The main point everyone seems to be making is that now with AI anyone can make a SaaS.
The initial reaction I think most people have to this is "SaaS companies/devs are in trouble."
I actually think the opposite is true.
With an outpouring of vibe-coded apps/SaaS, you have the new wave of vulnerabilities/leaks/problems that happen even with the best software. Except now, it's worse because it's being done on platforms "built" by people who haven't the slightest clue how they work.
This I imagine will, over time, erode trust in most apps/SaaS products. With that erosion of trust will come skepticism and with that, will come trust in the "old faithful" of SaaS products/companies. Basecamp is a good example of this.
I could be wrong on this one, but it seems to me those that have built credibility for privacy/security/competence will become more valuable in the AI age, not less.
This is basically why I buy the tech dip. When you pay for software, you pay for infrastructure, expertise, QA, consumer relations, having staff on call, etc. It was always possible to replace enterprise software by 2 guys coding a product in 6 months, but you still need everything around the code before serious clients will want to work with you, and at that point you're a regular software company. All these vibe coded products are one untested push away from getting dropped.
You know what's funny, less than a week ago I signed up for Basecamp.
Could I have asked Codex/Claude to whip me up a Basecamp clone with the exact features I want?
Of course. Do I want to deal with managing that codebase, even with AI? No.
The problem has been solved and the $15/mo. is well worth the time I will save not having to deal with managing that codebase and can instead focus my attention on things that bring in revenue.
SaaS companies/devs are in trouble - but for slightly different reason. That was the case already for something like 10 years.
Earlier if you had developers and no domain knowledge you were able to land a contract building application for a company and maybe spin it off to get more customers in that niche.
If you got lucky and you landed law firm and made case management for them you probably had nice little niche.
But as it turns out lawyers can also use JIRA, Trello, Basecamp or whatever and they really don't need Facebook for lawyers so those gigs dried up.
Main point is, software development alone is not going to bring as much money as it did earlier. You will have to have backing of domain experts to get the business going to offer something special in your SaaS. Like possibility to actually have call with those domain experts or their oversight on whatever it is you are doing but you not having budget or enough work to hire domain expert full time.
This is essentially what Jensen Huang (Nvidia CEO) was predicting a few months ago. Incumbents in most software spaces will probably see a lot of short and medium term benefits from the new tooling as being trustworthy and truly understanding the problem space.
> The main point everyone seems to be making is that now with AI anyone can make a SaaS.
I agree with you that is incorrect.
With AI, not everyone needs a SaaS.
They can make a bespoke tool for themselves with 5% of the SaaS features they actually need. If it's only used by authorized, internal, users and never exposed to the outside, many of the risks you mention disappear.
That's not to say everyone will vibe-code their Slack replacement, but a bar for relying on an external SaaS vendor will go up (and I think that's a good thing).
The default has been pay $x/month for every service. I've seen startups that require a dozen service accounts just to run the software, and dozens more to get onboarded org wide. One service for feature flags. One service for logs. One service for traces. One service for error handling. Another service for ticket tracking, which is completely separate from your planning, design, and CI services. Jesus. What do people hope to accomplish here besides just defering blame?
Replacing SAAS isn't about building a replacement services 1:1. It's about figuring out what you actually needed in the first place! Often we only use a tiny fraction of what the full-blown SAAS offers. IOW it's about eliminating the service entirely and building something that fits your actual needs, rather than following what some VC thinks your needs are.
AI or not, the "build vs buy" pendulum is now swinging hard to build. And IMO that's a real opportunity to consolidate, trim some fat, and actually apply engineering practices rather than just blindly signing up for every SAAS that crosses your path.
This fundamentally misunderstands a couple of things.
DIY software is "free" like a free yacht is free. It initially looks appealing but there's a lot of expensive hidden costs and upkeep and pitfalls and problems.
For one, this is a bad assumption:
> building something that fits your actual needs
Unless your business is very small and not growing, this is a moving target. Your needs are going to change as you grow and different groups in the org are going to have different needs.
You really don't want to be dicking around creating software that already exists instead of doing the shit that actually makes you money. Spending a few hundo thousand on a bunch of software is nothing, you spend that on one engineer.
You buy a SaaS product because you have a problem and want to throw money at someone else to deal with it.
You're most certainly wrong on this one. Superior models give superior products and security over time. Until every 3-6 months stops bringing a large improvement in coding capability and scaffolding, there's no reason to assume we are nearing a hard limit.
You also have to factor in that bespoke software is... bespoke. ie. much more suited to your org's use-cases than the primary solution is. Way less bloat. Way less vulnerability when you don't need an enterprise SaaS solution and instead can host on your private networks.
And as far as security considerations: Imagine you had a separate Opus 4.6 agent tasked with managing and monitoring and updating devoted to a specific slice of vulnerabilities. Of course this is highly inefficient, but it would take care of the vast majority of vulnerabilities that even enterprise SaaS have. This is simply a scaffolding issue at this point, not model ability. Scaffolding issues like this will continue to dominoe.
>Until every 3-6 months stops bringing a large improvement in coding capability and scaffolding, there's no reason to assume we are nearing a hard limit.
How much of that is better models, and how much is it AI companies throwing more resources at each one? E.g. larger context windows and higher token/s correlate with the better models.
I also suspect AI is going to make software more secure rather than less.
Even today it can probably find a lot of issues automatically. With basic knowledge of what to look for, it certainly helps in understanding data flow too.
I think people fundamentally misunderstand what "non-techies writing software" means.
Nobody is going to ask Chat GPT to write them an app. They'll just ask something like "show me all the nearby restaurants that don't have any shellfish products for allergic reasons and let me browse like on ubereats", and Chat GPT will (eventually) do the right thing.
People will just see some UI, which will be like many UIs they've seen before. They won't care whether that UI is some UberEats embed designed by a programmer in California, or something that Chat GPT just came up with on the spot to wrap some API.
"Eventually, everybody is going to use computers" was a ridiculous thesis in the age of mainframes. It was a slightly less ridiculous (if still unlikely) thesis in the age of minicomputers. In the DOS days, it started to look likely, in the Windows days, it seemed inevitable, iOS and Android is what made it actually happen.
100%. Saas isn’t going away, but the economics are changing drastically and that’s bad for one-size-fits-all tools, and excellent for niche solutions. But it’s still saas, just more specific.
Businesses don’t want to use dozens of hyper-specific tools from dozens of vendors, if they instead can use a single vendor that can already do 80% of it and can vibe-code them the remaining 20%. I don’t see how this favors niche vendors.
I don’t disagree if models stay as capable as they are today. But devils advocate: the point of the saaspocalypse isn’t just that anyone will be able to make their own software, it’s also that the AI will be good enough and interconnected enough to maintain it.
The world these investors are envisioning is not one where a software engineer gives a detailed spec to a model and reviews its output, deploys the resulting files and monitors said application. It’s where Jo-shmo at the law firm can tell the model “give me a new billing system”, and the AI does everything correctly and better than a team of software engineers, in a matter of minutes or hours. And that AI maintains it for them, better than the engineers would have
Imagine that you are a consultant. You get a call that starts with, "Hi, this is Joe Schmoe and Schmoe Law Firm. I need a new billing system. Can you build me one?"
And you respond by saying that you can, but you need to do a _lot_ of work with him to spec this billing system out. You can't just build "a new billing system" without any more details. You tell him that this will take many hours of work between the two of you where you ask him questions, write a spec, get his feedback, and repeat that a number of times.
At this point, he says "wow, that sounds like a ton of of work for me just get started", and he gives up.
AI does not fix any of this, and this is the thing that I think most people will not want to do, and that's why I think this blog post is making a very good point. The amount of work it takes to build a new software system, even with a super competent programmer as a partner, is still quite significant. And it requires thinking about hundreds of tiny little details in a way that drives a lot of people nuts. They will only do it if they _really_ have to do it.
Think payroll. I used to think payroll is relatively simple. Then I spent some time on government of Canada Phoenix pay system (go ahead... Google and weep). And it's... Insane. System has been live for a decade and still regularly gets hit with some weird scenario from some department that nobody foresaw, it wasn't captured in requirements, but upon review by business analysts is a valid scenario. Bob was a CS5 in department of defense and speaks French so gets bilingual bonus and his boss was away for half a day so Bob gets acting cs5 pay and is in public alliance union so these are the dues, and it is second Tuesday of a month and blue moon, but then Bob got moved to department of agriculture and then 3 months later realized that his previous manager at defense didn't put in his promotion on time so now you have to figure out his retro pay for when he was in defense even though everything on his file now has agriculture labour agreement and codes and rates etc etc etc. And this made up example is a fraction of the complex examples.
Clear and comprehensive Requirements are always the tricky bit, at least in business software. Twilight zone covered it perfectly and presciently decades before AI, with genies taking your requests literally and giving you unpredictable and usually negative outcomes.
Or AI won't fix diffusion of responsibility that you see in companies through outsourcing, offshoring or matrix organizations...Or to go through committees to know if they should change shh root access with abc123 as password.
I was being a little facetious - I really dont think AI systems are there yet. It would probably look more like an interview, and there will be some amount of human-required maintenance and subjectivity for a while.
But I think thats what the investors are envisioning.
Everything you describe is fantasy, though. It’s not real. It’s not possible to be real. “Give me a new billing system”?? No way is that going to produce a good result for the company or their clients. But the second that Joe Schmo has to start laying out all the ever-evolving requirements for his custom billing system, he will run back to traditional SaaS providers.
At best, if AI is supergenius enough to just intuit everything Joe needs, then the cost of running the AI to constantly maintain a billing system will far exceed the cost of just paying someone for their existing billing system SaaS.
I think the idea is you'd basically have it take a look at your current system, it would learn what features you're actually using at all, it'd check company emails for past and current pain points or stuff you wish was possible or just simpler, it'd Slack everyone in the company asking what their biggest wish and biggest pet peeves are currently, it'd do a small interview with Joe himself presenting the above to see if it's gotten the right idea, create a very detailed spec and then implement it.
Of course both models and tooling will need to be far more powerful for all this, but it doesn't exactly seem sci-fi to me.
Once system is built it could run detailed analysis on its usage and figure out what parts seem to be confusing or slow for users, and simply refine, deploy, keep analyzing, rinse and repeat.
The biggest upside is probably that workers could also simply request features, have Joe sign off on them (would get messy otherwise) and minutes later they actually roll out.
To me anyways most systems are a PITA because they do so much and your own organization only utilizes a small subset. Good systems actually let you turn off stuff you don't use so that users don't even know it's possible and don't have to drown in menu options, but that's still rare enough. And good luck getting dev focus on your specific requests regarding the parts of the system most important to your specific company, since there are a zillion other things and hundreds or thousands of other customers.
Something literally tailored to what you need will surely be the norm eventually. In five years or whatever I'm sure we'll be plenty on our way towards something like that.
But again just like LLM training in general this all requires having something existing to analyze and work off of. So yeah nobody will be going from paper to custom agent-built system.
Other idea:
Stay with SaaS, real devs, real core product, closed source, but each customer can (if they want and pay up) literally skip multi-tenant and being on the same codebase as everyone else, and get an interface to actually customize their own version to their liking.
Remove unneeded features, change UX, UI, add features. Some dev spends tiny amounts of time ensuring nothing gets too crazy, but apart from that it's basically an autonomous fork of the product, continuously tracking main.
I vibe coded a saas and it went nowhere because it wasn't a good enough idea to begin with. I consulted with multiple varied models along the way for competitive analysis, pricing structure etc.
AI doesn't solve for ideas and product market fit. But it did allow me to fail pretty fast before I sunk too much time into it. But also, I should have spoken to potential users earlier rather than vibe coding.
Why wouldn’t just make some AI generated user personas to talk to? Whatever their opinion is, it’s already been captured and is in the training data. You don’t need to talk to users.
> they're almost always people who already had some pull toward software
I think this is probably true, and basically how I got into software myself.
I always dabbled in writing software and things for the web, but for some reason I never thought studying computer science would be any fun and that a career as a software developer sounded boring. But then I got an actual full time office job and oh boy, did my perspective on things change fast.
That first job did not have anything to do with writing software at all. But I saw people struggle with things that seemed to me trivial to automate, such as making annotations on paper bank statements and entering them into the system line-by-line. The bookkeeping system did support electronic bank statements, but lacked features to match certain descriptions to certain cost places. In the end it was indeed faster to go the paper route... It took me a couple of hours to write something that saved hours every week and that basically kick started my software career.
Would AI have made much of a difference here? Yes, in terms of getting to the correct solution faster, but probably not in terms of who would have done that. People would still come to the person who came up with the solution to ask for maintenance and new features.
At the individual level, I think most people will be writing software, whether they realise it or not. Asking Claude to do something for you will often result in a purely generated script built for that one specific task. Some might even take it further, generating custom dashboards or whatever else they need to support their work.
At the company level though... most companies can hardly maintain opensource deployments, let alone write and maintain their own bespoke software. Pick any company that uses GitLab, they're probably a few major versions behind. It's across the board.
There's no doubt people will try to write more software.
But we've all seen how this plays out.
The smart engineer who built a weekend solution leaves, and nobody supports the software afterward. Coding agents certainly help, and only time will tell, but my bet is that for most organisations it will end up miserably.
The doomsday saying the past months that everyone will now vibecode a solution instead of paying for SaaS as a big logic error… These people could have switched to self-hosting an open source clone of any popular SaaS but they didn‘t! Why? Because they dont want to be the person maintaining this, so we should those people not self-hosting a free software now go one step further and also build those products?
> Because they dont want to be the person maintaining this, so we should those people not self-hosting a free software now go one step further and also build those products?
If one thinks AI can do this eventually then it makes sense. But I feel that is impossible the predict.
Competing on price was never a good strategy. Moreover, price segmentation is still a thing. You can buy Chinese Rolex knockoffs for $7, but people still buy $10k Rolex.
Why are SaaS margins apparently dropping? The cost of producing saas is going down, no one is investing in new saas companies, the build your own saas will crumble as described in this article, meanwhile corporate IT departments will be decimated to pay for ballooning AI costs leaving saas as the only option to run companies.
Are any saas companies actually reducing their prices?
Do some people really believe that SaaS margins are dropping because the public at large has discovered that to save 20$/month on some app they can instead vibe code their own and use that instead?
My current guess about the future is that the age of SaaS is coming stronger than ever. I expect many vibe coders to come up with half-assed prototypes that will be copiously replicated and improved by more qualified devs aided by LLMs. In a similar way, I also expect smaller qualified teams (3 to 5) to leverage LLMs to become more relevant competitors of medium to large SaaS players. By 2029, we'll have more, but smaller SaaS companies.
Margins wouldn't drop because every consumer is going to vibe code their own apps. It's going to bring down the barrier for competition creating natural price pressure in the market. That is of course if all other factors end up equal such as quality, security, performance, etc...
This of course will be software in general imho. It's not that the profession will disappear overnight. There is going to be this tight squeeze until all the margin/excess salaries/etc.. is gone. There is also going to be immense pressure to produce as much as possible and productivity expectations are going to go way up (even if it is unjustified).
Basically, the good days are over. It's going to be a miserable profession.
SaaS companies have a big dilemma. Agents make your SaaS app more useful if there's no lock in and data is easy to access, but if that is the case, you have no lock in and the app is easy to migrate away from.
If you've already figured out what features are actually needed and which workflows work best, somebody can use AI to replicate those. This is the part of programming that AI is best at and accelerates most. The hard part, coming up with ideas, is far harder to copyright.
I could be wrong, but my understanding is that when people talk about the Death of SaaS, they are not talking about $20/month consumer apps. They are talking about six-figure+ enterprise deals that are the source of so much profit for SaaS companies.
I couldn’t disagree with this more. And at least anecdotally I’ve seen the opposite. I have one friend who has built and launched an app for diagnosing skin disease from a photo. Do I think it’s a good idea? No, but he built it. I have another friend that has completely automated her job in accounts receivable. She literally doesn’t have to work anymore and her employer has no clue. And another friend of mine is cranking out a new consumer app every few days. None of these people are even remotely technical. This is just the beginning.
Strongly agree. Personal software is also personal responsibility. It’s fun to dream up features, much less fun to be responsible for their implementation and maintenance.
However, if you move from "bespoke" to just "very small niches", I think lower production costs of software may well open up opportunities that were earlier unprofitable.
I buy it. SaaS doesn’t have to go extinct for this to be true.
I’m building an app and many things I’d normally pay for like metrics and emailing I can just do myself.
A friend has a law firm employing 100+ people and they are building so many internal tools they would otherwise be delaying or paying salesforce consultants for.
With all due respect to what Jason Fried &co have achieved, this is wrong.
Bespoke sofware does exist. And yes, consultants small and large have built, deployed, and charged through the roof for bespoke software. And often it sucks. Here's why it sucks: because clients can't coherently describe what they need, don't have a budget, consultancies don't care and - critically - the person writing the spec (and controling the budget) isn't the same person that will use it. (here you also have "A Tragedy of EdTech" in one sentence, but that's a different post)
But there's another kind of bespoke software, which, for a lack of a better name, I'll unimaginatively call "internal tool". This is what VB6/Access/VBA/HyperCard enabled back in the day, what Retool tried to own recently, and what many Excel spreadsheets are secretly doing.
This is duct-taped-code-pasta that barely holds but does exactly what the business needs, and nothing more. I've seen and heard of many cases already of non-techies doing exactly that. It's not scalable, it's not maintainable, it doesn't follow best practices, it doesn't have tests or docs, but it doesn't matter, because it works and solves a biz problem.
The reason it works is that the person can iteratively narrow down to what they need, feedback is instant, iteration is minutes not days or weeks and is super cheap (compared to external developers).
No sane freelancer or agency would ship something like it - for many reasons: as a software engineer you want to ship quality product and charge appropriate amount of money. Many times, that's the right thing for the customers.
Often, it's overkill, and these types of smaller "quick win" projects never get started in the first place. And there's loads of potential projects like these!
So yeah, nobody will vibe-code a payroll system for 100+ person company, nor should they. But people absolutely will, and already do, whip up something that solves their niche problem. Now maybe they'll use AI instead of Excel.
Excel is 'free-at-point-of-use', i.e. once you've paid for it, to use it doesn't cost anymore. But LLMs do cost per use (unless we all go to local models). Either this cost is billed directly, or some sort of bundling occurs with 'fair use' limits.
Excel is deterministic, yes scary spaghetti-fied spreadsheets are routinely constructed, but, for example, sorting a result column somewhere can be done with a bit of poking in the right place. LLMs have a tendency to dangerously change many things if the prompting is a bit wrong (and even if it is a bit right).
> Most people don’t like computers. Nobody in tech wants to say that out loud. People tolerate computers. They use them because they have to. Given the choice, most would rather not think about them at all.
Which makes me think there's a lot more room for "virtual people." Imagine a very smart AI bot that could hold multiple conversations at once and remember a lot of things.
> So when someone suggests that AI means everyone will build their own custom tools, ask who "everyone" is. The three-person accounting firm drowning in client paperwork? They want the paperwork gone, not a new system to maintain. The regional logistics company with 40 trucks? They want the routes optimized, not Joe spouting off about this new system he’s been messing around with. The law firm billing 70-hour weeks? They want leverage on their time, not a software project to design.
What if there was a bot that was just smart enough to figure those things out, without needing traditional "software"?
To me, that's more what AI is, instead of adding chatbots to everything, and vibecoding everything.
that's a good way to frame it, but it boils down to: what is it that these entities or individuals do that is valuable and how do you replicate parts or the whole of it.
which is essentially the direction that were heading in: we're sequentially and iteratively building improvements.
what the logistics company did pre computers and even pre trucks was not all that different in many aspects.
Neither am I. It feels like the dotcom in the sense that people will be spitting out new apps all overt the place. Down the line they will have problems maintaining them (and will say it's not core to their business) and they will revert to SaaS. However i expect the SaaS apps to have super low margins compared to today. Instead of 20-30% it will be 5-7% and the companies will be a shadow of themselves.
The main point everyone seems to be making is that now with AI anyone can make a SaaS.
The initial reaction I think most people have to this is "SaaS companies/devs are in trouble."
I actually think the opposite is true.
With an outpouring of vibe-coded apps/SaaS, you have the new wave of vulnerabilities/leaks/problems that happen even with the best software. Except now, it's worse because it's being done on platforms "built" by people who haven't the slightest clue how they work.
One of many examples: https://dig.watch/updates/women-only-dating-app-tea-suffers-...
This I imagine will, over time, erode trust in most apps/SaaS products. With that erosion of trust will come skepticism and with that, will come trust in the "old faithful" of SaaS products/companies. Basecamp is a good example of this.
I could be wrong on this one, but it seems to me those that have built credibility for privacy/security/competence will become more valuable in the AI age, not less.
This is basically why I buy the tech dip. When you pay for software, you pay for infrastructure, expertise, QA, consumer relations, having staff on call, etc. It was always possible to replace enterprise software by 2 guys coding a product in 6 months, but you still need everything around the code before serious clients will want to work with you, and at that point you're a regular software company. All these vibe coded products are one untested push away from getting dropped.
This is exactly right.
You know what's funny, less than a week ago I signed up for Basecamp.
Could I have asked Codex/Claude to whip me up a Basecamp clone with the exact features I want?
Of course. Do I want to deal with managing that codebase, even with AI? No.
The problem has been solved and the $15/mo. is well worth the time I will save not having to deal with managing that codebase and can instead focus my attention on things that bring in revenue.
+ the domain expertise collected from multiple clients
From a market perspective bundling this into SaaS players is more efficient.
But: AI might enable niche applications which were to expensive to capture thus far
SaaS companies/devs are in trouble - but for slightly different reason. That was the case already for something like 10 years.
Earlier if you had developers and no domain knowledge you were able to land a contract building application for a company and maybe spin it off to get more customers in that niche.
If you got lucky and you landed law firm and made case management for them you probably had nice little niche.
But as it turns out lawyers can also use JIRA, Trello, Basecamp or whatever and they really don't need Facebook for lawyers so those gigs dried up.
Main point is, software development alone is not going to bring as much money as it did earlier. You will have to have backing of domain experts to get the business going to offer something special in your SaaS. Like possibility to actually have call with those domain experts or their oversight on whatever it is you are doing but you not having budget or enough work to hire domain expert full time.
This is essentially what Jensen Huang (Nvidia CEO) was predicting a few months ago. Incumbents in most software spaces will probably see a lot of short and medium term benefits from the new tooling as being trustworthy and truly understanding the problem space.
> The main point everyone seems to be making is that now with AI anyone can make a SaaS.
I agree with you that is incorrect.
With AI, not everyone needs a SaaS.
They can make a bespoke tool for themselves with 5% of the SaaS features they actually need. If it's only used by authorized, internal, users and never exposed to the outside, many of the risks you mention disappear.
That's not to say everyone will vibe-code their Slack replacement, but a bar for relying on an external SaaS vendor will go up (and I think that's a good thing).
The default has been pay $x/month for every service. I've seen startups that require a dozen service accounts just to run the software, and dozens more to get onboarded org wide. One service for feature flags. One service for logs. One service for traces. One service for error handling. Another service for ticket tracking, which is completely separate from your planning, design, and CI services. Jesus. What do people hope to accomplish here besides just defering blame?
Replacing SAAS isn't about building a replacement services 1:1. It's about figuring out what you actually needed in the first place! Often we only use a tiny fraction of what the full-blown SAAS offers. IOW it's about eliminating the service entirely and building something that fits your actual needs, rather than following what some VC thinks your needs are.
AI or not, the "build vs buy" pendulum is now swinging hard to build. And IMO that's a real opportunity to consolidate, trim some fat, and actually apply engineering practices rather than just blindly signing up for every SAAS that crosses your path.
This fundamentally misunderstands a couple of things.
DIY software is "free" like a free yacht is free. It initially looks appealing but there's a lot of expensive hidden costs and upkeep and pitfalls and problems.
For one, this is a bad assumption:
> building something that fits your actual needs
Unless your business is very small and not growing, this is a moving target. Your needs are going to change as you grow and different groups in the org are going to have different needs.
You really don't want to be dicking around creating software that already exists instead of doing the shit that actually makes you money. Spending a few hundo thousand on a bunch of software is nothing, you spend that on one engineer.
You buy a SaaS product because you have a problem and want to throw money at someone else to deal with it.
You're most certainly wrong on this one. Superior models give superior products and security over time. Until every 3-6 months stops bringing a large improvement in coding capability and scaffolding, there's no reason to assume we are nearing a hard limit.
You also have to factor in that bespoke software is... bespoke. ie. much more suited to your org's use-cases than the primary solution is. Way less bloat. Way less vulnerability when you don't need an enterprise SaaS solution and instead can host on your private networks.
And as far as security considerations: Imagine you had a separate Opus 4.6 agent tasked with managing and monitoring and updating devoted to a specific slice of vulnerabilities. Of course this is highly inefficient, but it would take care of the vast majority of vulnerabilities that even enterprise SaaS have. This is simply a scaffolding issue at this point, not model ability. Scaffolding issues like this will continue to dominoe.
>Until every 3-6 months stops bringing a large improvement in coding capability and scaffolding, there's no reason to assume we are nearing a hard limit.
How much of that is better models, and how much is it AI companies throwing more resources at each one? E.g. larger context windows and higher token/s correlate with the better models.
This is like saying I know how to do plumbing so now I’m going to do all my own plumbing.
Yet I will still pay for a plumber. I wonder why.
[dead]
I also suspect AI is going to make software more secure rather than less.
Even today it can probably find a lot of issues automatically. With basic knowledge of what to look for, it certainly helps in understanding data flow too.
When the cost of software drops, supply increases.
That means we'll see even more niche apps, and more custom apps.
That doesn't mean everyone becomes a builder.
It means that the people who can build can now do so much more cheaply. Custom apps that were previously too expensive may now be cost-effective.
I think people fundamentally misunderstand what "non-techies writing software" means.
Nobody is going to ask Chat GPT to write them an app. They'll just ask something like "show me all the nearby restaurants that don't have any shellfish products for allergic reasons and let me browse like on ubereats", and Chat GPT will (eventually) do the right thing.
People will just see some UI, which will be like many UIs they've seen before. They won't care whether that UI is some UberEats embed designed by a programmer in California, or something that Chat GPT just came up with on the spot to wrap some API.
"Eventually, everybody is going to use computers" was a ridiculous thesis in the age of mainframes. It was a slightly less ridiculous (if still unlikely) thesis in the age of minicomputers. In the DOS days, it started to look likely, in the Windows days, it seemed inevitable, iOS and Android is what made it actually happen.
100%. Saas isn’t going away, but the economics are changing drastically and that’s bad for one-size-fits-all tools, and excellent for niche solutions. But it’s still saas, just more specific.
Businesses don’t want to use dozens of hyper-specific tools from dozens of vendors, if they instead can use a single vendor that can already do 80% of it and can vibe-code them the remaining 20%. I don’t see how this favors niche vendors.
Isn’t the SaaS landscape already highly specific?
I don’t disagree if models stay as capable as they are today. But devils advocate: the point of the saaspocalypse isn’t just that anyone will be able to make their own software, it’s also that the AI will be good enough and interconnected enough to maintain it.
The world these investors are envisioning is not one where a software engineer gives a detailed spec to a model and reviews its output, deploys the resulting files and monitors said application. It’s where Jo-shmo at the law firm can tell the model “give me a new billing system”, and the AI does everything correctly and better than a team of software engineers, in a matter of minutes or hours. And that AI maintains it for them, better than the engineers would have
Imagine that you are a consultant. You get a call that starts with, "Hi, this is Joe Schmoe and Schmoe Law Firm. I need a new billing system. Can you build me one?"
And you respond by saying that you can, but you need to do a _lot_ of work with him to spec this billing system out. You can't just build "a new billing system" without any more details. You tell him that this will take many hours of work between the two of you where you ask him questions, write a spec, get his feedback, and repeat that a number of times.
At this point, he says "wow, that sounds like a ton of of work for me just get started", and he gives up.
AI does not fix any of this, and this is the thing that I think most people will not want to do, and that's why I think this blog post is making a very good point. The amount of work it takes to build a new software system, even with a super competent programmer as a partner, is still quite significant. And it requires thinking about hundreds of tiny little details in a way that drives a lot of people nuts. They will only do it if they _really_ have to do it.
Think payroll. I used to think payroll is relatively simple. Then I spent some time on government of Canada Phoenix pay system (go ahead... Google and weep). And it's... Insane. System has been live for a decade and still regularly gets hit with some weird scenario from some department that nobody foresaw, it wasn't captured in requirements, but upon review by business analysts is a valid scenario. Bob was a CS5 in department of defense and speaks French so gets bilingual bonus and his boss was away for half a day so Bob gets acting cs5 pay and is in public alliance union so these are the dues, and it is second Tuesday of a month and blue moon, but then Bob got moved to department of agriculture and then 3 months later realized that his previous manager at defense didn't put in his promotion on time so now you have to figure out his retro pay for when he was in defense even though everything on his file now has agriculture labour agreement and codes and rates etc etc etc. And this made up example is a fraction of the complex examples.
Clear and comprehensive Requirements are always the tricky bit, at least in business software. Twilight zone covered it perfectly and presciently decades before AI, with genies taking your requests literally and giving you unpredictable and usually negative outcomes.
Or AI won't fix diffusion of responsibility that you see in companies through outsourcing, offshoring or matrix organizations...Or to go through committees to know if they should change shh root access with abc123 as password.
That billing system will end glue code that links existing Big SaaS players which AI can handle just fine.
Consultants are not spinning up a bespoke SaaS product given the risk.
I really struggle to imagine that working. What does “a new billing system” even mean? How is it better than the old system?
I was being a little facetious - I really dont think AI systems are there yet. It would probably look more like an interview, and there will be some amount of human-required maintenance and subjectivity for a while.
But I think thats what the investors are envisioning.
The AI will link existing APIs with glue code; slurp up data, map it to appropriate CRUD ops with Intuit or something.
We are way beyond the 90s-early 00s wild west where billing can be some random consultants opinions.
Everything you describe is fantasy, though. It’s not real. It’s not possible to be real. “Give me a new billing system”?? No way is that going to produce a good result for the company or their clients. But the second that Joe Schmo has to start laying out all the ever-evolving requirements for his custom billing system, he will run back to traditional SaaS providers.
At best, if AI is supergenius enough to just intuit everything Joe needs, then the cost of running the AI to constantly maintain a billing system will far exceed the cost of just paying someone for their existing billing system SaaS.
I'm not saying its logical - I think this is what investors believe however.
I think the idea is you'd basically have it take a look at your current system, it would learn what features you're actually using at all, it'd check company emails for past and current pain points or stuff you wish was possible or just simpler, it'd Slack everyone in the company asking what their biggest wish and biggest pet peeves are currently, it'd do a small interview with Joe himself presenting the above to see if it's gotten the right idea, create a very detailed spec and then implement it.
Of course both models and tooling will need to be far more powerful for all this, but it doesn't exactly seem sci-fi to me.
Once system is built it could run detailed analysis on its usage and figure out what parts seem to be confusing or slow for users, and simply refine, deploy, keep analyzing, rinse and repeat.
The biggest upside is probably that workers could also simply request features, have Joe sign off on them (would get messy otherwise) and minutes later they actually roll out.
To me anyways most systems are a PITA because they do so much and your own organization only utilizes a small subset. Good systems actually let you turn off stuff you don't use so that users don't even know it's possible and don't have to drown in menu options, but that's still rare enough. And good luck getting dev focus on your specific requests regarding the parts of the system most important to your specific company, since there are a zillion other things and hundreds or thousands of other customers.
Something literally tailored to what you need will surely be the norm eventually. In five years or whatever I'm sure we'll be plenty on our way towards something like that.
But again just like LLM training in general this all requires having something existing to analyze and work off of. So yeah nobody will be going from paper to custom agent-built system.
Other idea: Stay with SaaS, real devs, real core product, closed source, but each customer can (if they want and pay up) literally skip multi-tenant and being on the same codebase as everyone else, and get an interface to actually customize their own version to their liking. Remove unneeded features, change UX, UI, add features. Some dev spends tiny amounts of time ensuring nothing gets too crazy, but apart from that it's basically an autonomous fork of the product, continuously tracking main.
That will probably come a lot sooner.
I vibe coded a saas and it went nowhere because it wasn't a good enough idea to begin with. I consulted with multiple varied models along the way for competitive analysis, pricing structure etc.
AI doesn't solve for ideas and product market fit. But it did allow me to fail pretty fast before I sunk too much time into it. But also, I should have spoken to potential users earlier rather than vibe coding.
Why wouldn’t just make some AI generated user personas to talk to? Whatever their opinion is, it’s already been captured and is in the training data. You don’t need to talk to users.
> they're almost always people who already had some pull toward software
I think this is probably true, and basically how I got into software myself.
I always dabbled in writing software and things for the web, but for some reason I never thought studying computer science would be any fun and that a career as a software developer sounded boring. But then I got an actual full time office job and oh boy, did my perspective on things change fast.
That first job did not have anything to do with writing software at all. But I saw people struggle with things that seemed to me trivial to automate, such as making annotations on paper bank statements and entering them into the system line-by-line. The bookkeeping system did support electronic bank statements, but lacked features to match certain descriptions to certain cost places. In the end it was indeed faster to go the paper route... It took me a couple of hours to write something that saved hours every week and that basically kick started my software career.
Would AI have made much of a difference here? Yes, in terms of getting to the correct solution faster, but probably not in terms of who would have done that. People would still come to the person who came up with the solution to ask for maintenance and new features.
Yes and no.
At the individual level, I think most people will be writing software, whether they realise it or not. Asking Claude to do something for you will often result in a purely generated script built for that one specific task. Some might even take it further, generating custom dashboards or whatever else they need to support their work.
At the company level though... most companies can hardly maintain opensource deployments, let alone write and maintain their own bespoke software. Pick any company that uses GitLab, they're probably a few major versions behind. It's across the board.
There's no doubt people will try to write more software.
But we've all seen how this plays out.
The smart engineer who built a weekend solution leaves, and nobody supports the software afterward. Coding agents certainly help, and only time will tell, but my bet is that for most organisations it will end up miserably.
The doomsday saying the past months that everyone will now vibecode a solution instead of paying for SaaS as a big logic error… These people could have switched to self-hosting an open source clone of any popular SaaS but they didn‘t! Why? Because they dont want to be the person maintaining this, so we should those people not self-hosting a free software now go one step further and also build those products?
> Because they dont want to be the person maintaining this, so we should those people not self-hosting a free software now go one step further and also build those products?
If one thinks AI can do this eventually then it makes sense. But I feel that is impossible the predict.
Who do you think they can blame in case of outage/security issues etc ?
It will be race to bottom for SaaS in terms of pricing, with lots of alternatives to every SaaS.
It’s not about personal software it’s about how 1-3 people team will deliver a SaaS that actually works at scale for the 1/10th of the price.
In terms of personal software, he’s absolutely right, it’s great for hobbyist and things like in house tooling but that’s it.
Competing on price was never a good strategy. Moreover, price segmentation is still a thing. You can buy Chinese Rolex knockoffs for $7, but people still buy $10k Rolex.
That's for signaling purposes, not for functional purposes. Who is buying SaaS for signaling purposes over function?
People who need support, and compliance? Things that AI generated slop can not provide
I think the AI would be exceptionally good at support, especially for a codebase that it owns, but you are right about compliance.
A lot of SaaS services have open source solutions that are the same, on par or better and yet people still pay for SaaS.
If they're not doing it now, why would they do it when they have to build it, too?
> It’s not about personal software it’s about how 1-3 people team will deliver a SaaS that actually works at scale for the 1/10th of the price.
Is this already happening?
Shouldn't it be?
> In terms of personal software, he’s absolutely right, it’s great for hobbyist and things like in house tooling but that’s it.
That much is true.
Why are SaaS margins apparently dropping? The cost of producing saas is going down, no one is investing in new saas companies, the build your own saas will crumble as described in this article, meanwhile corporate IT departments will be decimated to pay for ballooning AI costs leaving saas as the only option to run companies.
Are any saas companies actually reducing their prices?
Because interest rates gone up, tech bubble hype now almost exclusively focuses on AI, and SaaS was commodified a decade ago?
Do some people really believe that SaaS margins are dropping because the public at large has discovered that to save 20$/month on some app they can instead vibe code their own and use that instead?
My current guess about the future is that the age of SaaS is coming stronger than ever. I expect many vibe coders to come up with half-assed prototypes that will be copiously replicated and improved by more qualified devs aided by LLMs. In a similar way, I also expect smaller qualified teams (3 to 5) to leverage LLMs to become more relevant competitors of medium to large SaaS players. By 2029, we'll have more, but smaller SaaS companies.
Margins wouldn't drop because every consumer is going to vibe code their own apps. It's going to bring down the barrier for competition creating natural price pressure in the market. That is of course if all other factors end up equal such as quality, security, performance, etc...
This of course will be software in general imho. It's not that the profession will disappear overnight. There is going to be this tight squeeze until all the margin/excess salaries/etc.. is gone. There is also going to be immense pressure to produce as much as possible and productivity expectations are going to go way up (even if it is unjustified).
Basically, the good days are over. It's going to be a miserable profession.
SaaS companies have a big dilemma. Agents make your SaaS app more useful if there's no lock in and data is easy to access, but if that is the case, you have no lock in and the app is easy to migrate away from.
If you've already figured out what features are actually needed and which workflows work best, somebody can use AI to replicate those. This is the part of programming that AI is best at and accelerates most. The hard part, coming up with ideas, is far harder to copyright.
I could be wrong, but my understanding is that when people talk about the Death of SaaS, they are not talking about $20/month consumer apps. They are talking about six-figure+ enterprise deals that are the source of so much profit for SaaS companies.
I couldn’t disagree with this more. And at least anecdotally I’ve seen the opposite. I have one friend who has built and launched an app for diagnosing skin disease from a photo. Do I think it’s a good idea? No, but he built it. I have another friend that has completely automated her job in accounts receivable. She literally doesn’t have to work anymore and her employer has no clue. And another friend of mine is cranking out a new consumer app every few days. None of these people are even remotely technical. This is just the beginning.
Those aren't SaaS apps, that's (respectively) unlicensed practice, fraud, and a new hobby.
Based on my math, taking development in-house doesn’t make sense until we’re talking >$5K/mn.
https://forstarters.substack.com/p/for-starters-72-vibe-code...
Strongly agree. Personal software is also personal responsibility. It’s fun to dream up features, much less fun to be responsible for their implementation and maintenance.
People hating software is such a good clue.
Like that logistics company owner, he doesn't want to fill in your crappy insurance form, he wants to call insurance broker and have an insurance.
Doesn't matter if you put AI in the form, make 10% discount, polish UX to finest levels. He just doesn't want to spend time on it.
However, if you move from "bespoke" to just "very small niches", I think lower production costs of software may well open up opportunities that were earlier unprofitable.
I buy it. SaaS doesn’t have to go extinct for this to be true.
I’m building an app and many things I’d normally pay for like metrics and emailing I can just do myself.
A friend has a law firm employing 100+ people and they are building so many internal tools they would otherwise be delaying or paying salesforce consultants for.
Incentive check! The author seems to have a vested interest in people not making their own software. Curious
With all due respect to what Jason Fried &co have achieved, this is wrong.
Bespoke sofware does exist. And yes, consultants small and large have built, deployed, and charged through the roof for bespoke software. And often it sucks. Here's why it sucks: because clients can't coherently describe what they need, don't have a budget, consultancies don't care and - critically - the person writing the spec (and controling the budget) isn't the same person that will use it. (here you also have "A Tragedy of EdTech" in one sentence, but that's a different post)
But there's another kind of bespoke software, which, for a lack of a better name, I'll unimaginatively call "internal tool". This is what VB6/Access/VBA/HyperCard enabled back in the day, what Retool tried to own recently, and what many Excel spreadsheets are secretly doing.
This is duct-taped-code-pasta that barely holds but does exactly what the business needs, and nothing more. I've seen and heard of many cases already of non-techies doing exactly that. It's not scalable, it's not maintainable, it doesn't follow best practices, it doesn't have tests or docs, but it doesn't matter, because it works and solves a biz problem.
The reason it works is that the person can iteratively narrow down to what they need, feedback is instant, iteration is minutes not days or weeks and is super cheap (compared to external developers).
No sane freelancer or agency would ship something like it - for many reasons: as a software engineer you want to ship quality product and charge appropriate amount of money. Many times, that's the right thing for the customers.
Often, it's overkill, and these types of smaller "quick win" projects never get started in the first place. And there's loads of potential projects like these!
So yeah, nobody will vibe-code a payroll system for 100+ person company, nor should they. But people absolutely will, and already do, whip up something that solves their niche problem. Now maybe they'll use AI instead of Excel.
True. The reason why this works is because someone owns it and takes responsibility of it, usually for reasons other than making money.
However:
Excel is 'free-at-point-of-use', i.e. once you've paid for it, to use it doesn't cost anymore. But LLMs do cost per use (unless we all go to local models). Either this cost is billed directly, or some sort of bundling occurs with 'fair use' limits.
Excel is deterministic, yes scary spaghetti-fied spreadsheets are routinely constructed, but, for example, sorting a result column somewhere can be done with a bit of poking in the right place. LLMs have a tendency to dangerously change many things if the prompting is a bit wrong (and even if it is a bit right).
> Most people don’t like computers. Nobody in tech wants to say that out loud. People tolerate computers. They use them because they have to. Given the choice, most would rather not think about them at all.
Which makes me think there's a lot more room for "virtual people." Imagine a very smart AI bot that could hold multiple conversations at once and remember a lot of things.
> So when someone suggests that AI means everyone will build their own custom tools, ask who "everyone" is. The three-person accounting firm drowning in client paperwork? They want the paperwork gone, not a new system to maintain. The regional logistics company with 40 trucks? They want the routes optimized, not Joe spouting off about this new system he’s been messing around with. The law firm billing 70-hour weeks? They want leverage on their time, not a software project to design.
What if there was a bot that was just smart enough to figure those things out, without needing traditional "software"?
To me, that's more what AI is, instead of adding chatbots to everything, and vibecoding everything.
that's a good way to frame it, but it boils down to: what is it that these entities or individuals do that is valuable and how do you replicate parts or the whole of it.
which is essentially the direction that were heading in: we're sequentially and iteratively building improvements.
what the logistics company did pre computers and even pre trucks was not all that different in many aspects.
the future will be via evolution not revolution.
Neither am I. It feels like the dotcom in the sense that people will be spitting out new apps all overt the place. Down the line they will have problems maintaining them (and will say it's not core to their business) and they will revert to SaaS. However i expect the SaaS apps to have super low margins compared to today. Instead of 20-30% it will be 5-7% and the companies will be a shadow of themselves.
Right so teams will shrink and we'll have more "bespoke" SaaS with more small biz services making enough to support maybe 5-10?
Sounds just like the bespoke software product companies and consultants out there.
Thats my guess and I think this might happen in other domains like games as well.
I am +5 on browser extensions just for me.
SaaS dev says SaaS is the one true solution to software.
Self selecting biology gonna self select.
I think there's a potential path where most of your interactions with the computer are through software that rewrites itself to your exact task.
> Most people don’t like computers.
Ummmm, what? People love their smartphones, and do you know what those are?
Computers.
[dead]
[dead]