Google option offering them much more actual cash right now.
So the result of aggressively scrutinizing big tech acquisitions is acquihires, not a more competitive tech ecosystem with say more IPO’s.
The libertarian spin on this would be government should have never scrutinized acquisitions and the result is just worse for everyone.
The progressive spin would be to now ban acquihires somehow, and then whatever new legal invention will be created next. I can imagine the next step being, creating a consulting company out of your startup and then selling yourself as consultants to big techs. Now you are neither acquired nor technically acqui-hired and the whackamole continues.
At some point, we need to realize the solution is the culture of people involved. If the government could just ask to reduce acquisitions to make the ecosystem more competitive and companies tried following it in spirit to the best of their ability, we might have much better results than whatever we have now. When culture degrades, the govt can’t trust companies, the companies can’t trust the govt, everything just gets worse, regardless of what rules you write and enforce.
The culture of the people involved got us to this point, I’m not sure it’s the solution to the problem.
> The progressive spin would be to now ban acquihires somehow, and then whatever new legal invention will be created next.
Progressive has become a moving target, but the pro-competition view would be to break up the massively concentrated companies that are further consolidating markets. Thats what the Khan FTC was trying to do, but we need a Congress interested in a competitive marketplace, which we haven’t had in a while.
This wasn't a result of regulator scrutiny. The issue was that MS (owner of Copilot) was demanding access to the IP (due to their existing agreement with OpenAI), and OpenAI was resisting. In addition, Claude blocked access to Windsurf, which also damaged them as an acquisition target.
I find this hard to believe considering all the recent acquihires that happened recently like Character AI, Inflection, Covariant AI, Scale AI, context AI and so on. Maybe you’re right about the specifics of this situation, but my prior for this being an acquihire is very high and I would need to see very compelling evidence that that is not the case.
Not Windsurf… OpenAI. And OpenAI cares because they’re competing (in part) against Copilot, so if Msoft gets all the benefits of Windsurf then OpenAI would effectively be paying 3B to feed their competitor.
I don't know if you noticed but cursors language server aspect that runs the coding edits and stuff like that from a server to the workstation is a lot better than windsurf.
Windsurf phone's home on every code edit that you have and takes on 30% load on your servers or on your workstation depending on what you're running.
I would strongly discourage the use of windsurf on your systems.
Case in point their AI model that they just built.
Honestly there's no value that windsurf, cursor and all the other VSCode forks provide that couldn't be provided as an extension and even then - none of them perform as well for agentic coding as Cline / Roo Code (debates about the subscription pricing aside due to people often not realising their model limits, public US only based APIs, pay for useful API limits etc aside).
This isn’t a great look for OpenAI, but acquisitions fall through all the time.
The issue isn’t an acquisition not working out, it’s that the founding/exec team felt it appropriate to arrange their own exits and abandon their team before even communicating that their “successful exit” wasn’t actually happening.
Windsurf's value to OpenAI was for the latter to "see the whole chessboard" of context, which is helpful when you're training models to be good at coding.
But codex (and Claude Code) fulfill this from the CLI, and it's a first-party utility, not an acquisition.
What could Google have possible offered? Must have been astronomical. Sundar has been aggressive ever since OpenAI knocked the wind out of them by being first release gpt.
Wow must have seen the numbers and decideed they wanted to call it off. Not a good business model probably so the human talent is where you find the best amount. Still why would anyone ever want to work at Google? Don't they know they are contributing to a system that covertly disseminates information they want you to see. Especially the AI models. Has anyone ever wondered about the training data sometimes? What would people think of them if they knew they had the entire pestein list in their hands but decided its better to protect the ones that pay them. People need to reconsider what they believe from AI, it can be extremely abused to scale narratives.
Wow, this is a pretty fascinating twist. First OpenAI's $3B deal falls through, and now Google swoops in to poach the key talent anyway? Classic big-tech maneuvering
The founders fucked over the employees and the investors and sold out. I guess they don’t care if they are worth $200M each but they fucked every employee that poured their heart out into that company.
I wonder if this is a result of the previously reported clashes between OpenAI and Microsoft over access to the Windsurf IP (under their investment agreement)
Is Lina Khan to blame for this new acquihire meta? She was very aggressive in blocking any tech acquisition during her time and ever since we have seen more and more acquihires which I believe these companies are using to prevent themselves from getting sued.
Google is having a hard time acquiring Wiz for 32b, and if it's blocked they owe 3.2b to Wiz. So why risk it when you can just spend the money to hire the talent behind it and spend a few month building out a new product.
I mean that's exactly what is happening right now. Companies are buying the assets of companies instead of the companies themselves so they don't have to worry about being sued which is a timely and costly process that has become very unreliable due to how much politics are involved in the process now.
@dang - The title’s wording suggest that OpenAI’s CEO is leaving, not Windsurf. A more accurate title might be: “Windsurf’s deal with OpenAI is off, and its CEO is going to Google”
Can someone explain how this works financially for the acquihired? I know they aren’t joining like a regular employee with a high TC. Does Google offer them a giant multi-million (billion?) dollar signing bonus? Why would they tank the value of the company they own just to be another employee at Google?
https://web.archive.org/web/20250711213611/https://www.theve...
IMO other than the Microsoft IP issue, I think the biggest thing that has shifted since this acquisition was first in the works is Claude Code has absolutely exploded. Forking an IDE and all the expense that comes with that feels like a waste of effort, considering the number of free/open source CLI agentic tools that are out there.
Let's review the current state of things:
- Terminal CLI agents are several orders of magnitude less $$$ to develop than forking an entire IDE.
- CC is dead simple to onboard (use whatever IDE you're using now, with a simple extension for some UX improvements).
- Anthropic is free to aggressively undercut their own API margins (and middlemen like Cursor) in exchange for more predictable subscription revenue + training data access.
What does Cursor/Windsurf offer over VS Code + CC?
- Tab completion model (Cursor's remaining moat)
- Some UI niceties like "add selection to chat", and etc.
Personally I think this is a harbinger of where things are going. Cursor was fastest to $900M ARR and IMO will be fastest back down again.
Agreed on everything. Just to add, not only anthropic is offering CC at like a 500% loss, they restricted sonnet/opus 4 access to windsurf, and jacked up their enterprise deal to Cursor. The increase in price was so big that it forced cursor to make that disastrous downgrade to their plans.
I think only way Cursor and other UX wrappers still win is if on device models or at least open source models catch up in the next 2 years. Then i can see a big push for UX if models are truly a commodity. But as long as claude is much better then yes they hold all the cards. (And don't have a bigger company to have a civil war with like openai)
Not sure this is true. Inference margins are substantial and if you look at your claude code usage it's very clever at caching
as an example here's my usage. Massive daily usage for the past two months.> CC at like a 500% loss
Do you have a citation for this?
It might be at a loss, but I don’t think it is that extravagant.
The way I am doing the math with my Max subscription and assuming DeepSeek API prices, it is still x5 times cheaper. So either DeepSeek is losing money (unlikely) or Anthropic is losing lots of money (more likely). Grok kinda confirms my suspicions. Assuming DeepSeek prices, I've probably spent north of $100 of Grok compute. I didn't pay Grok or Twitter a single cent. $100 is a lot of loss for a single user.
Comparison should be with Claude API pricing. It doesn't matter what other models cost.
Claude API pricing has significant margin baked in. I think it's safe to assume that anthropic is getting 80% margin on their api and they are selling claude code for less than that.
I assumed I pay a premium for a Max account to not have to worry about api costs getting stuck in a forever loop.
You are not paying a premium for a max account. You are paying much less than what you would pay if you were using the api directly.
what?? sonnet/opus is way better than deepseek, how can you compare that to deepseek
also you probably talking about distilled deepseek model
I haven't tried deepseek but I've seen claude do crazy things if you are at the correct random.seed
what do you mean about correct random.seed?
It means “if you’re unlucky”.
To me, claude usually feels like a bumbling idiot. But in extremely rare cases it feels like a sentient super intelligence. I facetiously assumed that in those cases it ran on the correct RNG seed.
I'm also curious about this. Claude Code feels very expensive to me, but at the same time I don't have much perspective (nothing to compare it to, really, other than Codex or other agent editors I guess. And CC is way better so likely worth the extra money anyway)
I think GP is talking about Claude Code Max 100 & 200 plans. They are very reasonable compared to anything else that has per-use token usage.
I am on Max and I can work 5 hrs+ a day easily. It does fall back to Sonnet pretty fast, but I don't seem to notice any big differece.
Yes, my CC usage is regularly $50-$100 per day, so their Max plan is absolutely great value that I don’t expect to last.
Pretty easy to hit $100 an hour using Opus on API credits. The model providers are heavily subsidized, the datacenters appear to be too. If you look at the Coreweave stuff and the private datacenters it starts looking like the telecom bubble. Even Meta is looking to finance datacenter expansion - https://www.reuters.com/business/meta-seeks-29-billion-priva...
The reason they are talking about building new nuclear power plants in the US isn't just for a few training runs, its for inference. At scale the AI tools are going to be extremely expensive.
Also note China produces twice as much electricity as the United States. Software development and agent demand is going to be competitive across industries. You may think, oh I can just use a few hours of this a day and I got a week of work done (happens to me some days), but you are going to end up needing to match what your competitors are doing - not what you got comfortable with. This is the recurring trap of new technology (no capitalism required.)
There is a danger to independent developers becoming reliant on models. $100-$200 is a customer acquisition cost giveaway. The state of the art models probably will end up costing hourly what a human developer costs. There is also the speed and batching part. How willing is the developer to, for example, get 50% off but maybe wait twice as long for the output. Hopefully the good dev models end up only costing $1000-$2000 a month in a year. At least that will be more accessible.
Somewhere in the future these good models will run on device and just cost the price of your hardware. Will it be the AGI models? We will find out.
I wonder how this comment will age, will look back at it in 5 or 10 years.
Your excellent comments make me grateful that I am retired and just work part time on my own research and learning. I believe you when you say professional developers will need large inference compute budgets.
Probably because I am an old man, but I don’t personally vibe with full time AI assistant use, rather I will use the best models available for brief periods on specific problems.
Ironically, when I do use the best models available to me it is almost always to work on making weaker and smaller models running on Ollama more effective for my interests.
BTW, I have used neural network tech in production since 1985, and I am thrilled by the rate of progress, but worry about such externalities as energy use, environmental factors, and hurting the job market for many young people.
I've been around for a while (not quite retirement age) and this time is the closest to the new feeling I had using the internet and web in the early days. There are simultaneously infinite possibilities but also great uncertainty what pathways will be taken and how things will end up.
There are a lot of parts in the near term to dislike here, especially the consequences for privacy, adtech, energy use. I do have concerns that the greatest pitfalls in the short terms are being ignored while other uncertainties are being exaggerated. (I've been warning on deep learning model use for recommendation engines for years, and only a sliver of people seem to have picked up on that one, for example.)
On the other hand, if good enough models can run locally, humans can end up with a lot more autonomy and choice with their software and operating systems than they have today. The most powerful models might run on supercomputers and just be solving the really big science problems. There is a lot of fantastic software out there that does not improve by throwing infinite resources at it.
Another consideration is while the big tech firms are spending (what will likely approach) hundreds of billions of dollars in a race to "AGI", what matters to those same companies even more than winning is making sure that the winner isn't a winner takes all. In that case, hopefully the outcome looks more like open source.
The SOTA models will always run in data centers, because they have 5x or more VRAM and 10-100x the compute allowance. Plus, they can make good use of scaling w/ batch inference which is a huge power savings, and which a single developer machine doesn’t make full use of.
> Pretty easy to hit $100 an hour
I don’t see how that can be true, but if it is…
Either you, or I are definitely use Claude Code incorrectly.
It’s definitely easy with an API key I hit 200$ in an evening. I didn’t think that could be possible. Horrifying
To be clear, this is a lot of full-scale reading and (re)writing, without any rules, promots, "agents"/code to limit your resource usage, right?
Nobody's asking for $200 in single-line diffs in less than a day - right?
Right, this is having Claude Code just running as an agent doing a lot of stuff. Also tool use is a big context hog here.
It’s not about a single line diff but the same prompts with cursor does not end costing that much
Yes - do you understand why?
Yes I do, it’s just for new comers who are used to cursor where, without careful prompting you just lock yourself out of premium requests, that’s not immediately a given that CC is more dangerous and does not work the same way at all. Of course it requires careful planning but the trap is easy to fall into.
In my experience, it's more about the tool's local indexing and aggressive automatic upload and model usage limitations to avoid them (and you) overpaying.
People are recreating this with local toolchains now.
This is around what what Cursor was costing me with Claude 4 Opus before I switched to Claude Code. Sonnet works fine for some things, but for some projects it spews unusable garbage, unless the specification is so detailed that it's almost the implementation already.
> unless the specification is so detailed that it's almost the implementation already.
You mean… it’s almost exactly like working with interns and jr developers? ;)
Yeah, it's like an idiot savant intern who has memorized all the API docs in the world, but still somehow struggles to work independently.
And never remembers what we talked about 5 minutes ago on a different thread.
This is where something like Perplexity's "memory" feature is really great. It treats other threads similarly to web resources.
I would love to understand better just how Perplexity is able to integrate up-to-date sources like other theads (and presumably recent web searches, but I haven't verified this, they could be just from the latest model) into it's query responses. It feels seamless.
Why “no capitalism required”? Competition of this kind is only possible with capitalism.
Have you been human before? competition for resources and status is an instinctive trait.
It rears its head regardless of what sociopolitical environment you place us in.
You’re either competing to offer better products or services to customers…or you’re competing for your position in the breadline or politburo via black markets.
Even in the Soviet Union there were multiple design bureaus competing for designs of things like aircraft. Tupolec, Ilyushin, Sukhoi, Mikoyan-Gyurevich (MiG), Yakolev, Mil. There were quite a lot. Several (not all, they had their specialisations) provided designs when a requirement was raised. Not too different from the US yet not capitalist.
Unfortunately it's called war and it appears to be part of human nature.
Not really, it's possible with any market economy, even a hypothetical socialist one (that is, one where all market actors are worker-owned co-ops).
And, since there is no global super-state, the world economy is a market economy, so even if every state were a state-owned planned economy, North Korea style, still there would exist this type of competition between states.
Worker owned coops are not socialist unless the government forces it.
That is true, but if all business are worker owned Co-ops then it has to be forced
Worker-owned co-ops is the basic idea of socialism (that is what "workers owning the means of productions" means in modern language).
I mean, if you wanna get technical, many companies in Silicon Valley are worker-owned (equity compensation)
They are not worker owned, they have some small amount of worker ownership. But the majority of stock is never owned by workers, other than the CEO.
Consider also that VC funds often have pension funds as their limited partners. Workers have a claim to their pension, and thus a claim to the startup returns that the VC invests in.
So yeah it basically comes down to your definition of "worker-owned". What fraction of worker ownership is necessary? Do C-level execs count as workers? Can it be "worker-owned" if the "workers" are people working elsewhere?
Beyond the "worker-owned" terminology, why is this distinction supposed to matter exactly? Supposing there was an SV startup that was relatively generous with equity compensation, so over 50% of equity is owned by non-C-level employees. What would you expect to change, if anything, if that threshold was passed?
> Supposing there was an SV startup that was relatively generous with equity compensation, so over 50% of equity is owned by non-C-level employees. What would you expect to change, if anything, if that threshold was passed?
If the workers are majority owners, then they can, for example, fire a CEO that is leading the company in the wrong direction, or trying to cut their salaries, or anything like that.
>If the workers are majority owners, then they can, for example, fire a CEO that is leading the company in the wrong direction, or trying to cut their salaries, or anything like that.
Why wouldn't the board fire said CEO?
The most common reason to cut salaries is if the company is in dire financial straits regardless. Co-ops are more likely to cut salary and less likely to do layoffs.
Because the board doesn't understand the business at the level that employees do. Or because the board has different goals for the business than employees do. Or because the board is filled with friends of the CEO who let them do whatever.
Also, lots of companies reduce salaries or headcount if they feel they can get away with it. They don't need to be in dire financial straights, it's enough to have a few quarters of no or low growth and to want to show a positive change.
It’s not “your definition”. Worker owned means the workers own the means of production. What you’re talking about is not that at all.
What changes is democracy in the work place.
You are confusing owning minority equity with what actual control gives you —- actual ownership of capital/MoP/assets/profits
How specifically would you expect a typical SV corp's policies to change if employee equity passes from 49% to 51%?
Remember, if employees own 49%, if they can persuade just 2% of the other shareholders that a change will be positive for the business, they can make that change. So minority vs majority is not as significant as it may seem.
Can you give me an idea of how much interaction would be $50-$100 per day? Like are you pretty constantly in a back and forth with CC? And if you wouldn’t mind, any chance you can give me an idea of productivity gains pre/post LLM?
Yes, a lot of usage, I’d guess top 10% among my peers. I do 6-10hrs of constant iterating across mid-size codebases of 750k tokens. CC is set to use Opus by default, which further drives up costs.
Estimating productivity gains is a flame war I don’t want to start, but as a signal: if the CC Max plan goes up 10x in price, I’m still keeping my subscription.
I maintain top-tier subscription to every frontier service (~$1k/mo) and throughout the week spend multiple hours with each of Cursor, Amp, Augment, Windsurf, Codex CLI, Gemini CLI, but keep on defaulting to Claude Code.
I am curious what kind of development you’re doing and where your projects fall on the fast iteration<->correctness curve (no judgment). I’ve used CC Pro for a few weeks now and I will keep it, it’s fantastically useful for some things, but it has wasted more of my time than it saved when I’ve experimented with giving it harder tasks.
It's interesting to work with a number of people using various models and interaction modes in slightly different capacities. I can see where the huge productivity gains are and can feel them, but the same is true for the opposite. I'm pretty sure I lost a full day or more trying to track down a build error because it was relatively trivial fpr someone to ask CC or something to refactor a ton of files, which it seems to have done a bit too eagerly. On the other hand, that refactor would have been super tedious, so maybe worth it?
Thank you for your perspective. I’ve been staring at Claude Code for a bit and I think I will just pull the trigger.
It’s a wild frontier, but as a recent convert to CC, I would say go for it.
It’s so stupid fast to get running that you aren’t out anything if you don’t like it.
There was no way I was going to switch to a different IDE.
Mostly to save money (I am retired) I mostly use Gemini APIs. I used to also use good open weight models on groq.com, but life is simpler just using Gemini.
Ultimately, my not using the best tools for my personal research projects has zero effect on the world but I am still very curious what elite developers with the best tools can accomplish, and what capability I am ‘leaving on the table.’
I am curious what kind of code development you are doing with so many subscriptions?
Are you doing front end backend full stack or model development itself?
Are you destilling models for training your own?
I have never heard someone using so much subscription?
Is this for your full time job or startup?
Why not use qwen or deep seek and host it yourself?
I am impressed with what you are doing.
I’m a founder/CTO of an enterprise SaaS, and I code everything from data modeling, to algos, backend integrations, frontend architecture, UI widgets, etc. All in TypeScript, which is perfectly suited to LLMs because we can fit the types and repo map into context without loading all code.
As to “why”: I’ve been coding for 25 years, and LLMs is the first technology that has a non-linear impact on my output. It’s simultaneously moronic and jaw-dropping. I’m good at what I do (eg, merged fixes into Node) and Claude/o3 regularly finds material edge cases in my code that I was confident in. Then they add a test case (as per our style), write a fix, and update docs/examples within two minutes.
I love coding and the art&craft of software development. I’ve written millions of lines of revenue generating code, and made millions doing it. If someone forced me to stop using LLMs in my production process, I’d quit on the spot.
Why not self host: open source models are a generation behind SOTA. R1 is just not in the same league as the pro commercial models.
> If someone forced me to stop using LLMs in my production process, I’d quit on the spot.
Yup 100% agree. I’d rather try to convince them of the benefits than go back to what feels like an unnecessarily inefficient process of writing all code by hand again.
And I’ve got 25+ years of solid coding experience. Never going back.
> data modeling, to algos, backend integrations, frontend architecture, UI widgets, etc. All in TypeScript, which is perfectly suited to LLMs because we can fit the types and repo map into context without loading all code.
Which frameworks & libraries have you found work well in this (agentic) context? I feel much of the js lib. landscape does not do enough to enforce an easily-understood project structure that would "constrain" the architecture and force modularity. (I might have this bias from my many years of work with Rails that is highly opinionated in this regard).
When you say generation behind, can you give a sense of what that means in functionality per your current use? Slower/lower quality, it would take more iterations to get what you want?
Context rot. My use case is iterating over a large codebase which quickly grows context. All LLMs degrade with larger context sizes, well below their published limits, but pro models degrade the least. R1 gets confused relatively quickly, despite their published numbers.
I think Fiction LiveBench captures some of those differences via a standardized benchmark that spreads interconnected facts through an increasingly large context to see how models can continue connecting the dots (similar to how in codebases you often have related ideas spread across many files)
https://fiction.live/stories/Fiction-liveBench-May-22-2025/o...
i've been using llm-based tools like copilot and claude pro (though not cc with opus), and while they can be helpful – e.g. for doc lookups, repetitive stuff, or quick reminders – i rarely get value beyond that. i've honestly never had a model surface a bug or edge case i wouldn’t have spotted myself.
i've tried agent-style workflows in copilot and windsurf (on claude 3.5 and 4), and honestly, they often just get stuck or build themselves into a corner. they don’t seem to reason across structure or long-term architecture in any meaningful way. it might look helpful at first, but what comes out tends to be fragile and usually something i’d refactor immediately.
sure, the model writes fast – but that speed doesn't translate into actual productivity for me unless it’s something dead simple. and if i’m spending a lot of time generating boilerplate, i usually take that as a design smell, not a task i want to automate harder.
so i’m honestly wondering: is cc max really that much better? are those productivity claims based on something fundamentally different? or is it more about tool enthusiasm + selective wins?
Yeah completely agree.
Honestly reading some of the comments here makes me want to do some sort of course on using them properly. I feel like I'm using them incorrectly.
Approx 250 working days in a year. 25 years coding. Just one million lines would be phenom output, at 160 lines per day forever. Now you are claiming multiple millions? Come on.
It's impossible as an IC on a team, or working where a concept of "tickets" exists. It's unavoidable as a solo founder, whether you're building enterprise systems or expanding your vision. Some details -
1. Before wife&kids, every weekend I would learn a library or a concept by recreating it from scratch. Re-implementing jQuery, fetch API via XHR, Promises, barebones React, a basic web router, express + common middlewares, etc. Usually, at least 1,000 lines of code every weekend. That's 1M+ over 25 years.
2. My last product is currently 400k LOCs, 95% built by me over three years. I didn't one-shot it, so assuming 2-3x ongoing refactors, that's more than 1M LOCs written.
3. In my current product repo, GitHub says for the last 6 months I'm +120k,-80k. I code less than I used to, but even at this rate, it's safely 100k-250k per year (times 20 years).
4. Even in open source, there are examples like esbuild, which is a side project from one person (cofounder and architect of Figma). esbuild is currently at ~150k LOCs, and GitHub says his contributions were +600k,-400k.
5. LOCs are not the same. 10k lines of algorithms can take a month, but 10K of React widgets is like a week of work (on a greenfield project where you know exactly what you're building). These days, when a frontend developer says their most extensive UI codebase was 100k LOCs in an interview, I assume they haven't built a big UI thing.
So yes, if the reference point is "how many sprint tickets is that", it seems impossible. If the reference point is "a creative outlet that aligns with startup-level rewards", I think my statement of "millions of lines" is conservative.
Granted, not all of it was revenue-generating - much was experimental, exploratory, or just for fun. My overarching point was that I build software products for (great) living, as opposed to a marketer who stumbled into Claude Code and now evangelizes it as some huge unlock.
100-200 lines per day, written, debugged, tested and deployed, is normal performance, isn't it? I think I could do it if worked for 8 hours.
No, it’s not. At all. At the overwhelming majority of companies I’ve worked for or heard of, even 400-500 lines fully shipped in a week, slightly less than your figure here, would be top quartile of output - but further, it isn’t necessarily the point. Writing lines of code is a pretty small part of the job at companies with more than about 5-6 engineers on staff, past that it’s a lot more design and architecture and LEGO-brick-fitting - or just politicking and policying. Heck, I know folks who wish they could ship 400 lines of code a month, but are held back by the bureaucracies of their companies.
Uh... Totaling +1000 at the end of a work week is an easy thing to do, especially if working on a new/evolving product.
Now extrapolate. That’s maybe 50k a year assuming some PTO.
10 years would make 500k and you just cross a million at 20.
So that would have to be 20 years straight of that style of working and you’re still not into plural millions until 40 years.
If someone actually produced multiple millions of lines in 25 years, it would have to be a side effect of some extremely verbose language where trivial changes take up many lines (maybe Java).
Maybe there's some copying and pasting involved. ;-)
Just fitting types and repo map into context eh? That's great. Any other tips?
Re productivity gains, CC allows me to code during my commute time. Even on a crowded bus/train I can get real work done just with my phone.
Unless you're getting paid for your commute, you're just giving your employer free productivity. I would recommend doing literally anything else with that time. Read a book, maybe.
Everywhere I've worked as a programmer you're just paid to do your job. If you get some of it done on your commute what difference does it make?
If you can't do your job in your 8 hours then you're either not good enough or the requirements are too much and the company should change processes and hire.
Right, I'm not saying anyone should actually be in the office 40 hours a week that sounds terrible. And even with all the RTO of the last couple years that doesn't seem to be expected many places.
It's for a paid side gig.
How do you use Claude Code via your phone?
Personally I use dev containers on a server and I have written some template containers for quickly setting up new containers that has claude code and some scripts for easily connecting to the right container etc. Makes it possible to work on mobile,but lots of room for improvement in the workflow still.
vibetunnel.sh perhaps
What's your workflow if I may ask? I've been interested in the idea as well.
The project is just a web backend. I give Claude Code grunt work tasks. Things like "make X operation also return Y data" or "create Z new model + CRUD operations". Also asking it to implement well-known patterns like denouncing or caching for an existing operation works well.
My app builds and runs fine on Termux, so my CLAUDE.md says to always run unit tests after making changes. So I punch in a request, close my phone for a bit, then check back later and review the diff. Usually takes one or two follow-up asks to get right, but since it always builds and passes tests, I never get complete garbage back.
There are some tasks that I never give it. Most of that is just intuition. Anything I need to understand deeply or care about the implementation of I do myself. And the app was originally hand-built by me, which I think is important - I would not trust CC to design the entire thing from scratch. It's much easier to review changes when you understand the overall architecture deeply.
you can easily reach 50$ per day. by force switching model to opus /model opus it will continue to use opus eventhough there is a warning about approaching limit.
i found opus is significantly more capable in coding than sonnet, especcially for the task that is poorly defined, thinking mode can fulfill alot of missing detail and you just need to edit a little before let it code.
wow. haven't tried Opus but Sonnet 4 is already damn good.
Is there a cheap version for hobbyists? Or what’s the best thing for hobbyists to use, just cut and paste?
Claude Code with a Claude subscription is the cheap version for current SOTA.
"Agentic" workflows burn through tokens like there's no tomorrow, and the new Opus model is so expensive per-token that the Max plan pays itself back in one or two days of moderate usage. When people reports their Claude Code sessions costing $100+ per day, I read that as the API price equivalent - it makes no sense to actually "pay as you go" with Claude right now.
This is arguably the cheapest option available on the market right now in terms of results per dollar, but only if you can afford the subscription itself. There's also time/value component here: on Max x5, it's quite easy to hit the usage limits of Opus (fortunately the limit is per 5 hours or so); Max x20 is only twice the price of Max x5 but gives you 4x more Opus; better model = less time spent fighting with and cleaning up after the AI. It's expensive to be poor, unfortunately.
>less time spent fighting with and cleaning up after the AI.
I've yet to use anything but copilot in vscode, which is 1/2 the time helpful, and 1/2 wasting my time. For me it's almost break-even, if I don't count the frustration it causes.
I've been reading all these AI-related comment sections and none of it is convincing me there is really anything better out there. AI seems like break-even at best, but usually it's just "fighting with and cleaning up after the AI", and I'm really not interested in doing any of that. I was a lot happier when I wasn't constantly being shown bad code that I need to read and decide about, when I'm perfectly capable of writing the code myself without the hasle of AI getting in my way.
AI burnout is probably already a thing, and I'm close to that point already. I do not have hope that it will get much better than it is, as the core of the tech is essentially just a guessing game.
I tend to agree except for one recent experience: I built a quick prototype of an application whose backend I had written twice before and finally wanted to do right. But the existing infrastructure for it had bit-rotted, and I am definitely not a UI person. Every time I dive into html+js I have to spend hours updating my years-out-of-date knowledge of how to do things.
So I vibe coded it. I was extremely specific about how the back end should operate and pretty vague about the UI, and basically everything worked.
But there were a few things about this one: first, it was just a prototype. I wanted to kick around some ideas quickly, and I didn't care at all about code quality. Second, I already knew exactly how to do the hard parts in the back end, so part of the prompt input was the architecture and mechanism that I wanted.
But it spat out that html app way way faster than I could have.
Claude Code pro is ~$20USD/ month and is nearly enough for someone like me who can’t use it at work and is just playing around with it after work. I’m loving it.
If you are a hobbyist, just use Google’s gemini-cli (currently free!) on a half dozen projects to get experience.
cursor on a $20/month plan (if you burn thru the free credits) or gemini-cli (free) are 2 great ways to try out this kinda stuff for a hobbyist. you can throw in v0 too, $5/month free credits. susana’s free tier can give you a db as well.
*supabase
I've been enjoying Zed lately
Zed is fantastic. Just dipping my toes in agentic AI, but I was able to fix a failing test I spent maybe 15 minutes trying to untangle in a couple minutes with Zed. (It did proceed to break other tests in that file though, but I quickly reverted that.)
It is also BYOA or you can buy a subscription from Zed themselves and help them out. I currently use it with my free Copilot+ subscription (GitHub hands it out to pretty much any free/open source dev).
Cursor at 20$/M is pretty great
Shhh don’t say that. I love Max. I don’t want it to go anywhere.
You can tell Claude Code to use opus using /model and then it doesn't fall back to Sonnet btw. I am on the $100 plan and I hit rate-limits every now and then, but not enough to warrant using Sonnet instead of Opus.
Before they announced the Max plans, I could easily hit 10-15$ of API usage per day (without even being a heavy user).
Since they announced that you can use the Pro subscription with Claude Code, I've been using it much more and I've never ever been rate limited.
This is what I don’t get about the cost being reported by Claude code. At work I use it against our AWS Bedrock instance, and most sessions will say 15/20 dollars and I’ll have multiple agents running. So I can easily spend 60 bucks a day in reported cost. Our AWS Bedrock bill is only a small fraction of that? Why would you over charge on direct usage of your API?
Anthropic has costs beyond their AWS bill ....
API prices are way higher than actual inference cost.
You can spend $200 worth of tokens in a single day using the Max $200/mo fixed cost plan.
Seems like the survival strategy for cursor would be to develop their own frontier coding model. Maybe they can leverage the data from their still somewhat significant lead in the space to make a solid effort.
I don’t think that’s a viable strategy. It is very very hard and not many people can do it. Just look at how much Meta is paying to poach the few people in the world capable of training a next gen frontier model.
Why are there actually only a few people in the world able to do this?
The basic concept is out there.
Lots of smart people studying hard to catch up to also be poached. No shortage of those I assume.
Good trainingsdata still seems the most important to me.
(and lots of hardware)
Or does the specific training still involves lots of smart decisions all the time?
And those small or big decisions make all the difference?
The basic concept plus a lot of money spent on compute and training data gets you pretraining. After that to get a really good model there’s a lot more fine-tuning / RL steps that companies are pretty secretive about. That is where the “smart decisions” and knowledge gained by training previous generations of sota models comes in.
We’d probably see more companies training their own models if it was cheaper, for sure. Maybe some of them would do very well. But even having a lot of money to throw at this doesn’t guarantee success, e.g. Meta’s Llama 4 was a big disappointment.
That said, it’s not impossible to catch up to close to state-of-the-art, as Deepseek showed.
I’d also add that no one predicted the emergent properties of LLMs as they followed the scaling laws hypothesis. GPT showed all kinds of emergent stuff like reasoning/sentiment analysis when we went up an order of magnitude on the number of parameters. We don’t don’t actually know what would emerge if we trained a quadrillion param model. SOTA will always be mysterious until we reach those limits, so, no, companies like Cursor will never be on the frontier. It takes too much money and requires seeking out things we haven’t ever seen before.
Why are there so few people in the world able to run 100m in sub 10s?
The basic concept is out there: run very fast.
Lots of people running every day who could be poached. No shortage of those I assume.
Good running shoes still seem the most important to me.
1. Cost to hire is now prohibitive. You're competing against companies like Meta paying tens of millions for top talent.
2. Cost to train is also prohibitive. Grok data centre has 200,000 H100 Graphics cards. Impossible for a startup to compete with this.
"Impossible for a startup to compete with this."
its funny to me since xAI literally the "youngest" in this space and recently made an Grok4 that surpass all frontier model
it literally not impossible
I mean, that's a startup backed by the richest man in the world who also was engaged with OpenAI in the beginning.
I assume startup here means the average one, that has a little bit less of funding and connections.
The richest man in the world, who could also divert the world's biggest GPU order from his other company
so is Meta(fb) and Apple but that doesn't seem to be the case
money is "less" important factor, I don't say they don't matters but much less than you would think
xAI isn’t young. The brand, maybe. Not the actual history / timeline. Tesla was working on AI long ago.
xAI was just spun out to raise more money / fix the x finance issues.
Most startups don't have Elon Musk's money.
Because it’s not about “who can do it”, it’s about “who can do it the best”.
It’s the difference between running a marathon (impressive) and winning a marathon (here’s a giant sponsorship check).
There are plenty of people theoretically capable of doing this, I secretly believe some of the most talented people in this space are randos posting on /r/LocalLlama.
But the truth is to have experience building models at this scale requires working at a high level job at a major FAANG/LLM provider. Building what Meta needs is not something you can do in your basement.
The reality is the set of people who really understand this stuff and have experience working on it at scale is very, very small. And the people in this space are already paid very well.
It's a staggeringly bad deal. It's a hugely expensive task where unless you are the literal best in the world, you would never even see any usage. And even for those who are BOTH best and well known they have to be willing to lose billions on repeat with no end in sight.
It's very very rare to have winner takes all to such an extreme degree as code llm models
I don't think it's literally "winner takes all" - I regularly cycle between Gemini, DeepSeek and Claude for coding tasks. I'm sure any GPT model would be fine too, and I could even fall back to Qwen in a pinch (exactly what I did when I was in China recently with no ability to access foreign servers).
Claude does have a slight edge in quality (which is why it's my default) but infrastructure/cost/speed are all relevant too. Different providers may focus on one at the expense of the others.
One interesting scenario where we could end up is using large hosted models for planning/logic, and handing off to local models for execution.
I'd recommend reading some of the papers on what it takes to actually train a proper foundation model, such as the Llama 3 Herd of Models paper. It is a deeply sophisticated process.
Coding startups also try to fine-tune OSS models to their own ends. But this is also very difficult, and usually just done as a cost optimization, not as a way to get better functionality.
You need a person that can hit the ground running. Compute for LLM is extremely capital intensive and you’re always racing against time. Missing performance targets can mean life or death of the company.
Windsurf has developed their own fron tier model. It's pretty good. It's not sota but it's very well aligned with their tool call formatting etc.
interestingly windsurf have done this (I'm not sure how frontier this model is...but it's their own model) https://windsurf.com/blog/windsurf-wave-9-swe-1 but AFAIK cursor have not.
> to develop their own frontier coding model
Uh, the irony is that this is exactly what Windsurf tried.
As an actual user of Windsurf model, I don't think "tried" is fair. I sometimes use it. It's not as smart as Gemini but it iterates quicker and is very well aligned with their tool calls
Why did they fail?
It's both hard AND expensive.
Where is a citation on Anthropic increasing cost to cursor? I had not seen that news, but it would make sense.
But Cursor is also offering OpenAI and Google models.
Which is interesting because Sonnet is cheap and Opus is not on par with o3 for tasks where you want to deploy it.
It probably doesn’t cost them all that much? Maybe they were offering the API at a 500% markup, and code is just breaking even.
If open models become big, open coding agents would be bigger at that point. Even more motivation as well.
you mean the plans are subsidized? pay-per-use doesn't look subsidized to me, I can spend 5$ a day on a medium sized codebase easily.
- Forking VSCode is very easy; you can do it in 1 hour.
- Anthropic doesn't use the inputs for training.
- Cursor doesn't have $900M ARR. That was the raise. Their ARR is ~$500m [1].
- Claude Code already support the niceties, including "add selection to chat", accessing IDE's realtime warnings and errors (built-in tool 'ideDiagnostics'), and using IDE's native diff viewer for reviewing the edits.
[1] https://techcrunch.com/2025/06/05/cursors-anysphere-nabs-9-9...
The cost of the fork isn't creating it, it's maintaining it. But maybe AI could help :/
The cost of vscode fork is that microsoft has restricted extension marketplace for forks. You have to maintain separate one, that is the real dealbreaker
https://open-vsx.org/
Eclipse maintains a public repo.
Forking Linux is very easy; you can do it in 1 hour.
Is this $900M ARR a reliable number?
Their base is $20/mth. That would equal 3.75M people paying a sub to Cursor.
If literally everyone is on their $200/mth plan, then that would be 375K paid users.
There’s 50M VS Code + VS users (May 2025). [1] 7% of all VS Code users having switched to Cursor does not match my personal circle of developers. 0.7% . . . Maybe? But, that would be if everyone using Cursor were paying $200/month.
Seems impossibly high, especially given the number of other AI subscription options as well.
[1] https://devblogs.microsoft.com/blog/celebrating-50-million-d...
Maybe the OP got confused with Cursor's $900mil raise? https://cursor.com/blog/series-c
Last disclosed revenue from Cursor was $500mil. https://www.bloomberg.com/news/articles/2025-06-05/anysphere...
It’s probably due to the top comment citing that number
Yeah that’s probably it!
That's the same order of magnitude though.
The $20/month cursor sub is heavily limited though, for basic casual usage that's fine but you VERY soon run into its limits when working at any speed.
The base plan limit is not hard to hit. Then you're on the usage based rocket.
Enterprise pay more.
I also just prefer CC's UX. I've tried to make myself use Copilot and Roo and I just couldn't. The extra mental overhead and UI context-switching took me out of the flow. And tab completion has never felt valuable to me.
But the chat UX is so simple it doesn't take up any extra brain-cycles. It's easier to alt-tab to and from; it feels like slacking a coworker. I can have one or more terminal windows open with agents I'm managing, and still monitor/intervene in my editor as they work. Fits much nicer with my brain, and accelerates my flow instead of disrupting it
There's something starkly different for me about not having to think about exactly what context to feed to the tool, which text to highlight or tabs to open, which predefined agent to select, which IDE button to press
Just formulate my concepts and intent and then express those in words. If I need to be more precise in my words then I will be, but I stay in a concepts + words headspace. That's very important for conserving my own mental context window
Claude Code is just proving that coding agents can be successful. The interface isn’t magic, it just fits the model and integrates with a system in all the right ways. The Anthropic team for that product is very small comparatively (their most prolific contributor is Claude), and I think it’s more of a technology proof than a core competency - it’s a great API $ business lever, but there’s no reason for them to try and win the “agentic coding UI” market. Unless Generative AI flops everywhere else, these markets will continue to emerge and need focus. The Windsurf kerfuffle is further proof that OpenAI doesn’t see the market as must-win for a frontier model shop.
And so I’d say this isn’t a harbinger of the death of Cursor, instead proof that there’s a future in the market they were just recently winning.
I was being hyperbolic saying their ARR will go to zero. That's obviously not the case, but the point is that CC has revealed their real product was not "agentic coding UI", it was "insanely cheap tokens". I have no doubt they will continue to see success, but their future right now looks closer to being a competitor to free/open tools like cline/roo code, as well as the CLI entrants, not a standalone $500M ARR juggarnaut. They have no horse in the race in the token market, they're a middleman.
They either need to create their own model and compete on cost, or hope that token costs come down dramatically so as to be too cheap to meter.
Digging in here more... why would you say it isn't in Anthropic's interest to win the "agentic coding UI" market?
My mental model is that these foundation model companies will need to invest in and win in a significant number of the app layer markets in order to realize enough revenue to drive returns. And if coding / agentic coding is one of the top X use cases for tokens at the app layer, seems logical that they'd want to be a winner in this market.
Is your view that these companies will be content to win at the model layer and be agnostic as to the app layer?
My intuition is that their fundamental business is executing on the models, and any other products are secondary and exist to drive revenue that they can use to compete against Google/OpenAI/Meta as well as to ensure - and demonstrate - that their models are performant in these new markets. Claude needs to be great at coding, but Anthropic doesn’t need to own Coding. Claude Code is growing their core business, just like a Claude Robotics or a Claude Scheduling might, but they cant focus on robotics or scheduling because that takes them away from the core business of models. A strategic relationship with Cursor might have been enough to accomplish this, but it wasn’t - maybe Cursor couldn’t execute fast enough, or didn’t align on priorities, or whatever. I’ve watched a bunch of interviews with the CC team and I very much get the impression that it was more “holy shit, this works great” than a product strategy.
You may be right about “they need to invest in and win” in order to have __enough__ revenue to outcompete the nation-state sized competition, but this stuff is moving way to fast for anyone know.
> A strategic relationship with Cursor might have been enough to accomplish this, but it wasn’t
It’s a huge risk as Cursor can get acquired, just like what this news article is about.
Cursor see it coming - it's why they're moving to the web and mobile[0]
The bigger issue is the advantage Anthropic, Google and OpenAI have in developing and deploying their own models. It wasn't that long ago that Cursor was reading 50 lines of code at a time to save on token costs. Anthropic just came out and yolo'd the context window because they could afford to, and it blew everything else away.
Cursor could release a cli tomorrow but it wouldn't help them compete when Anthropic and Google can always be multiples cheaper
[0] https://cursor.com/blog/agent-web
> Anthropic just came out and yolo'd the context window because they could afford to
I don’t think this is true at all. The reason CC is so good is that they’re very deliberate about what goes in the context. CC often spends ages reading 5 LOC snippets, but afterwards it only has relevant stuff in context.
Background of how it works: https://kirshatrov.com/posts/claude-code-internals
Prompt: https://gist.github.com/transitive-bullshit/487c9cb52c75a970...
Also check out claude-trace, which injects fetch hooks to get at the data: https://github.com/badlogic/lemmy/tree/main/apps/claude-trac...
I'm always surprised how short system prompts are. It makes me wonder where the rest of the app's behavior is encoded.
Lots and lots of reinforcement training.
I’ve definitely observed that CC is waaaay slower than cursor
Heard a lot of this context bs parroted all over HN, don't buy it. If simply increasing context size can solve problem, Gemini would be the best model for everything.
Gemini tends to be better at bug hunting, but yes everything else Claude is still superior
I think this is an interesting and cool direction for Cursor to be going in and I don't doubt something like this is the future. But I have my doubts whether it will save them in the short/medium term:
- AI is not good enough yet to abandon the traditional IDE experierence if you're doing anything non-trivial. Hard finding use cases for this right now.
- There's no moat here. There are already a dozen "Claude Code UI" OSS projects with similar basic functionality.
I have a whole backlog of trivial tasks I never get around to because I’m working on less trivial things.
Strictly speaking about large, complex, sprawling codebases, I don't think you can beat the experience that an IDE + coding agent brings with a terminal-based coding agent.
Auto-regressive nature of these things mean that errors accumulate, and IDEs are well placed to give that observability to the human, than a coding agent. I can course correct more easily in an IDE with clear diffs, coding navigation, than following a terminal timeline.
You can view and navigate the diffs made by the terminal agent in your IDE in realtime, just like Cursor, as well as commit, revert, etc. That’s really all the “integration” you need.
> I don't think you can beat the experience that an IDE + coding agent brings with a terminal-based coding agent.
CC has some integration with VSC it is not all or nothing.
Honestly, I think the Claude Code integration in VS Code is very close to the « nothing » part of the spectrum!
> I don't think you can beat the experience that an IDE + coding agent brings with a terminal-based coding agent.
I resisted moving from Roo in VS Code to CC for this reason, and then tried it for a day, and didn't go back.
Some excellent points. On “add selection to chat”, I just want to add that the Claude Code VS code extension automatically passes the current selection to the model. :)
I am genuinely curious if any Cursor or Windsurf users who have also tried Claude Code could speak to why they prefer the IDE-fork tools? I’ve only ever used Claude Code myself - what am I missing?
Cursor's tab completion model is legitimately fantastic and for many people is worth the entire $20 subscription. Lint fixes or syntax-level refactors are guessed and executed instantly with TAB with close to 100% accuracy. This is their final moat IMO, if Copilot manages to bring their tab completion up to near parity, very little reason to use Cursor.
Idk. When you're doing something it really gets it's super nice, but it's also off a lot of times and it's IMO super distracting when it constantly pop up. No way to explicitly request it instead - other than toggling, which seems to also turn off context/edit tracking, because after toggling on it does not suggest anything until you make some edits.
While Zed's model is not as good the UI is so much better IMO.
Just to offer a different perspective, I use Cursor at work and, coming from emacs (which I still use) with copilot completions only when I request them with a shortcut, Cursor’s behavior drives me crazy.
Which Emacs Package do you use for CoPilot, i tried using Copilot.el a long while ago, but had problems with it. Is there something new or does copilot.el fulfill your needs?
Just copilot.el (I checked to be sure and I'm pulling from github.com/zerolfx/copilot.el)
My config is quite simple:
I only get suggestions if I use that key (the prefix looks huge but I have a Keyboardio Model 100 and I have that bound to the Any key, so I intentionally picked a crazy-long prefix hoping to avoid collision with other shortcuts) and that's the way I like to use these tools (which is why the behaviour in Cursor drives me crazy, though I admit I haven't spend time looking at its configuration, maybe it's something that can be turned off).(Edit: formatting)
I’ll give it another look, thanks! Seems so easy to do!
I haven't used Cursor or Claude much, how different is it from Copilot? I bounce between desktop ChatGPT (which can update VS Code) and copilot. Is there an impression that those have fallen behind?
IME, one of execution. Copilot is like having your cousin who works at Bestbuy try and help you code - it knows what a computer is, and speaks english, but is pretty bad at both
The story I've heard is that Cursor is making all their money on context management and prompting, to help smooth over the gap between "you know what I meant" and getting the underlying model to "know what you meant"
I haven't had as much experience with Claude or Claude Code to speak to those, but my colleagues speak of them highly
Github Copilot just added that about a week ago.
<https://forum.cursor.com/t/i-made-59-699-lines-of-agent-edit...>
It's quite interesting how little the Cursor power users use tab. Majority of the posts are some insane number of agent edits and close to (or exactly) 0 tabs.
At my company we have an enterprise subscription and we're also all allowed to see the analytics for the entire company. Last I checked, I was literally the number one user of Tab and middle of the pack for agent.
It's interesting when I see videos or reddit posts about cursor and people getting rate limited and being super angry. In my experience tab is the number one feature, and I feel like most people using agent are probably overusing it tasks that would honestly take less time to do myself or using models way smarter than they need to be for the task at hand.
I use cursor strictly for agent edits and do anything else in a proper IDE meaning in a Jetbrains product that I run in a separate window.
Many of my co-workers do the same. VC Code is vastly inferior when it comes to editing and actual IDE feature so it is a non-starter when you do programming yourself.
I once tried AI tab-complete on Zed and it was all right but breaks my flow. Either the AI does the editing or I do it but mixing both annoys me.
I find tab extremely distracting and it was the first thing I turned off. I have no idea how people can tolerate it.
I'd like to ask the opposite question: why do people prefer command line tools? I tried both and I prefer working in IDE. The main reason is that I don't trust the LLMs too much and I like to see and potentially quickly edit the changes they make. With an IDE, I can iterate much faster than with the command line tool.
I haven't tried Claude Code VS Code extension. Did anyone replaced Cursor with this setup?
I replaced. My opinion: Cursor sucks as an IDE. Cursor may have a average to above average quality in IDE assistance - but the IDE seems to get in the way. It's entire performance is based on the real-time performance and latency from their servers and sometimes it is way too slow. The TAB autocomplete that was working for you in the last 30 minutes suddenly doesn't work randomly, or just experiences severe delays that it stops making sense.
Besides that, the IDE seems poorly designed - some navigation options are confusing and it makes way too many intrusive changes (ex: automatically finishing strings).
I've since gone back to VS Code - with Cline (with OpenRouter and super cheap Qwen Coder models, Windsurf FREE, Claude Code with $20 per month) and I get great mileage from all of them.
You're looking at (coloured) diffs in your shell is all when it comes to coding. It's pretty easy to setup MCP and have claude be the director. Like I have zen MCP running with an OpenRouter API key, and will ask claude to consult with pro (gemini) or o3, or both to come up with an architecture review / plan.
I honestly don't know how great that is, because it just reiterates what I was planning anyways, and I can't tell if it's just glazing, or it's just drawing the same general conclusions. Seriously though, it does a decent job, and you can discuss / ruminate over approaches.
I assume you can do all the same things in an editor. I'm just comfortable with a shell is all, and as a hardcore Vi user, I don't really want to use Visual Studio.
I also use vim heavily and I've found that I'm really enjoying Cursor + VS Code Vim extension. The cursor tab completion works very nicely in conjunction with vim navigate mode.
JetBrains has CC integration where CC runs in a terminal window but uses the IDE (i.e., Pycharm) for diffing. Works well.
heh, including "for diffing" is selling short when our new job as software developers now seems to be reviewing code, of which looking at a diff is only one tiny part. That goes infinitely more for dynamically typed languages, where there is no compiler to catch dumb typos. If I have to actually, no kidding, review code then I want all the introspections, find references, go to declaration, et al for catching the intern trying to cheat me
I can roll back to different checkpoints with Cursor easily. Maybe CC has it but the fact that I haven’t found it after using it daily is an example of Cursor having a better UX for me.
Or Cursor just gave him a better deal?
I like using Claude Code through Roo Code (vscode extension). I find it easier to work with text using a mouse, vscode diff viewer etc. I guess if you're very good at vim shortcuts etc you can use that in Claude Code instead of selecting text with a mouse. Claude Code has a vscode extension too so I feel that using Claude Code through vscode just adds a better UI.
It already does this btw, when you use Cc from the vscode terminal and select things it adds it to cc context automatically
As does Copilot
I never understood why those tools need to fork Visual Studio Code. Wouldn't an extension suffice?
Cline and Roo Code (my favorite Cline fork) are fantastic and run as normal VS Code extensions.
Occasionally they lose their connection to the terminal in VSCode, but I’ve got no other integration complaints.
And I really prefer the bring-your-own-key model as opposed to letting the IDE be my middleman.
Using cline for a bit made me realize cursor was doomed. Everything is just a gpt/anthropic wrapper of fancy prompts.
I can do most of what I want with cline, and I've gone back from large changes to just small changes and been moving much quicker. Large refactors/changes start to deviate from what you actually want to accomplish unless you have written a dissertation, and even then they fail.
I agree with all you’ve said but with regards to writing a dissertation for larger changes : have you tried letting it first right a plan for you as markdown (just keep this file uncommitted) and then let it build a checklist of things to do?
I find just referencing this file over and over works wonders and it respects items that were already checked off really well.
I can get a lot done really fast this way in small enough chunks so i know every bit of code and how it works (tweaking manually of course where needed).
But I can blow through some tickets way faster than before this way.
IIRC problem is that VS Code does not allow extensions to create custom UI in the panels areas except for WebViews(?). It makes for not a great experience. Plus Cursor does a lot with background indexing to make their tab completion model really good - more than would be possible with the extensions APIs available.
> Wouldn't an extension suffice?
Not if you want custom UI. There are a lot of things you can do in extension land (continue, cline, roocode, kilocode, etc. are good examples) but there are some things you can't.
One thing I would have thought would be really cool to try is to integrate it at the LSP level, and use all that good stuff, but apparently people trying (I think there was a company from .il trying) either went closed or didn't release anything note worthy...
When the Copilot extension needs a new VS Code feature it gets added, but it isn't available to third party extensions until months later... Err, years later... well, whenever Microsoft feels like it.
So an extension will never be able to compete with Copilot.
As part of this whole drama, the APIs that Copilot uses are being opened up https://code.visualstudio.com/blogs/2025/06/30/openSourceAIE...
I've really struggled with using the extensions - their UI/UX is worse, they're much more limited in what they can do and they're much more unstable (in IntelliJ at least).
Then again writing mostly kotlin I cannot get along with the VS Code forks as they're just not that great outside of typescript projects in my experience.
I tend to prompt in cursor/windsurf and refactor in IntelliJ which is okay but a bit of a pain.
I use Augment extensively and find it superior to cursor in every way - and operates as an extension. It has a really handy task planning interface and meta prompt refinement feature and the costs are remarkably low. The quality of output implantation is higher IMO and I don’t have to do a lot of model selection and don’t get Max model bill explosions. If there’s something Cursor provided that Augment doesn’t via extension it was not functionally useful enough to notice.
I think Augment has been flying under the radar for many people, and really reserve better marketing.
I've been using Augment for over a year with IntelliJ, and never understood why my colleagues were all raving about Cursor and Windsurf. I gave Cursor a real try, but it wasn't any better, and the value proposition of having to adopt a dedicated IDE wasn't attractive to me.
A plugin to leverage your existing tools makes a lot more sense than an IDE. Or at least until/if AI agents get so smart that you don't need most of the IDE's functionality, which might change what kinds of tooling are needed when you're in the passenger seat rather than the driver's seat.
It was so they could close source it.
You can ship clised source extensions
One competitor to Claude Code that I don't hear much about is Jetbrains Junie. From my experience, the code it generates is as good as CC, and if you've purchased a Jetbrains license you probably have some amount of free Junie every month.
I'll fill in some context. I think the value of Cursor as an IDE is probably somewhat ephemeral. It's mostly combating Microsoft's ambitions to keep other players busy and box them out of the market. Agents gain a lot of value from model context protocol and there's an amazingly short list of clients that fully support the protocol, but the VSCode Chat window is one of them: https://modelcontextprotocol.io/clients
I actually do prefer the view that having the agent built into an IDE brings me but I'll be damned if I'm forced to use CoPilot/OpenAI. Second to that, the agent does have access to a lot more contextual tools by being built into the editor like focused linting errors and test failures. Of course that demands your development environment is setup correctly and could be replicatable with Claude Code to some extent.
I never got the valuation. I (and many others) have built open source agent plugins that are pretty much just as good, in our free time (check out magenta nvim btw, I think it turned out neat!)
The Microsoft investments in both VSCode and GitHub are looking incredibly prescient.
Turns out, Balmer was right.
> with a simple extension for some UX improvements
What are the UX improvements?
I was using the Pycharm plugin and didn’t notice any actual integration.
I had problems with pycharm’s terminal—not least of which was default 5k line scroll back which while easy to change was worst part of CC for me at first.
I finally jumped to using iterm and then using pycharm separately to do code review, visual git workflows, some run config etc.
But the actual value of Pycharm—-and I’ve been a real booster of that IDE has shrank due to CC and moving out of the built in terminal is a threat to usage of the product for me.
If the plugin offered some big value I might stick with it but I’m not sure what they could even do.
#1 improvement for VS Code users is giving the agent MCP tools to get diagnostics from the editor LSPs. Saves a tremendous amount of time having the agent run and rerun linting commands.
This is a great point. Now I'm wondering if there's a way to get LSPs going with the terminal/TUI interface.
opencode has that
>> Claude Code has absolutely exploded
Does anyone have a comparison between this and OpenAI Codex? I find OpenAI's thing really good actually (vastly better workflow that Windsurf). Maybe I am missing out however.
Codex CLI is very bad, it often struggles to even find the file and goes on a rampage inside the home directory trying to find the file and commenting on random folders. Using o3/o4-mini in Aider is decent though.
It isn't a cli thing, it is available in the ChatGPT ui. I've been using it for a few weeks.
never met anyone who used codex lol
Maybe that is why they added it to ChatGPT in the regular $20 / month plan. It seems pretty good but again maybe CC is way better.
> What does Cursor/Windsurf offer over VS Code + CC?
Cursor's @Docs is still unparalleled and no MCP server for documentation fetching even comes close. That is the only reason why I still use Cursor, sometimes I have esoteric packages that must be used in my code and other IDEs will simply hallucinate due to not having such a robust docs feature, if any, which is useless to me, and I believe Claude Code also falls into that bucket.
> Cursor's @Docs is still unparalleled and no MCP server for documentation
I strongly disagree. It will put the wrong doc snippets into context 99% of the time. If the docs are slightly long then forget it, it’ll be even worse.
I never use it because of this.
What packages do you use it for? I honestly never had that issue, it's very good in my use cases to find some specific function to call or to figure out some specific syntax.
Claude code can get pretty far simply calling `go doc` on packages.
That's great for Go, but unfortunately not all languages have such an option.
Context7 mcp
Tried it, doesn't work that great
just curious because I'm inexperienced with all the latest tools here
> - Tab completion model (Cursor's remaining moat)
What is that? I have Gemini Code Assist installed in VSCode and I'm getting tab completion. (yes, LLM based tab completion)
Which, as an aside I find useful when it works but also often extremely confusing to read. Like say in C++ I type
The editor might show And it's nearly impossible to tell that I didn't enter that `;` so I move on to the next line instead of pressing tab only to find the `;` wasn't really there. That's also probably an easy example. Literally it feels like 1 of 6 lines I type I can't tell what is actually in the file and what is being suggested. Any tips? Maybe I just need to set some special background color for text being suggested.and PS: that tiny example is not an example of a great tab completion. A better one is when I start editing 1 of 10 similar lines, I edit the first one, it sees the pattern and auto does the other 9. Can also do the "type a comment and it fills in the code" thing. Just trying to be clear I'm getting LLM tab completion and not using Cursor
This feeling of, “what exactly is in the file?” is why I have all AI turned off in my IDE, and run CC independently.
I get all AI or none, so it’s always obvious what’s happening.
Completions are OK, but I did not enjoy the feeling of both us having a hand on the wheel and trying to type at the same time.
It gets even worse when all three of IntelliSense, AI completion, and the human are all vying for control of the input. This can be very frustrating at times.
Tab completion in cursor lets you keep hitting tab and it will jump to next logical spot in file to keep editing or completing from.
Do people actually use this?
My experience with this has been hair-pullingly frustrating
I use Windsurf so I remain in the driver's seat. Using AI coding tools too much feels like brain rot where I can't think sharply anymore. Having auto complete guess my next edit as I'm typing is great because I still retain all the control over the code base. There's never any blocks of code that I can't be bothered to look at, because I wrote everything still.
Just cancelled my Cursor sub due to claude code, so heavily agree.
> - Tab completion model (Cursor's remaining moat)
My local ollama + continue + Qwen 2.5 coder gives good tab completion with minimal latency; how much better is Cursor’s tab completion model?
I’m still weary of letting LLM edit my code so my local setup gives me sufficient assistance with tab completion and occasional chat.
I often use the same setup. Qwen 2.5 coder is very good on its own, but my Emacs setup doesn’t also use web search when that would be appropriate. I have separately been experimenting with the Perplexity Sonar APIs that combine models and search, but I don’t have that integrated with my Emacs and Qwen setup - and that automatic integration would be very difficult to do well! If I could ‘automatically’ use a local Qwen, or other model, and fall back to using a paid service like Perplexity or Gemini grounding APIs just when needed that would be fine indeed.
I am thinking about a new setup as I write this: in Emacs, I explicitly choose a local Ollama model or a paid API like Gemini or OpenAI, so I should just make calling Perplexity Sonar APIs another manual choice. (Currently I only use Perplexity from Python scripts.)
If I owned a company, I would frequently evaluate privacy and security aspects of using commercial APIs. Using Ollama solves that.
What kind of hardware are you using?
I use a laptop with 4090 16GB VRAM, core i9 and 96GB RAM for low latency work and Mac mini M4 for tasks which doesn’t require low latency.
I had written a blog on how I run LLM locally a while back[1] I’ll update the information on models & Mac mini soon.
[1] https://abishekmuthian.com/how-i-run-llms-locally/
Have you tried to hook up your local setup to Cline/Roo or similar stuff?
I haven’t heard of them, can you explain how it could be useful for those run LLMs locally ?
They're agentic IDEs. They don't "do" anything to Llama, they run on top.
And open source tools like aider are, of course, even more validated and get more eyes.
Plus recently launched OpenCode, open source CC is gaining traction fast.
There was always very little moat in the model wrapper.
The main value of CC is the free tool built by people who understand all the internals of their own models.
Unless I’m understanding it wrong, the Tab Completion in Cursor isn’t a moat anymore.
VSCode & CoPilot now offer it.
Is it as good? Maybe not.
But they are really working hard over there at Copilot and seem to be catching up.
I get an Edu license for Copilot, so just ditched Cursor!
I agree it has a good chance of catching up, but the difference in quality is pretty noticeable today. I'd much rather stick with vscode, because I hate all the subtle ways Cursor changes the UI; like taking over the keyboard shortcut for clearing the scrollback in the terminal. But I find it's pretty hard to use Copilot's tab completion after using Cursor for a while.
Windsurf big claim to fame was that you could run their model in airgap and they said they did not train on GPL code. This was an option available for Enterprise customers until they took it away recently to prevent self hosting
I think CC is just far more useful; I use it for literally everything and without MCP (except puppeteer sometimes) as it just writes python/bash scripts to do that far better than all those hacked together MCP garbage bins. It controls my computer & writes code. It made me better as well as now I actually write code, including GUI/web apps, that's are always fully scriptable. It helps me, but it definitely helps CC; it can just interrogate/test everything I make without puppeteer (or other web browser control, which is always brittle as hell).
CC would explode even further if they had official Team/Enterprise plan (likely in the work, Claude Code Waffle flag), and worked on Windows without WSL (supposedly pretty easy to fix, they just didn't bother). Cursor learnt the % of Windows user was really high when they started looking, even before they really supported it.
They're likely artificially holding it back either because its a loss leader they want to use a very specific way, or because they're planning the next big boom/launch (maybe with a new model to build hype?).
They quietly released an update to CC earlier today so it can now be run natively on Windows.
Conversely Cursor is still broken on WSL2.
Kilo Code CEO here. This is why open source is so important!!
Cursor's multi-file tab completion and multi-file diff experience are worth $20 easily IMO.
I truly do not understand people's affinity for a CLI interface for coding agents. Scriptability I understand, but surely we could agree that CC with Cursor's UX would be superior to CC's terminal alone, right? That's why CC is pushing IDE integration -- they're just not there yet.
> surely we could agree that CC with Cursor's UX
I can't stand the UX, or VS Code's UX in general. I vastly prefer having CC open in a terminal alongside neovim. CC is fully capable of opening diffs in neovim or otherwise completely controlling neovim by talking to its socket.
Fair enough. I guess a better way to put it is: for people who like Cursor's UX, but prefer Claude Code's performance as an agent, the combination of both would be the true killer app. Having to choose between these feels like a temporary gap in the evolution of these tools, and I'm ready for us to get past it.
I sympathize with your feeling that there is a "gap", but I'm fairly certain that both my ideal workflow and your ideal workflow are unlikely to be anything more than evolutionary dead ends, like early automobiles that inherited the shape of horse-drawn carriages.
I don't know where the evolution of coding agents will take us in the next couple of years, but I would not be shocked if it looks more like a GitHub issue/PR tracker than a code+chat interface with autocomplete, etc. I'm already noticing the I'm starting to rely on tux + multiple CC instances with independent work trees instead of babysitting each proposed change.
I strongly agree with you. I’m more of a CLI guy, and Claude Code just works. Most good projects have a CLI anyway (gcloud, GitHub CLI, Vercel, etc.). I prefer CLI vs MCP’s. I’m on the $200 plan, and it’s absolutely worth it (never thought I’d say this for a CLI app).
But CC is more work to setup, especially if you want to use multiple models. Not even sure it supports all the models that the IDE products do.
Almost all of this was true before they even announced the purchase. I was so shocked and now I’m not surprised it fell through
I've tried all the CLI and vscode with agent mode (and personally I prefer o4-mini) is the best thing out there.
I don’t see how there will be any money to be made in this industry once these models are quantized and all local. It’s going to be one of the most painful bubble deflations we have ever seen and the biggest success of open source in our lifetimes.
The forked IDE thing I don't understand either, but...
During the evaluation at a previous job, we found that windsurf is waaaay better than anything else. They were expensive (to train on our source code directly) but the solution the offered was outperforming others.
>What does Cursor/Windsurf offer over VS Code + CC?
A lot of devs are not superstar devs.
They don't want a terminal tool, or anything they have to configure.
A IDE you can just download and 'it just works' has value. And there are companies that will pay.
You don't need to be a superstar dev to use CC. If you can use chat window you can use CC.
CC _is_ that took. npm install, login, give tasks. Diff automatically appears in your IDE (in VSC/Intellij at least).
A lot of engineers underestimate the learning curve required to jump from IDE to terminal. Multiple generations of engineers were raised on IDEs. It's really hard to break that mental model.
Yes.
A simple example is GIT. Many insist on using standalone GUI/IDE panels for a simple fetch/push rather than just using terminal.
thats not a fair comparision CC is
agentic tool + anthropic subsidized pricing.
Second part is why it has "exploded"
Good analysis. And Claude code itself will be mercilessly copied, so even if another model jumps ahead, small switching cost.
That said, the creator of Claude Code jumped to Cursor so they must see a there there.
does claude code have a privacy mode with zero data retention?
Haven’t looked recently but when it came out, the story was that it was private by default. It uses a regular API token, which promises no retention.
Cursor - co-pilot/AI pair programming usecases.
Claude Code - Agentic/Autonomous coding usecases.
Both have their own place in programming, though there are overlaps.
I do agree, Claude Code absolutely changed the game. It is outstanding.
Is Claude code expensive? Can you control the costs or can it surprise you.
On a subscription, it is 100% predictable: $20, $100, or $200/month
Is the case for using Claude Code much weaker now that Gemini CLI is out?
no. CC is not just a cli. Its cli + their pro/max plan.
gemini cli is very expensive.
Isn't Gemini CLI 1000 requests/day free?
https://blog.google/technology/developers/introducing-gemini...
It's false advertising - 1000 model requests, not 1000 Gemini 2.5 Pro requests. It drops to Flash after 3-5 requests and Flash is useless.
Flash is very very good (better than 4.1 at faster speeds). Just not anywhere near 2.5 pro.
Theoretically, practically it is riddled with rate limiting issues
https://github.com/google-gemini/gemini-cli/issues/1502
We are working to resolve this. It's still in preview but expect to see healthy Gemini 2.5 Pro allocations for subscription customers in the near future.
Wait a minute, have you often run out of the gemini cli free daily quota? Their free quota is very generous because they are trying to get market/mind share.
it switches to flash almost immediately like in 10-15 mins . flash sucks.
And even switching is not smooth either. for me when the switch happens it just get stuck sitting there so i have to restart cli.
OK, thanks, I understand now what is happening. I use gemini-cli for one specific task at a time and sometimes just one 5 to 10 minute session a day. If I use long work sessions then I will add my API key and pay.
They do have a subscription: it's $22/month, but the whole pricing and instructions is very confusing, it took me 15 min to figure it all out.
Since then OpenAI has released Codex as well (the web one)
for those who seldom use the terminal, is Claude Code still usable? I heard it doesn't do tab autocomplete in IDE like Cursor
I think lots of issues with the integration of CC or other TUI with graphical IDEs will be solved with stuff like the Agentic Coding Protocol that the guys at Zed at working on https://www.npmjs.com/package/@zed-industries/agentic-coding...
I trust zed to get it right over cursor with their continual enshittification.
Claude Code is totally different paradigm. You don't edit your files directly so there is no tab autocomplete. It's a chat session.
There are IDE integrations where you can run it in a terminal session while perusing the files through your IDE, but it's not powering any autocomplete there AFAIK.
are people viewing file diffs in the terminal? surely people aren't just vibing code changes in
Yes. I manually read the diff of every proposed change and manually accept or deny.
I love CC, but letting it auto-write changes is, at best, a waste of time trying to find the bugs after they start compounding.
it seems like CC is king at the moment from what I read.
I currently have a Copilot subscription that has 4.1 for free but Sonnet 4 and Gemini Pro 2.5 with monthly limits. Thinking to switch to CC
I am curious to know which Claude Code subscription most people are using... ?
Yes or running claude code in the cursor/vscode terminal and watching the files change and then reviewing in IDE. I often like to be able to see an entire file when reviewing a diff, rather than just the lines that changed. Plus it's nice to have go-to-definition when reviewing.
Yes, it shows you the file diff. But generally, the workflow is that you git commit a checkpoint, then let it make all the changes it wants freely, then in your IDE, review what has changed since previous commit, iterate the prompts/make your own adjustments to the code, and when you like it, git commit.
Depending on what I'm doing with it I have 3 modes:
Trivial/easy stuff - let it make a PR at the end and review in GitHub. It rarely gets this stuff wrong IME or does anything stupid.
Moderately complex stuff - let it code away, review/test it in my IDE and make any changes myself and tell claude what I've changed (and get it to do a quick review of my code)
Complex stuff - watch it like a hawk as it is thinking and interrupt it constantly asking questions/telling it what to do, then review in my IDE.
I review and modify changes in Zed or Emacs.
Apparently they are, which is crazy to me. Zed agent mode shows modified hunks and you can accept/reject them individually. I can't imagine doing it all through the CLI, it seems extremely primitive.
I use windsurf to check the diff from Claude Code.
If there’s a conflict you just back out your change and do it again.
I just accept all and review in my editor.
that's what lazygit in another terminal tab is for
I too am an engineer that thinks CLI's are cool, but if you believe it's even remotely as useful as an IDE then give me a break.
As far as I can tell, terminal agents are inferior to hosted agents in sandboxed/imaged environments when it comes to concurrent execution and far inferior to assisted ide in terms of UX so what exactly is the point?. The "UI niceties" is the whole point of using cursor and somehow, everyone else sucks at it.
Not sure what you mean. "Hosted agents in sandboxed/imaged environments"? The entire selling point of CC is that you can do
- > curl -fsSL http://claude.ai/install.sh | bash
- > claude
- > OAuth to your Anthropic account
Done. Now you have a SOTA agentic AI with pretty forgiving usage limits up and running immediately. This is why it's capturing developer mindshare. The simplicity of getting up and going with it is a selling point.
Plus it’s straightforward to make Claude Code run agents in parallel/background just like Codex and Cursor, in local sandboxes: https://github.com/dagger/container-use
You’re missing the point tho. The point of the cli agent is that it’s a building block to put this thing everywhere. Look at CCs GitHub plugin, it’s great
CC on github just looks like Codex. I see your point, but it seems like all the big players basically have a CLI agent and most of them think that its just an implementation detail so they dont expose it.
Forking an IDE is not expensive if it's the core product of a company with a $900M ARR.
I doubt MS has ever made $900M off of VS Code.
"The same editor you already use for free, but with slightly nicer UI for some AI stuff" is not a $900M ARR product.
$900m in revenue is easy if you're selling a dollar for <$1. Feels like that's what cursor's $20/m "unlimited" plan is
"Some AI stuff" can well be worth that.
It's another Character.ai situation [0]. Unfortunate for any employees who aren't founders or researchers, as they don't get any payout or a nice new job from this exit structure. In fact they lose their whole time invested at the company.
What a harsh time to work for an AI startup as a rank and file employee! I wonder how the founders justify going along with it inside their mind.
[0] Character.ai CEO Noam Shazeer Returns to Google https://news.ycombinator.com/item?id=41141112 - 11 months ago (87 comments)
Edit: Thank you @jonny_eh for the clarification. I can't imagine it feels awesome being a leftover but at least you vested out. "Take the money and leave" is still a bit raw when the founders and researchers are now getting the initial payout + generous Google RSU's.
The “leftover” employees at Character were NOT screwed over. Options were converted to cash at the deal’s valuation.
Hopefully Windsurf employees are treated well here.
Note: I worked at Character until recently.
On the flipside, I’m pretty sure the investors got screwed.
The investors made money too. The valuation at the last round was $1B, and Google paid them out at a valuation of $2.5B as part of the agreement [1].
[1] https://www.theverge.com/2024/8/2/24212348/google-hires-char...
To quote an analyst from the Dotcom bubble era: "Everbody's happy, everybody's making money, something's wrong here"
This is how it's supposed to work. Nobody has to get screwed. It's not a zero-sum game.
2.5x in 10 months. With returns like that - if I were the full time chef of the investors spare private jet I would be updating my CV and looking for a new gig.
That’s a deep burn.
The full time masseuse should be able to help with that.
Honestly depends on when they got in. Seed investors? They're probably fine with their preferences. Series B and beyond? That's where it gets messy. What round you thinking?
It's literally the opposite - seed investors get paid last with the exception of common.
Hopefully. The world is healing.
Whose cash? OpenAI isn't paying, and Google isn't paying, and Windsurf investors already paid.
I wasn’t referring to Windsurf. But if there was no cash involved here, then ya, the employees were screwed. Do we know that’s the case though?
> I wonder how the founders justify going along with it
$2.4 billion.
This reads like a Dr. Evil plot.
Everything since ~2016 reads like a Dr. Evil plot! I swear it feels like the world is getting dumber around me.
100% agreed.
You've reminded me of when I first watched Idiocracy in 2006. At the time, I delighted in the comedic, sophomoreish, and seemingly ridiculous take on a possible trajectory of humanity. But now much of it is actually coming to pass. It's sad.
P.s. As a sidenote, apparently I love all of Mike Judge's productions, which also includes Office Space, and Beavis and Butthead.
Mike Judge does not judge wrong
He also made Silicon Valley.
The rank and file equity pitch is quickly falling apart…
Think it started that way... I'm currently in a vesting/allocation situation where the incentive is to drive the share price down.
Geeeeeze
Always has been
This.
Character.ai reached out to me for an opportunity, but they've already been carved up.
I think it's great that the rank and file got some of their equity cash-out (based on the other comment), but I imagine it isn't an attractive prospect as a start-up to join at this point.
I just ignored the recruiter. I can't imagine their would be a second liquidity event.
FYI, It wasn't taken the money and leave, a lot of them got absorbed into GDM.
Source: I was in GDM when character was acquired.
Do you mean Google Deep Mind? Curious what use deep mind had for the leftovers (kubernetes and web scraping experts, etc)?
Otherwise why not merge all of engineering into ElGoog?
> Unfortunate for any employees who aren't founders or researchers, as they don't get any payout or a nice new job from this exit structure. In fact they lose their whole time invested at the company.
Windsurf’s value didn’t go to $0 overnight. The company will continue and their equity is likely still worth a decent amount wherever the company ends up.
Obviously a disappointing outcome for the people who thought life changing money was right around the corner, but they didn’t lose everything.
Just like with Character I'm assuming the employees get something. Whatever nonsense "licensing" fee Google is paying to not cause an antitrust investigation should be paid out straight to employees
The general character of capital markets is to pay as little as possible. Otherwise you lose out to those who are more ruthless. It is plausible that Windsurf employees really are getting very little value for their work. We need to see details of the deal.
Not really true, I believe the "acquiring" (i.e. Google) company buys some equity from the employees (windsurf).
Edit: the people downvoting this clearly can't read, I made the exact same point as jonny_eh.
The acquisition of Windsurf was cancelled.
Instead they are paying 2.4B to "license" windsurfs IP. Still a loss vs OpenAI but at least the employees will get cash not openai stock.
This might be the beginning of the end of tech VC startups in general.
High interest rates make VC funding more expensive and now bigtech can swoop in, poach all the necessary staff and deprive investors of an exit.
What is the point any more?
Isn't there not some contractual agreement between the VCs and the founders? (I understand that a non-compete might not apply [in CA], but taking VC money is a little different that simply getting hired).
Were I a Windsurf investor, I'd be pissed right now and calling my lawyer.
the founder is on a vesting schedule set with the vc. walking away forfeits his ownership in the company (not sure of the specifics of this weird deal, but this is true in 99% of situations) which returns his ownership to the VCs either directly or functionally.
the only reason he'd walk away is because he thinks other opportunities are higher EV. if he believes this, a) the investors investment is likely worth virtually 0 anyway and b) if it's not, removing a leader who doesn't want to be there probably increases P(success) for the company and further increases the value of the investment.
founder departure isn't good for the narrative, but it's a symptom of an investment going bad, not often a cause.
Presumably the founder(s) is/are getting a better deal by walking away in this case. If they've been through a few round of funding, they may have been diluted to the point when this sort of exit is better (for them).
The low level employees are screwed. Basically they lost their job. Not cool.
I went from Emacs to VS Code, then to Cursor, next to Claude Code, which is so good that I feel like I am having half a dozen junior devs at my fingertips, 24/7.
Since Claude Code is cli based, I reviewed my cli toolset: Migrated from iTerm2 to Ghostty and Tmux, from Cursor to NeoVim (my God is it good!).
Just had a 14h workday with this tooling. It’s so good that I complete the work of weeks and months within days! Absolutely beast.
At this point I am thinking IDEs do not reflect the changing reality of software development. They are designed for navigating project folders, writing / changing files. But I don’t review files that much anymore. I rather write prompts, watch Claude Code create a plan, implement it, even write meaningful commit messages.
Yes I can navigate the project with neovim, yes I can make commits in git and in lazygit, but my task is best spent in designing, planning, prompting, reviewing and testing.
I'm curious to see what you've built with all that extra productivity.
I work at a company with over 700 employees. And there are tons of use cases where a simple CRUD app is sufficient. Or where glue code needs to be written / changed for legacy systems. Or where an OS system like Camunda is deployed and needs to be configured, workflows developed, etc
The reality of companies out there is much simpler than the challenges of a startup that needs to build systems that are state of the art, scale for millions of users, etc There are companies out there that make millions, in areas you‘ve never heard of, and their core business does not depend on software development best practices.
In our company we have an IT team with the median age of fifty, team members who never have developed software, just maintain systems, delegate hard work to expensive consultants.
Now in that setting someone coming from a startup background is like someone coming from the future. I feel like a wizard who can solve problems in days, instead of weeks or months waiting for a consultant to solve.
Fair enough. There are valid use cases for vibe coding scripts and simple CRUD apps, which current AI tools are fairly competent at producing.
The thing is that those don't typically take weeks and months to build with conventional tooling. And I find it hard to believe that all you're doing is this type of integration work. But I suppose there are companies that need such roles.
> There are companies out there that make millions, in areas you‘ve never heard of, and their core business does not depend on software development best practices.
That is true.
I do think that this cowboy coding approach is doing these companies a disservice, especially where tech is not their main product. It's only creating more operational risk that on-call and support staff have to deal with, and producing more technical debt that some poor soul will inevitably have to resolve one day. That is, it all appears to work until one edge case out of thousands brings down the entire system. Which could all be mitigated, if not avoided, by taking the time to understand the system and by following standard software development processes, even if it does take longer to implement.
What you describe isn't new. This approach has existed long before the current wave of AI tooling. But AI tools make the problem worse by making it easier to ship code quickly without following any software development practices that ensure the software is robust and reliable.
So, it's great that you're enjoying these tools. But I would suggest you adopt a more measured approach and work closely with those senior and junior engineers, instead of feeling like a wizard from the future.
> I do think that this cowboy coding approach is doing these companies a disservice, especially where tech is not their main product. It's only creating more operational risk
I absolutely agree with you! Alas, management in those companies (I have worked for some) does not understand the necessity of solid software engineering work. It’s way above their head. That’s why they have systems that constantly cause problems. And then expensive consultants are called in to fix symptoms, not solve root causes. Shocking to see that in lots of otherwise very successful companies.
Who's reviewing all the code you are churning out with ai? If everyone is used to maintaining not developing software it doesn't sound like they'd be best suited to have to review lots of complex pull requests.
It sounds like you are moving very fast and probably have people just clicking "approve".
Good luck for the future to who ever owns your company!
When I setup systems, I thoroughly document them, test them, develop them according to architectural best practices. My AI assisted code generation is lightyears ahead of what I see in companies I have worked for. The best they —the companies—-do is hire expensive consultants. Who sell them preconfigured system. And when you look into those systems you won’t believe your eyes either. Because you instantly realise that those devs do not know much about architectural patterns, aout systems design, about software development best practices. Yet they sell their systems as well, because they offer a niche product where they have only a handful competitors.
In that setting someone with solid software engineering background using AI to solve problems is like a wizard from the team‘s perspective.
When I worked for startups I was constantly panicking to miss the latest tech trends, and I feared that I would be not marketable in case I didn’t catch up. But in mature companies things work much slower. They work with decades old technology. In that setting not the latest tech counts but being able to solve problems, with whatever means you can.
did i miss it or did you still not answer
> Who's reviewing all the code you are churning out with ai?
I don’t review every single line that AI generates. I glance over the files to see if they meet my standards, prompt it to rewrite this or that portion, when necessary. Or change it myself.
Writing code is the most tedious part, not reviewing.
oh gotcha. you are a solo dev on the work you are doing. makes sense.
I am an IT project manager ;) But apparently you can solve lots of problems with code.
He didn’t answer because he didn’t even read your comment. Likely a bot
Yeah, I am a bot.
Gemini of course! /s
Personally I've had mixed experience when I let Sonnet 3.7 document my (and its) code and write commit messages, for some wip stuff it's alright but it soon gets out of hand and because it doesn't really have a direct view in my mind it ends up documenting what it sees instead of the intention behind it, which is totally fair but eh.
So yeah, mileage varies and agentic tools usually spit out more code and redundant comments than I'd like to review. I'm still waiting for a company to develop some sanity check for this somehow but snapshot testing and manual review aren't enough sadly.
No one ever shares their great and shipped products. AI built slop is for generating hype not revenue or users.
Man, who sucked the joy out of your life. Just try the damn thing. I have the staunchest anti-hypsters in my org and even they are using these tools heavily now.
I build most of not all of my stuff for work, and I ain't sharing that.
It's no panacea, but is there something to be had there? Abso-fucking-lutley. All of this would have been complete scifi at the beginning of this decade.
I’m super pro AI. I’ve been using ChatGPT since the day it released. I use an agent coder at work semi-regularly to reasonable levels of success. Big fan.
But I am exceedingly tired of phrases like “complete the work of weeks and months within days”. If AI is making devs 5x to 10x faster then I’d like to see some actual results. Internet is full of hypesters that make bombastic claims of productivity but never actually shown anything they’ve made.
100%!
Not the OP but this is smth I've vibecoded using cursor: https://bestphoto.ai/ MRR ~$150. It basically started as a clone of my other site: https://aieasypic.com (MRR 2.5k, 5-8k/mo rev) since I was having trouble keeping code context in mind and claude was pretty bad at doing full features with the tech stack I used for that site(Django BE, NextJS FE) making adding new features a pain, so I completely switched to a stack that claude is very good at NextJS fullstack(trpc BE) and now it can basically one-shot a feature request.
Just putting this here because a lot of times AI coding seems to be dismissed as smth that can't do actual work ie generate revenue, while its more like making money as a solo dev is already pretty rare and if you're working in a corp. instead you're not going to just post your company name when asked for examples on what you're using AI for.
Those are exactly the kind of AI slop products I would expect to be vibe coded. You've created yet another wrapper around LLM APIs where the business model is charging a premium over existing services. Your revenue depends on the ignorance of customers to not realize they can get the same or better service for cheaper from companies that actually do the hard work. I bet SEO hacking is really important to you.
It's irrefutable that AI tools can be used to create software that generates revenue. What's more difficult is using them to create something that brings actual value into the world.
> yet another wrapper around LLM APIs
Patio11 famously built, ran for a number of years (profitably) and then sold a "wrapper for a random number generator" (bingocardcreator.com)
Value is in the eye of the beholder, and only tangentially related to the technical complexity or ingenuity.
I'm not arguing in favor of technical complexity or ingenuity.
My point is that the perceived value of a service or product is directly related to its competitive advantage, product differentiation, and so on. When the service is made from the same cookie cutter template as all the others, the only value that can be extracted from it is by duping customers who don't know better.
There are entire industries flooded with cheap and poorly made crap from companies that change brand names every week. Code generation tools have now enabled such grifters to profit from software as well.
> When the service is made from the same cookie cutter template as all the others,
You're conflating "underlying tech" and "cookie cutter template".
The fact that something uses the same LLM under the hood as some other product doesn't mean it's not differentiated.
It's people configuring WordPress with various themes from Envato at jacked-up prices in a trench coat /s
I'm only half joking.
Eh, it is more like an extended/better UI. Plenty of people are willing to pay for just that.
There are lots of people that only use LLMs in whatever UI the model companies are providing. I have colleagues that will never venture outside the ChatGPT website, even though with some effort they could make their tooling richer by using the API and building some wrapper or UI for it.
Sure man, any product you don't like is just "another wrapper". I guess every website is just a wrapper over postgres or wordpress too. I run my own serverless GPU containers on runpod with a combination of comfy and my own fastapi servers using diffusers, not that it'd even matter if I just used some third party APIs. It originally even started as smth that was hacked together using 4x 4070ti supers in my basement that I then moved to runpod. Indiehacking is mostly marketing, nobody cares if you built some technically beautiful thing.
Also its easy to criticize from the sidelines but, do you have products that you made by yourself that are used by hundreds of thousands of people? I have 5 such sites, 2 of which I named above
Hey, don't blame me for the fact that your sites are indistinguishable from hundreds of others that offer the same service. Everything I said is logical to assume, since all these sites look the same.
Good on you for learning how AI tools work, but there's no way for anyone to tell whether your backend is self-managed or not, and practically it doesn't really matter. I reckon your users would get better results from proprietary models that expose an API than self-hosted open source ones, but then your revenue would probably be lower.
> Also its easy to criticize from the sidelines but, do you have products that you made by yourself that are used by hundreds of thousands of people? I have 5 such sites, 2 of which I named above
That's a lazy defense considering anyone is free to criticize anyone else's work, especially if they're familiar with the industry. Just like food and film critics don't need to be chefs and movie producers.
But I'll give you credit for actually building and launching something that generates revenue. I admit that that is more than I have managed with my personal projects.
I love examples like these. I eventually want to start a bunch of these too.
thanks for sharing.
The other day someone was gloating they'd created a 30k LoC code base in a few weeks with a similar setup.
I'd consider that a liability, not an asset, but they were pretty happy with it.
It does exude a strong scent of astroturfing.
That's not always the case.
AI is often used to pump out sites and apps that scam users, SEO spam, etc. So there is definitely a revenue stream that makes scammers and grifters excited for AI. These tools have increased the scope and reach of their scams, and provide a huge boost to their productivity.
That's partly why I'm curious about OP's work. Nobody who's using these tools while following best software engineering practices would claim that they're making them that much more productive. Reviewing the generated code and fixing issues counteracts whatever time is saved by generating code. But if they're vibe coding and don't even look at the code...
> But I don’t review files that much anymore.
Say no more.
I review PRs/commits, not files. Given the right cage to lock the agent inside, and guardrails built around, and conventions and guidelines, and agentic flows so it can pull in what's needed.. the need to look at every line and file during implementation is significantly lessened. So then I review the final output (which is a unit of work/task wrapped in a PR).
yes vibecoding is addicting like that. but if you are not reviewing any code and simply vibing then in my expreience you'll eventually get stuck in "its still not working" loops beause you have no other context or insight to provide it other than that. Then you have either accept what you have or throw the whole thing out and/or actually read the code . kind of rules out last option because code is now just too far gone with too many special cases hardcoded because AI sucks at abstraction or real software engineering.
> I don’t review files that much anymore
You don't review the code? Just test it works?
At work we’re encouraged to use AI, so I do. For me the one thing that works well is using it to write one off scripts that do stuff and would be a chore to write.
Usually in 2-3 prompts I can get a python or shell script that reads some file list somewhere, reads some json/csv elsewhere. Combines it in various ways and spits out some output to be ingested by some other pipeline.
I just test this code if it works it’s good.
Never in my life would I put this in a critical system though. When I review these files they are full of tiny errors that would blow up in spectacular manner if the input was slightly off somewhere.
It’s good for what it is. But I’m honestly afraid of production code being vibe coded by these tools.
When generating code that is often wrong and needing to review it, and IDE is demonstrably better, this isn't an argument
> But I don’t review files that much anymore.
they don't review files anymore though.
Curious - how did moving from iTerm2 to Ghostty help? I currently use iTerm2 and have never used Claude Code
Ghostty is gpu accelerated. It’s super fast, and tmux in it is a joy to use. That combined with NeoVim gives me an increadibly smooth dev experience in the terminal, something I had never have with iTerm2 and emacs.
iTerm is also GPU accelerated.
Configuration in XML, instead of YAML or TOML, it feels bloated, compared to Ghostty.
It's plist, that's what Mac apps are going to use
Try again. No self respecting Emacs user would ever call vim “good”.
Haha :) I lived inside Emacs, used orgmode for everything, have written tons of Elisp, used org-roam as my second brain, used vanilla Emacs shortcuts instead of Evil (with a special keyboard settup using Karabiner Elements), did even my googling from Emacs, used emacs calc instead of my calculator, but in the end I spent more time tinkering my Emacs setup than doing real work. Emacs was a lifestyle. At some point I realized: Unix and the terminal are what Emacs try to be: It tries to be a one-stop shop offering you everything: Surfing the web, writing emails, word processor, calculator, planner, terminal. Unix and the terminal offer me all of that. Plus any scripting language. Why miss all the beautiful apps, just to be an Emacs zealot? The editor in emacs is just one usecase. Neovim does it just as well, if not better.
But relax, noone is taking your Emacs from you :) I still like it, but am not a disciple anymore ;)
> It’s so good that I complete the work of weeks and months within days
and yet you're pulling 14 hour workdays..
That doesn't negate that he is compressing his backlog.
i get it... i find the productivity is extremely addictive
Well you can't risk Claude quitting overnight. It forgets everything it did the day before and now you have to start over ... must ... finish ... tonight ... within ... context ... window.
Fortunately LLMs are stateless thus not affected by passage of time - your context stays exactly as it was while the tool maintaining it is running.
(Prompt caches are another thing; leaving it for the night and resuming the next day will cost you a little extra on resume, if you're using models via API pay-as-you-go billing.)
So half a dozen junior devs plus 14h workday. That's a ton of surplus value right there. Hope he's getting a cut!
IDEs were a crutch, and now that crutch has been replaced by a semi-autonomous bot that can fetch and carry.
>Migrated from iTerm2 to Ghostty and Tmux
Would love to hear more.
Do you do all this switching during the workday?
Cursor (and Garry Tan’s X post) has shown us that the VC money is propping up these companies astounding growth, the only way for them to become profitable is to increase the cost per a request, which means they need to innovate like crazy.
The moat is paper thin.
GitHub has open sourced copilot.
The open source community is working hard on their own projects.
No doubt Cursor is moving fast to create amazing innovations, but if the competition only focuses on thin wrappers they are not worth the billion dollar valuations.
I love watching this space as it is moving extremely fast.
There is no moat. If you’re a true believer that strong agents are around the corner, then all of these add on companies will be obsolete in a few years. The first company to strong agents can trivially rebuild Cursor or Windsurf.
If you believe AGI is around the corner, doesn’t it mean it’ll replace ALL products?
If AGI is around the corner, I don't believe one single company will "own" that tech. It will be like it is today, where you have multiple models competing.
And after that, AGI will be open source.
In the end, ownership of data and compute will be the things that define the victors.
That would be true if the product was the goal. In my experience,
Even with AGI in hand, there will still be competition between offerings based on externalities, inertia, or battle-testedness, or authority. Maybe super-intelligence would change the calculus of that, but you'd still probably find opportunities beyond just letting your pool of agents vibe code it.Sounds like the kind of thing Windsurf & Cursor had as their most valuable asset - mindshare and authority?
If you believe AGI is round the corner (I don't), then you face the dilemma of investing in a company that will be the one profitable survivor in a disintegrating world.
This would only be true if it was cheap to run and would return results quickly. If AGI only has compute to serve 1 customer per hour then their is an upper bound of market share it can take from other products.
So grifting for investor cash and revenue right now is the obvious play either way
We don’t eat intelligence: bikes, bottles, food, energy… have room for improvement but I hardly see how AGI would replace them.
Same for physical services like labors, miners and cooks, even taxi/bus drivers for +99% of the world. Automation immensely improve their efficiency and the Modern Times is the past for half of the globe, but AGI isn’t the main facilitator.
Replace all (most*) Silicon Valley -and cousins- similar "products" and services, perhaps yes !
You don’t see how robots would replace all these roles? I take it you never seen a Waymo or read about how agriculture needs a tiny fraction of humans in the US vs a couple of decades ago? And those aren’t even “AGI-level” solutions…
Waymo aren’t build, maintained and disposed by robots only. Even if they were we’ve still to see how that can be deployed (and make sense economically) to more than a couple percentiles of the richest people in the top cities. That has nothing in common with "replacing the humans".
Agriculture efficiency mainly comes from last century fertilizer. The modern robot trucks are also limited to some richest countries and farmers and have nothing to do with AGI.
Your vague dream of AGI solutionism missed the barriers to technologie deployment: raw ressources availability, politics, demography, energy… IA can have a small impact there but it certainly won’t replace ALL products. How would a computer replace a mining truck or a reactor turbine?
You’re missing the point so completely, I don’t see why I should keep engaging.
If I missed your point you’d be wonderful to clarify it, perhaps by detailing the first post "replacing all products":
- what do you consider a product?
- ALL as in Everything (obviously not I guess)?
Regarding "don’t you see how robots…" I clearly see the theory but also can’t see how it would happen in reality for the reasons I mentions in my precedent post: mainly access to ressources/energy to build/maintain those robots for a signifiant time scale and socio-politico barriers.
I’d be glad you reconsider explaining your point because I only see vague allusions, but if you want to stop the conversation here that’s all up to you.
Google spent 2.4 billion dollars
I am saying the subsidies tokens are the paper thin moat.
I think the recent Grok release and considering xAI was relatively late to the game shows that the only moat to training giant models is how many GPUs you can buy. ChatGPT was earth-shattering and it took less than two years for multiple credible competitors to match or exceed them. Making these models profitable is proving extremely difficult in the face of so much competition and such unsustainable expectations being set. Google seems to be most likely to sustain themselves through this melee. Them and the Chinese companies.
It feels like we're watching a hype cycle in real time
Cursor just committed mass consumer fraud at worst, and at best pissed off all their best customers. I feel really sorry for those who invested at a 9bb valuation.
> I feel really sorry for those who invested at a 9bb valuation.
Because they didn't do their jobs properly?
>Cursor just committed mass consumer fraud at worst, and at best pissed off all their best customers.
What happened?
flip-flopping on pricing has led users to feel nickel-and-dimed
i like cursor fine, but check out the forum/subreddit to see people talking like addicts, pissed their fix is getting more expensive
i think this aggressive reaction is more pronounced for non-programmers who are making things for the first time. they tasted a new power and they don't want it taken away.
I agree with your take, but I still don't excuse anti consumer practices like that. It annoys me because this is a repeat problem in this space, where these companies don't take into account the market dynamics, or costs of their service. From the start I've been looking at these $20.00 subscriptions, and then my own personal api per token costs and been wondering how they aren't all bankrupt.
Look no further than founders in the sports betting space, like the fanduel founders. Borrow a bunch of money at huge valuations because of hype and ignore the fact, that despite it being exciting and popular, the margins are like <5%. Fanduel founders sold for 400 something million, walked away with nothing. Its now a multibillion dollar company when the new owners realized the product was marketing, not the vig. These AI companies are shifting towards their "marketing" eras.
I think any "value add" business that has a primary product built on top of another larger business' non-commodity service(s) runs the risk of having to re-do their pricing in ways that are outside their control.
This is nothing new. I'm not sure if it's "anti-consumer" as much as it's just a risky play from a brand and customer happiness viewpoint. Because your prices can be forced up by your supplier, and your customers will be mad at you, not at your supplier.
I do also think it is on consumers - in some part - to go into it with eyes open and do their research.
Thankfully a product like Cursor is a monthly sub and not a big up-front investment so if you don't like - or can't afford - the new pricing, you can just stop paying.
I think my primary concern are the customers who paid $220 for a yearly subscription, only for it to get nerfed and asked to pay $200.00 a month to get similar output.
What surprises me is just how much they've missed the mark.
I'm not an extreme user of Cursor. It has become an essential part of my workflow, but I also probably on the lower/medium section of users. I know that a lot of my friends were spending $XXX amounts/month on extra usage with them, while I've never gone beyond 50% included premium credits usage.
After their changes I'm getting hit with throttling multiple times a day, which likely means that the same thing happens to almost every Cursor user. So that means one or more of:
- They are jacking up the prices, to squeeze out more profit, so it looks good in the VC game
- They had to jack up the prices, so that they aren't running at a loss anymore (that would be a bad indicator regarding profitability for the whole field)
- They are really incompetent about simulating/estimating the impact of their pricing decisions, which also isn't a good future indicator for their customers
My guess is that it's your second scenario there (avoid running at a loss). In the start-up game scale/growth is the most important thing and profits really aren't that important. you want to show to later stage VCs that your idea has traction and there's a large addressable market.
Whilst profits aren't important you also can't burn all your current capital, so if the burn rate gets too high you have to put up prices, which seems likely to be what Cursor is doing.
If the people are still paying the increased prices, it's a success. Rugpulling literal addicts is a great business model. Remember that business profits are primary, not consumer opinion.
I want that entitled attitude to spread. Destroy profitability.
That's where the real test lies for Cursor and programming LLMs.
Will users feel that a $200 subscription is worth it or not?
I think what everyone, including those programmers advocating coding agents, are forgetting is that if you can have a full time programmer for $200/m, then that becomes the new value of programming labour in the open market.
IOW, the market will slowly but surely drive the labour rate for programming down to the cost of the cheapest coding agent.
So, sure, boasting about a 10x speedo on boilerplate has good metrics, but let's not delude ourselves that programmers are going to be paid enough to afford the $200/m coding agent in the future.
Hooray!
So how do I fight this as a programmer?
Because I have no interest in devaluing my skills with this crap, but unfortunately many others are all in!
> So how do I fight this as a programmer?
I wish I knew. I get a lot of joy out of programming and almost none from copy/pasting.
> Because I have no interest in devaluing my skills with this crap, but unfortunately many others are all in!
We are certainly living in Interesting Times.
No I’m a programmer and I’m better about the rug pull also.
I seriously cannot keep up. I fell a bit behind and now I feel I need a primer to know who owns/acquired/developed all these additional things surrounding the ai space
That information is as important as knowing about soap opera characters.
That is refreshing to hear and a great perspective shift for me, haha
Not so, you need to know how long you have until they shut it down.
I really hope Sonny doesn't go through with killing Marco!
What was Garry's post?
https://x.com/garrytan/status/1941553682736439307
The thesis is that once you’re paying $200 a month, you’re beholden and won’t pay and compare it with anything else.
Until something else comes around for similar price and is much better.
Good thing for consumers who use AI coding tools is that there is no lock-in like in Photoshop or similar software where you hone your skills for years to use particular tool. Switching from Cursor to any other platform would literally take 10 minutes.
I have same question
Gary gives off a grifter vibe to me. Such a shame seeing how YC has fallen
He blocked me (a relative nobody) on X for remarking on the number of people I know who’ve made it to YC on completely fraudulent credentials.
His reaction seems entirely appropriate. He could ignore you, but then you might keep replying to his posts and potentially spread incorrect but damaging information. He is losing close to zero by blocking you, but preventing a potential big loss. Why did you make that remark, if not to damage YC's reputation? This does not seem like the correct approach, if you wanted to improve their selection process.
The hilarious part is that I never interacted with him directly at all. I was just commenting something to a mutual on X, the thread blew up, he snooped it, and went on a blocking spree. It may have been different if I were directly accosting him in replies to his posts.
> Why did you make that remark, if not to damage YC's reputation?
Seems harsh and cultish to assume malice. He didnt say you parents have false credentials
I would say calling out people and institutions like that is important so as to keep them honest, and if they arent honest and are trying to grift/defraud people then they deserve the reputation loss
> He is losing close to zero by blocking you, but preventing a potential big loss.
Thats great for gary, but the rest of the world isnt there waiting to be optimized for his benefit. If people trust YC to incubate good talent, but feel its becoming a hub for grifters, then some accountability is in order. Institutions are beholden to their public stakeholders, even private institutions, because they still have people who are using and supporting them
Well, incorrect or not, now that person has a very strong motivation to talk bad about YC. Smart "reaction."
This whole situation feels shockingly close to the Meta/Scale situation, where founders and specific employees were plucked out, and effectively gutted any future prospects for the company.
At least in the Scale case there seemed to be some form of payout to employees and equity holders, but this takes it a whole lot further by just throwing out all other employees.
There is supposed to be the concept that “all common stock is the same”. These fake-acquisitions completely undermine that.
Yep, if investors and early employees keep getting left out in the cold while execs get a soft landing at Big Tech, it's going to shake a lot of trust in the startup game
I don’t think anyone trusts any tech company much these days. It’s been a steep decline in the past 5 years, from arbitrary mandates to the constant talk about firing everyone and hiring an AI. Even as an investor, it’s hard to trust that the “honor system” that once existed is still in play.
So Google, Meta, and Microsoft will just hollow out the best AI startups of their talent instead of buying them - out of fear of monopoly lawsuits I'm assuming?
Nice plan I guess. Kind of obvious to spot though.
Likely cheaper too. Nothing to pay the original shareholders
Can shareholders sue? I presume the only avenue is IP since that belongs to the company? Or the non-exclusive license somehow negates that? Brutal.
I actually don't know if there's much that can be done unless there's some non-competes in those employees' contracts which are usually not very enforceable outside of finance iirc.
Non competes aren’t enforceable in California but the company owns the IP so I’m curious about this license loophole they are using.
Non competes can definitely be enforceable in California for executives and those with fiduciary responsibilities to a company.
They’re just not enforceable against “rank and file” employees.
The only situation I know of is during a sale of business if the seller agrees. Which is clearly not the case here.
Is there any IP that's actually valuable without the team ? I sincerely doubt it.
The whole point of funding a company is for the company to build IP that makes the company valuable. Founders can't take investor money and then just go start another company -- that's specifically barred in most docs. There have been a lot of these weird "loopholes" lately that are completely against the spirit of company building.
The real IP is between the ears...
“We underpaid you relative to what price you were able to command on the market, and you left, how DARE you!”
“We spent our time and money helping you and now you leave taking everything with you and leaving us with nothing”
You think the only people in a company that matter are a few founders? It’s ok to screw over everyone else?
I’m honestly just surprised that the CEO and co-founder decided to walk away from the company and leave behind all these employees he was leading. Especially considering many of them probably joined for lower pay, hoping for a big upside.
Maybe there’s more to the story.
When you want to make a big impact for a big payday, why would this surprise you?
Gentle reminder that more startups die by suicide than homicide, and that an early-stage startup is a total crapshoot.
Yes, startups are always a bit of a gamble, but this feels like a captain abandoning ship while it’s still full of sailors (many of whom have families depending on them).
This really is a whole new level of getting screwed.
This is why advice is always to treat options for a non-public company as if they're near zero in value.
Because for most people, they will end up being worth exactly zero in value. Less if they went and exercised those options prior to a liquidity event that may never happen.
Isn't this the case for pretty much every startup that gets sold?
Founders get a big pay day and leave within a couple years while 100 employees share a 1% of the company between themselves.
The difference between 1% and 0% is painful.
rule 1: never believe a word a founder says
You're surprised that a CEO did something that massively financially benefitted them personally at the expense of rank and file employees?
You sweet summer child.
It's been working with software developers with no issues.
“Buying the startup” just means handing over megabucks to do-nothing investors. If Google isn’t buying any product or technology, why should investors get a talent fee?
The investors are getting paid in this deal. Just the employees getting screwed.
Do nothing investors who enabled the company to reach this point? Employees who chose lower salaries in expectation of shares being worth something? Come on now.
> Do nothing investors who enabled the company to reach this point?
Were you under the impression that venture capital is anything more than rent-seeking?
Very edgy so cool
Sure if you want to be negative about it and only look at the worst VCs. But the best VCs provide significant value outside capital and can be instrumental in a startups success or failure.
The big players know regulators are watching, so they're doing everything but the formal acquisition
This is the direct result of regulations. As usual regulations backfire. Expect more regulations to address this, surely they won’t backfire as well.
This is overly reductionist. The are plenty of laws that work well.
Any time I hear someone talk about more or less regulation, instead of talking about better or worse regulation, I suspect they are ideologists and trying to shift the narrative, or else they would be able to criticise based on actual merit.
There are many AI startups and we are just in the beginning of learning how to use them. There will be some stupid company like those you’ve listed that figures out a way to use AI that is far better than any other implementation, and Google, Meta, and Microsoft may go the way of Yahoo and AOL, but we’ll see
Doesn’t seem like it. Antitrust has no teeth so the mega corps are just buying all the talent with life changing cash.
The “talent” is not very talented, trust me. These are the short term whims of very large, increasingly bloated organizations. A leaner startup that knows what it has will not sell so quickly. At least, the odds will soon be in favor of whoever first decides to take that bet.
I never knew anyone who used Windsurf. These AI acquisitions have been unbelievable(in a bad way). WIX acquired some garbage Lovable.dev clone for 80 million. I think many of us are waiting for this bubble to pop(economy will likely pop too)
It was barely better than Cursor and they got shafted by Anthropic because of the takeover announcement so nobody really used it anymore because let's face it - Claude Sonnet is just the best coding model. Design-wise the chat panel and autocomplete integration was a bit nicer than in Cursor but not by much. Subscription for Windsurf was/is also 5$ cheaper.
i don't think it was better than or comparable to cursor at all. except for the month prior to the OpenAI Acquisition news where some minor influencers on X were calling it better.
if it was better it would have survived.
> if it was better it would have survived.
Not sure how you can claim this when:
1. It is still very much alive, and
2. The whole point GP is making is that what made it better got stripped from it because of the acquisition announcement
Everyone has a niche, Windsurf is the only large provider if you are a Jetbrains shop.
There are some alternatives like continue.dev or Jetbrains own AI offering but no Cursor or Claude Code ( Sonnet 3.7/4) you can get through Jetbrains plugin or others, but Anthropic does not provide support same with cursor.
Jetbrain's Junie works incredibly well. I much prefer it over cursor's or continue's UI.
What does it offer that's better than running CC with Pycharm/Jetbrains?
I just tried it out. It does feel better than CC for some tasks .
I think what Junie does differently it has better context on what the IDE knows or sees , from what files are open or typedefs are loaded, I could achieve some simpler tasks with lesser prompts and less tokens and cycles than with CC or even windsurf.
Jetbrains has the best incentive to integrate deeply I imagine, it has always been question of execution, it seems a fairly robust system.
I agree with you on the integration.
The biggest problem for me is that I can't link my Claude Pro (or Max) account to it. So now I'm paying for two LLM accounts (even if I used Junie for coding I use Claude for other things, research, etc., so wouldn't drop it).
Been using AugmentCode on Jetbrains. By far the best. I tried Junie and Windsurf
claude code has a Jetbrains plugin which is delightful!
Seems a recent launch in beta just in June .
Thanks for the share !
Check out sweep. Completely unaffiliated, their only offering is the jetbrains plugin so it gets a lot more focus than windsurf. Only downside is that Claude code is still a better agent, but at least its tab complete is some of the best
GitHub copilot now has agents in jetbrain (not sure about stable - my nightly does).
Jetbrains Junie is supposedly the same thing but no Rider and that's my current project so didn't get into that yet.
Windsurf was just disappointingly bad in intellij (like any other plugin I've tried so far)
The copilot agent stuff in IntelliJ works relatively well in my experience, they managed to implement a quite cursor-like “accept/reject” UI in a plugin, you know, forking IDEA. There are some areas like getting it to use git tools where cursor works more smoothly but you can coax Copilot into producing the same results. I’m just generally happier working in IntelliJ vs VSCode so I’ve tended to favour Copilot.
Never tried Windsurf in it’s recent form but we did evaluate it when it was still called Codeium and everyone liked Copilot better.
Augment Code is great on JetBrains
It’s the modern day dotcom boom and all these agents are the modern day Dreamweaver.
Windsurf was also used by enterprises because of their on-prem plan. They gutted that after OpenAI acquisition was announced and since then I am sure none of those enterprises that used it will switch to their cloud offering and look for other venues.
I was on Windsurf's grandfather $10 per month plan, it was really good during the Sonnet 3.5 and 3.7 days
I am still a paid subscriber but most of my usage is claude code now becaue Windsurf does not Sonnet 4 included in their plan.
I used it for about six weeks in the spring, at which point I tried cursor and found out that it was far better for my use case (backend Go): better and faster autocomplete that can work across files, less buggy UI, and it actually follows your rules.
I’m not bothered by this news — it’d be pretty shortsighted to assume that any of these disposable AI assistants will continue to exist forever.
The first time they hit the news, I’ve tried to open their website to see what it was all about and it froze my phone lol
Well I use Windsurf. It's a good alternative to GitHub Copilot. The free tier is on par with Copilot's paid plan.
...which no one talks about anymore. Okay I guess you have a point.
Base44 is absolutely not garbage. I’ve tried it and can say it’s as good or better of a vibe-builder than Lovable or Bolt. Have you benchmarked it against the competition or can you otherwise substantiate the “garbage” claim? FWIW I do know one amazing engineer using Windsurf
all those projects are garbage and just create half bake prototypes that never see the light of day
Agree in principle, but when evaluated against the competition and likely acquisition targets of Wix, it's certainly not garbage. I've seen it vibe code an entire app that was -- admittedly mostly working -- and deploy it with a prompt of 5 words, in about 2 minutes.
All of this game of thrones is going to create an amazing documentary if AI capabilities taper off and valuations vaporize.
Are the AI capabilities tapering-off, or commoditized? Building the next Windsurf (iteration 0) doesn't feel it's quite niche anymore.
I think the current valuations imply at least 2 magnitudes of improvement over existing functionality.
If by Documentary, you mean a new Silicon Valley sitcom, yes , all the ingredients are there: The AGI believers, the doomers, the "cure all diseases" people, the board drama, the money grabbers, the VC dance , the poaching, the lawsuits for copyrights ... there s a whole new universe of caricatures
Even a very risky attempt at "cure all diseases“ is worth all this economic upheaval, though.
And AI applied to biomedicine arguably already delivered some acceleration.
I know David Fincher is jumping on his seat
Obviously these things are difficult to tell from the outside
Apparently somebody missed crypto mania between 2019-2022
AI has nothing in common with crypto other than it being hyped a lot. The better comparison is the dot-com bubble.
> AI has nothing in common with crypto other than it being hyped a lot.
Don’t forget all the GPUs. Nvidia always gets its cut.
How did I forget about the GPUs! I have made a grave mistake, please forgive me.
Well it's also being slammed into everything everywhere, its just more useful so it has even more places where it's being put
Gary-Marcus-eating-popcorn.gif
> if AI capabilities taper off
AI growth has slowed to a crawl, and it's priced it self out vs cost of compute.
NVIDIA feels a lot like SUN.
> amazing documentary
Been there, done that: 2001, Startup Dot Com https://www.youtube.com/watch?v=cP4PGjnZwJE
Works out for Google and the C-suite. Horrible for the employees. These fake-acquisitions are effectively arbitrage against employees, who get left holding nothing. Should be illegal and regulated.
Not sure how the VCs get their cut. I'm guessing that Google can balance it out by participating in rounds for other startups in that VC's porfolio.
It is always the case that the employees get screwed
I recently read a reddit thread about how somebody made 12M on a stock that IPOd. But yeah the majority opinion is that employees get shafted.
Windsurf and Cursor are in the business of reselling ChatGPT and Claude at a loss, but the tech itself is not impressive at all
Those wrappers are gonna go away now that there's Claude Code and Googles CLI thing. They are that much better.
Are they?
Cursor's Accept / Reject feature for each change it makes in each file is nice whereas I have to use a diff tool to review the changes in Claude Code.
Also, if I go down a prompt alley that's a dead end, Cursor has the Restore Checkpoint feature to get back to the original prompt and try a different path. With Claude Code, you had better have committed the code to git, otherwise you end up with a mess you didn't want.
My company pays for both, but I mostly use Cursor unless I know I am doing a new project or some proof of concept, which Claude Code might have an edge on with a more mature TODO list feature.
Gemini CLI uses a shadow git repo and commits after every change, won't be long before Claude Code has that too.
That’s a neat idea!
None of these features are very deep though, there’s dozens of OSS clones for them already.
RooCode/Cline, etc
I got burned too many times from that Restore Checkpoint thing not working right, maybe it's been fixed by now but seems silly to rely on something thats not a literal tool built for the job (version control), not a good shortcut.
It has worked perfectly for me every time, and it’s such a great feature.
I agree. I use claude desktop with MCP and Gemini CLI exclusively. I have 20+ years of writing code, and this is awesome!
I was so surprised (or shocked) to hear that Windsurf was getting acquired for 3 billion dollars, I made an HN post asking about the truth of that news - https://news.ycombinator.com/item?id=43933825. HN's system didn't like my tone I guess and removed it, lol.
But in any case, I just can't see how AI code editors like Windsurf or Cursor, without any proprietary model, can be valued at billions. What's the underlying IP that justifies these valuations?
Similar thing to what we’ve witnessed with crypto coins. It’s AI season and those with money invest in it, pump it and will exit post IPO. Difference here is, that besides value that those products “hold”, it’s possible also to provide AI as a service, making Google / Microsoft etc interested.
> I just can't see how AI code editors like Windsurf or Cursor, without any proprietary model, can be valued at billions.
Windsurf used to be GPU rental company before they decided to make a text editor. I imagine they own a lot of silicon they are running their coding assistant on.
https://www.afp.com/en/infos/windsurf-launches-first-gpu-clu...
It's not a rinky-dinky startup that just forked VS Code in a month.
you meant Exafunction? Nothing of Exafunction was discussed in OpenAI or Google deals.
Windsurf did start as a rinky-dinky starup, just as a plugin (Codeium). And in its current state, it still has no moat.
The only reason big tech leaders are throwing money at this is - they need to show they are doing something to stay ahead. And right now they are going on a hiring frenzy as a show for their investors.
I have been there when shopping businesses were being bought for billions of dollars, and then they all got bundled up into a mess swapped under the rug silently. I don't think this is going to be any different.
Codium was built on Exafunction infrastructure.
https://www.unite.ai/varun-mohan-co-founder-ceo-of-codeium-i...
It's really a simple story, they built infrastructure years ago, initially they rented it, so companies outsourced their computing tasks to them. But eventually came up with the idea what product could they put on top of their infrastructure to sell it for dollars rather than pennies. And they did.
They don't need a moat because they provide turnkey utility to a lot of very rich corporations. They are not in the business of selling the vs plug-in to you. They are in the business of selling developer experience powered by their own vast compute to a lot of very rich corporations.
They had millions in revenue before they hired their first sales rep. I don't think just a VS plug-in startup could do that.
https://www.saastr.com/cro-confidential-how-codium-built-a-b...
They sell stuff that actually works, and people who use it convince people who pay money to pay for it.
They both have proprietary models.
Anyone can build a proprietary model in under a week; but the point is - Cursor and Windsurf don't have proprietary models that remove dependency on external models, and/or have capabilities that competitors don't have.
Maybe Google sees something under the hood
> Google will instead hire Windsurf CEO Varun Mohan, cofounder Douglas Chen, and some of Windsurf’s R&D employees and bring them onto the Google DeepMind team, [...] Google will not have any control over nor a stake in Windsurf, but it will take a non-exclusive license to some of Windsurf’s technology. [...] Google didn’t share how much it was paying to bring on the team. OpenAI was previously reported to be buying Windsurf for $3 billion.
Why not an acquisition?
How did Google get Windsurf and investors to agree to this maneuver that decapitated the leadership and key talent, without a big exit event for everyone?
My read of the article: "Here's x% of what OpenAI offered you, you waive legal challenges while we cherry-pick your people and license the tech in their heads, and you can keep the company, and everyone left behind can promote themselves to fill the vacancies."
If they acquire a company they might need approval due to anti-trust.
If the people instead just quit their jobs and start working at Google … nothing to see here.
https://medium.com/@villispeaks/the-blitzhire-acquisition-e3... explores this some more. The remaining employees need to make it look like the business is still viable
Thanks, that's a great article.
I wonder whether the left-behind small-fry employees make out comparably to how they would with an IPO or big acquisition.
Or if the deal is more like: "We'll pay you double what you could make elsewhere for the 2 years we want you to Potemkin Village, or you can walk and get paid zero; we don't care, because enough people will stay."
And everyone who didn't get a Google employee badge agreed because "x%" was big enough?
I guess VCs can't force founders to stay (the only penalty for joining Google is loosing some/all of their Windsurf equity, but I'm sure they chose what's better for them), and employees didn't need to agree (they have no vote).
I can guess a few angles and causes for legal action. I'm wondering what the deal was to incentive people not to take that legal action.
I'm not surprised. I started using Windsurf when it came out because I liked its UX better than Cursor's.
However, while Cursor and GH Copilot improved, Windsurf went in the opposite direction. On each update, I started to get more and more issues. The agent often tried to run shell commands, and it hung up, or I found minor UI bugs. One day, I decided to give GH Copilot another chance, and I was surprised by how it evolved, to the point that it worked better than Windsurf for my usage. I don’t know what happened internally at Windsurf, but I notice the degradation as a user. If my case indicates what happened to other users, maybe OpenAI saw declining subscriptions and canceled the deal.
It's unclear if OpenAI cancelled the deal, or Google poached them? Either way, this season of "OpenAI Drama" is wild. First Meta, now Google. Your turn Amazon / Microsoft.
Apparently OpenAI allowed the deal to expire; likely Google had already been in discussion with Windsurf as I'm sure they knew the deal was likely to die well before today.
MS probably killed the deal, MS wanted access to Windsurf to make Co Pilot better while OpenAI did not want to give them access.
Update:
> Google hires Windsurf CEO Varun Mohan, others in $2.4 billion AI talent deal
https://www.cnbc.com/2025/07/11/google-windsurf-ceo-varun-mo...
> OpenAI’s deal to buy Windsurf is off, and Google will instead hire Windsurf CEO Varun Mohan, cofounder Douglas Chen, and some of Windsurf’s R&D employees and bring them onto the Google DeepMind team, Google and Windsurf announced Friday.
> Mohan and the Windsurf employees will focus on agentic coding efforts at Google DeepMind and work largely on Gemini. Google will not have any control over nor a stake in Windsurf, but it will take a non-exclusive license to some of Windsurf’s technology.
Sounds to me like they're "hiring" them like one "hires" a consultant?
Why the quotes? Consultants are indeed hired for consulting work to be done.
Wasn’t meant in any negative way, just ESL.
I don't know anyone who heard or used Windsurf outside the Bay Area. Even Cursor feels very Bay Area bubbly (although that is the market to go after if you're in ai dev tools).
Cursor does add value but it's just a thin layer on top of VSCode so companies could just build that in-house and don't need to acquire. There's no moat there.
Cursor has custom tab and embedding models. And has a lot of distribution / paying users already.
Arguably they have the strongest product moat, and I wouldn’t be surprised if they beat OpenAI in a vertical coding model from that. Easy for them to have users generate evals and have model product feedback loop here.
The tab completion is fast and the best available right now but is still so garbage that I turn it off 99% of the time because the suggestions are mostly noise.
I have the opposite experience, it’s at least 90% correct. For example, if I start writing the name of a function that I just added in a different file, tab will suggest the function, then jump to the top of the file to import it. If I’m changing the way something is called in 5 places, if I change it in the first place, tab will jump to make the same change in the other places. It’s honestly pretty spot on.
Zed tab is a lot worse in comparison (partly because it’s slow)
> I have the opposite experience, it’s at least 90% correct. For example, if I start writing the name of a function that I just added in a different file, tab will suggest the function, then jump to the top of the file to import it
Intellisense has done this for ages. A decade at least.
Before being known as Windsurf it was Codeium - the only good free autocomplete extension for VS Code and Jetbrains ides.
I did not see this coming. Wow. The game of thrones in SV.
I wonder what happened with the OpenAI deal. Anyone have any guesses? My first guess is "Look at Claude Code, we can do this ourselves." But, I am likely thinking too simply.
edit: does this mean that Windsurf and its users will stop being iced-out by Anthropic? Or, is this the end of Windsurf?
Derp. Weird IP sharing issues.
> I did not see this coming. Wow. The game of thrones in SV.
You must be new around here.
I have been using Windsurf for few months. They even have their own AI model SWE-1 model. I really liked using Windsurf. They also have integrations with other IDEs ex: jetbrains, VS code, etc.
This week I have been using Claude Code and Windsurf side by side. I would make change with one, stash it, ask the other for similar change and then would diff it.
Overall Windsurf was pretty on a par with Claude code.
Funny to see this today.
I'm a rank and file dev at a non-big tech company and I got a call from a Windsurf sales rep this week who I had connected with on LinkedIn the day before (I never gave them my number). They told me my company was in talks with Windsurf about a licensing deal but that they would give me a 30 day trial of an enterprise account for use on personal projects to let me try it in advance. I guess the idea for them is to build enthusiasm among devs in the company?
Is this a standard sales strategy for products like this? It seems pretty aggressive to me but I'm just an engineer so I wouldn't know.
Very standard yep. Sales folks are sort of trained /indoctrinated into telling white lies like that in order to get in the door. There are loads of examples of using fake momentum to close deals. If its a senior person it’s “My CEO asked me to personally reach out to you” or a fake email from the CEO forwarded by the rep. If one person at the company uses it, it’s “we’re negotiating a company wide license” or “we already have a group license with extra seats” or “one of your teammates sent us a list of priority teammates” yada yada.
Nice to get a sanity check that confirms Windsurf was never really worth $3B to all those who thought that number was ridiculous.
Google is paying $2.4billion to them
Are these "acquihire & license" the new M&A...? I recall hearing that this was a "hack" to avoid DOJ and FTC scrutiny over acquisitions, but I have no clue how such deals are structured. Anyone care to chime in?
OpenAI needs to up their game on Codex to be on par with Claude Code. o3 is a better planner relative to Opus.
Where do you think Codex lags behind Claude Code?
The big one is that they do not offer "unlimited" plans where you can forget about the tokens and just use it.
UI is also worse compared to Claude.
They still have some work to do if they want to compete with Claude TBH.
This deal always looked strange in the first place. The usage of Windsurf was significantly lower than Cursor and Copilot and somehow it was worth $3B.
Given the release of Claude Code, it was already over for them.
For the love of God, can we get a reboot of the Silicon Valley television show? Just on AI. Like when they wrapped it, they wrapped it on AI usage. So, it's got the perfect arc for a reboot that focuses perfectly on AI.
The "Son of Anton" was an coding agent that deleted the entire codebase. Not so far off from Cursor's YOLO mode, is it?
https://news.ycombinator.com/item?id=44262383
Why ? we get to watch the original reality show in real time, for free!
Even the original Silicon Valley didn’t match the zaniness of real life. Why do we need a reboot? Just check HN!
My favorite thing about that series was watching it with friends who weren't from the Bay Area. Often they'd be laughing at the sheer absurdity of a situation, and I'd get to point out that it was barely exaggerated from real life.
My wife refuses to watch it because it hit too close to home.
That's what my friends not from SF said. "This is insane, this would never happen"
Dude, I saw a lot crazier things happen on a monthly basis. And don't even get me started on the personal lives and partying that the show didn't display.
There’s a lot of talk about Claude Code in here, and I agree it’s a great agentic coding tool. One of the benefits of Cursor & Windsurf is/was the ease for smaller companies to setup Team accounts and have control over spend.
Claude Code I think misses this. You can get an enterprise account if you commit to over, what.. 70 seats annually?
If you’re an individual you can get Max 5x/20x ..
But for smaller companies, I don’t think they are addressing that space. Am I wrong? Are there any Agentic tools like Claude Code that can provide a fixed cost per user?
Enterprise is where you make the money. I’m sure they will have something soon.
For now, i can just piggybacking on bedrock or vertex which is peanuts comparing to production cost.
> piggybacking on bedrock or vertex Can you elaborate? I’m not familiar with this approach.
A small company can just pay the $250 a month for X number of employees to each have CC max plan. Not that complicated
I thought that was against the terms of service from Anthropic and they have the rights to disable/ban accounts doing that.
Kilo Code CEO here. We’d like to welcome ex-Windsurf users by offering you $100 in credits. :D We'd love to show how through open source there is a better way (better community, more transparent pricing, won't mess with your or the product). https://blog.kilocode.ai/p/windsurf-is-over-switch-to-open-s...
This certainly aligns with my own usage. I'm currently using OpenAI's own Codex 50:1 compared to Windsurf. For me, I'd rather take some time to create a good quality prompt and have it work away for a few minutes and create a material delta. It isn't always perfect, and I often have to make a few tweaks myself, but it is much nicer and waiting around and watching Windsurf bang around on a tiny part of the solution. Windsurf is still nice to use for quick UI iteration however.
There are probably some extremely unhappy VC's, angels and employees over at Windsurf right about now.
Windsurf's Funding Rounds
Windsurf has raised a total of $243M over 2 funding rounds. One was Series C round of $150M from Investors like General Catalyst and Kleiner Perkins and the other was Series B round of $65M from Investors like Kleiner Perkins and General Catalyst. Here is the list of all funding rounds of Windsurf: Round Details Series C, $150M, Aug 29, 2024 Post money valuation - Revenue multiple - Investors General Catalyst, Kleiner Perkins, Greenoaks Round Details Series B, $65M, Jan 30, 2024 Post money valuation - Revenue multiple - Investors Kleiner Perkins, General Catalyst, Greenoaks
> Bloomberg reports that Google is paying $2.4 billion to license Windsurf’s technology and hire its top employees.
Sounds like they're getting a payoff but probably not what they had in mind with the much larger numbers being thrown around for AI.
Pour one out for the regular employees not getting absorbed by Google and suddenly not millionaires like they imagined they were a week ago.
Maybe the expectation that a job leading to an equity windfall is something people should be more cautious about.
It's something you should never assume is true until the wire hits your account. I had a deal where I was going to make $15 million called off 36 hours before closing.
Aw man, I'm really sorry to hear you had to experience that.
Like in WeCrashed (2022)?
This my friends is how the next iteration of venture capital contract templates becomes even longer...
Otherwise, normally with the amount of capital raised by Windsurf, the founders must have signed some kind of non-compete for the event of a bad-leaver (which this obviously is). Guess covering these penalties was just part of Google's deal, hm?
How does this happens?
They raised A, B, and C round (according to CrunchBase), and then the founders just walk away and get a job/deal at Google?
Perhaps it as combination of how much founders were diluted and how much they are being offered upfront. We are hearing about $100M signing bonuses.
It is hard to say no when Google/Meta gives you say $100M upfront and hundreds more if not Billion+ in RSUs. After 3 rounds it is not unreasonable to have only 5-10%.
10% of a company worth a few billion burning a lot of cash, that needs to keep raising more rounds i.e more dilution, may have less value than RSUs from multi-trillion dollar publicly traded liquid tech company today.
It is also quite hard to raise $5-10+Billion in cash. There are only handful of startups which have ever done so
Very few funds/investors can afford to do so large rounds. This was SoftBank's thesis for most of last decade, compete by just outfunding competing products in a market.
The deal for the founders may not have been as good as what Google offered. They may only hold 10% after those rounds, a serious part of the acquisition price could go to liquidation preferences of the VCs and the deal is mostly in OpenAI stock instead of cash. Not that hard to imagine the Google option offering them much more actual cash right now.
Nepotism.
The same set of rules that apply to you and me are not universal.
How long does Windsurf go on now? Losing your CEO to a poach job not even an acquihire must blow up any fund raising plans.
When Claude kind of cut them off, they realized these AI Agentic tools are as good as your model, little to no moat here.
And it was a crazy deal to begin with, for reference JetBrains who's building IDEs for 24 years are evaluated at $7 billions
So the result of aggressively scrutinizing big tech acquisitions is acquihires, not a more competitive tech ecosystem with say more IPO’s.
The libertarian spin on this would be government should have never scrutinized acquisitions and the result is just worse for everyone.
The progressive spin would be to now ban acquihires somehow, and then whatever new legal invention will be created next. I can imagine the next step being, creating a consulting company out of your startup and then selling yourself as consultants to big techs. Now you are neither acquired nor technically acqui-hired and the whackamole continues.
At some point, we need to realize the solution is the culture of people involved. If the government could just ask to reduce acquisitions to make the ecosystem more competitive and companies tried following it in spirit to the best of their ability, we might have much better results than whatever we have now. When culture degrades, the govt can’t trust companies, the companies can’t trust the govt, everything just gets worse, regardless of what rules you write and enforce.
The culture of the people involved got us to this point, I’m not sure it’s the solution to the problem.
> The progressive spin would be to now ban acquihires somehow, and then whatever new legal invention will be created next.
Progressive has become a moving target, but the pro-competition view would be to break up the massively concentrated companies that are further consolidating markets. Thats what the Khan FTC was trying to do, but we need a Congress interested in a competitive marketplace, which we haven’t had in a while.
This wasn't a result of regulator scrutiny. The issue was that MS (owner of Copilot) was demanding access to the IP (due to their existing agreement with OpenAI), and OpenAI was resisting. In addition, Claude blocked access to Windsurf, which also damaged them as an acquisition target.
Nothing to do with regulators.
I find this hard to believe considering all the recent acquihires that happened recently like Character AI, Inflection, Covariant AI, Scale AI, context AI and so on. Maybe you’re right about the specifics of this situation, but my prior for this being an acquihire is very high and I would need to see very compelling evidence that that is not the case.
According to The Information, Microsoft gaining access to Windsurf’s IP if OpenAI acquired them was a factor: https://www.theinformation.com/articles/openai-windsurf-brea... (paywalled)
I don't understand why Windsurf would care after they've exited.
Not Windsurf… OpenAI. And OpenAI cares because they’re competing (in part) against Copilot, so if Msoft gets all the benefits of Windsurf then OpenAI would effectively be paying 3B to feed their competitor.
This would also happen if OpenAI developed the same thing internally, right? I don't see how not acquiring them improves anything.
I guess building it internally is cheaper
If that's true today it was true the day they made the original offer. Microsoft didn't change that.
Does that ip deal expire at some point?
Only if OpenAI declares they have achieved AGI.
I can’t wait for these companies to start laying people off so I can buy their house. They are inflating real estate prices with their dumb AI money.
I don't know if you noticed but cursors language server aspect that runs the coding edits and stuff like that from a server to the workstation is a lot better than windsurf.
Windsurf phone's home on every code edit that you have and takes on 30% load on your servers or on your workstation depending on what you're running.
I would strongly discourage the use of windsurf on your systems.
Case in point their AI model that they just built.
Honestly there's no value that windsurf, cursor and all the other VSCode forks provide that couldn't be provided as an extension and even then - none of them perform as well for agentic coding as Cline / Roo Code (debates about the subscription pricing aside due to people often not realising their model limits, public US only based APIs, pay for useful API limits etc aside).
This sounds terrible if they're just taking management and key employees?
Imagine backing this startup and the founder team takes a parachute...
It's clearly designed to avoid anti-trust
https://centreforaileadership.org/resources/opinion_startups...
https://archive.is/Rdt3z
Just curious - would this negatively affect OpenAI's ability to acquire companies in the future?
This isn’t a great look for OpenAI, but acquisitions fall through all the time.
The issue isn’t an acquisition not working out, it’s that the founding/exec team felt it appropriate to arrange their own exits and abandon their team before even communicating that their “successful exit” wasn’t actually happening.
They have got get their act together from a structure standpoint or these types of acquisitions are going keep failing.
Bullet dodged.
Windsurf's value to OpenAI was for the latter to "see the whole chessboard" of context, which is helpful when you're training models to be good at coding.
But codex (and Claude Code) fulfill this from the CLI, and it's a first-party utility, not an acquisition.
If this isn't some kind of sign of the times, idk what is. This is too far.
What could Google have possible offered? Must have been astronomical. Sundar has been aggressive ever since OpenAI knocked the wind out of them by being first release gpt.
Wow must have seen the numbers and decideed they wanted to call it off. Not a good business model probably so the human talent is where you find the best amount. Still why would anyone ever want to work at Google? Don't they know they are contributing to a system that covertly disseminates information they want you to see. Especially the AI models. Has anyone ever wondered about the training data sometimes? What would people think of them if they knew they had the entire pestein list in their hands but decided its better to protect the ones that pay them. People need to reconsider what they believe from AI, it can be extremely abused to scale narratives.
It's somewhat telling that the most valuable part of an company is the people, some things don't change
Wow, this is a pretty fascinating twist. First OpenAI's $3B deal falls through, and now Google swoops in to poach the key talent anyway? Classic big-tech maneuvering
The founders fucked over the employees and the investors and sold out. I guess they don’t care if they are worth $200M each but they fucked every employee that poured their heart out into that company.
I hope no one works for them again.
This is why working for startups is not worth it.
I guess masks are completely off now. We can see who sells out to the highest bidder and who won't sell because they care more about the mission.
I really think that Apple is smart to sideline this shitshow.
This.
The title made sense until the comma, and then it didn’t :)
I wonder if this is a result of the previously reported clashes between OpenAI and Microsoft over access to the Windsurf IP (under their investment agreement)
Anyone know what the deal was? Can it be scrapped like that? I expected to read more info about that but it's not even mentioned.
Could be conditional on DD and deliverables
Sounds like the death knell for Open AI. They can't outswim the FAANG sharks. Once Microsoft is out, it's over for them.
https://archive.is/SkCKz
Is Lina Khan to blame for this new acquihire meta? She was very aggressive in blocking any tech acquisition during her time and ever since we have seen more and more acquihires which I believe these companies are using to prevent themselves from getting sued.
Google is having a hard time acquiring Wiz for 32b, and if it's blocked they owe 3.2b to Wiz. So why risk it when you can just spend the money to hire the talent behind it and spend a few month building out a new product.
The police is to blame for trying to enforce the law, it makes criminals innovate is exactly the kind of take I come here for.
I mean that's exactly what is happening right now. Companies are buying the assets of companies instead of the companies themselves so they don't have to worry about being sued which is a timely and costly process that has become very unreliable due to how much politics are involved in the process now.
Smart move, I always wonder if they have disposable money to spend on stuff like this and figure out what to do with it after.
Does anyone know which side cancelled the deal?
i would be so pissed if I was an employee who got nothing out of the deal and left to dry now
What if OpenAI is buying Cursor instead?
Not out of the question after a week of Cursor just absolutely torching goodwill
Could you elaborate or provide more context for those who don't use Cursor?
I'm assuming it's this:
https://news.ycombinator.com/item?id=44538243
Could anyone explain the implications of this for Windsurf as a company? Are they going to close?
Nothing for certain, but yes. The IP payout is to give the investors something.
Not every exit is a liquidity event these days. Startup employees take note.
and Windsurf employees have worthless equity and no CEO
loool dead
dang, I feel like my Bay to Breaker tote bag value just went up 10x right?
What is the source Verge? Give us a link more than "Google and Windsurf announced Friday"
I wonder how these two events came to be declared in the same news release.
nonexclusive proprietary licensing at its zenith
Zuck swoops in and hire them 100mm a piece.
Looks like Sama can’t catch a break.
I only learned this week that “sama” is “Sam Altman” and not the first name of some other ai startup ceo
It's his hackernews username.
Well, more actively nowadays, it's his X username...
sama is nothing without drama
Good.
Lol.
I commented on the OG thread something like "weird since MSFT owns VS Code" and got downvoted to oblivion.
Yet here we are, always right :).
https://archive.today/urwCT
@dang - The title’s wording suggest that OpenAI’s CEO is leaving, not Windsurf. A more accurate title might be: “Windsurf’s deal with OpenAI is off, and its CEO is going to Google”
Ok, thanks! (Submitted title was ""OpenAI’s Windsurf deal is off — and its CEO is going to Google"")
Can someone explain how this works financially for the acquihired? I know they aren’t joining like a regular employee with a high TC. Does Google offer them a giant multi-million (billion?) dollar signing bonus? Why would they tank the value of the company they own just to be another employee at Google?
C-level executives get paid. Labor gets stuck grinding at Google. What a waste. Google will probably shelve/hoard the IP from Windsurf.