>Workspace AI includes things like email summaries in Gmail, generated designs for spreadsheets and videos, an automated note-taker for meetings, the powerful NotebookLM research assistant, and writing tools across apps.
Maybe I'm just an old curmudgeon stuck in my ways, but I haven't found much compelling value in these use cases in my day-to-day work. For summaries and note-taking specifically, I feel they're solving the wrong problem: it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
The real solution to too much email is fewer and higher-priority emails. The real solution to too many meetings is fewer and more-focused meetings. These tools paper over the root cause of the problem, which is that people/organizations cannot (or are unwilling to) be clear about communication priorities and say "maybe this email/meeting isn't a good use of time after all."
There's a cartoon going around where in the first frame, one character points to their screen and says to another: "AI turns this single bullet point list into a long email I can pretend I wrote".
And in the other frame, there are two different characters, one of them presumably the receiver of the email sent in the first frame, who says to their colleague: "AI makes a single bullet point out of this long email I can pretend I read".
Right now at Amazon we are going through the annual feedback cycle where you have to write strengths and growth areas for your colleagues. You will usually have to do ~12 of those.
I don't use ChatGPT for those, but it is the epitome of what you are describing, people will take a single sentence, ask some LLM to blow it up the correct length and in the process make it a complete waste of time for everyone.
My guess is that with long-form text losing value due to LLMs, we will see a return of very succint 1-2 lines employee feedback.
I just exited the toilet following 2.5 hours of back-to-back meetings, and was looking forward to actually getting some work done when the product owner grabbed me for a conversation about priorities for the sprint planning session that's scheduled in a couple of hours.
In this week so far (first week back from Christmas / New Year leave) I've spent maybe half a day total on work that could be classified as "progress". The rest of the time has been meetings and the required meeting follow-up work.
There's no point in Sprint Planning or considering adding priorities to the current plate. It's full. But nobody has time to eat things off the plate because we're always in meetings to work out how we can eat off the plate more efficiently.
/rant
I've come back from holidays angry. Things gotta change.
I had a few use cases with searching and organizing emails I would have used. For example, I wanted a table of all my Lyft rides from a certain year with distances driven, start/end locations, cost, etc. All that info is available in the email you get after riding, so I figured Gemini could read my emails and organize the info.
Turns out it doesn't work at all. It gave me a random selection of rides, was missing info in some of them, and worst didn't realize it was giving me bad info. Pretty disappointing.
I find AI meeting transcripts and summaries to be one of the most genuinely useful things to come out of this era of LLM tools. Being able to see a quick summary of what was decided or who was supposed to do what next is just so helpful, either for refreshing your memory after the weekend or just because people aren’t all that great at taking and sharing notes.
These LLMs are excel at making more. More emails with more words. More blog posts with more fluff. Making it open to more people means more usage means more numbers being more which means more money for the people building these systems.
I don't see what I get out of 80% of these products. It's just more noise.
> For summaries and note-taking specifically, I feel they're solving the wrong problem: it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
I think this really encapsulates something that I hadn't been able to put my finger on in regards to LLM summarization. What it seems to indicate is that, if you need a computer to summarize a large amount of text that someone has sent to you, there are two likely possibilities:
1) The information is incredibly dense/important/technical/complex. This necessitates the extreme length of the message - (think: technical documents, research papers, a rough draft of a legal notice, or your will.) For these sorts of things, you should not rely on an LLM to summarize it, because it may miss key details of the message.
2) The person sending it to you is bad at communicating, in which case the solution is help them learn better communication, rather than "de-noising" their clumsy wording into something comprehensible.
"But what if its number 2, but it's coming from your boss?"
Then I see two obvious points to consider:
First, you should absolutely be telling them about the problem, regardless of the position that they hold. You can phrase it in a way that isn't rude. "Hey boss, I saw (message) but I'm not 100% the intent. I've actually noticed that with (other time)...I usually try to front-load the action items up front, and put the specifics lower down. Anyway, to make sure I'm tracking, you're talking about (action) on (thing), right?"
Second, until (or unless) their communication style is de-noised, then part of your job is being able to "translate" their instructions. Using an AI to do that for you is a bad idea because, at some point or another they're going to be trying to speak to you in-person, or by phone.
Not having dealt with their mannerisms in an unfiltered way might lead to you being "out of practice" and struggling much harder to figure out what they're trying to convey.
Well, that’s because you’re thinking as someone who likely has a stake in quality/specific outcomes actually happening. Or was raised/grew up in an environment where that was important.
Notably, in my experience there is a high correlation with that background and being curmudgeonly. Mainly because that means someone has been responsible for outcomes, regardless of feelings. And something often has to give, and it’s usually feelings. It’s also hard to not be cranky or even angry if someone has to constantly be the one ‘not having fun’ or cleaning up messes so the whole thing doesn’t fall apart.
There is huge market demand exactly for what you’re complaining about, which is faking things happening as convincingly as possible, precisely because being clear/concise, etc. helps with seeing the root cause of problems, and if someone is worried (or is legitimately) a root cause of the problem, of course they’ll consider that bad.
For example, a good sign of a badly led organization is that it’s always busy, but never seems to get anything done. Everything is an emergency, so nothing really gets fixed, etc.
Or there are constant meetings and emails, but nothing gets decided.
People will pay good money for the right kind of wallpaper that makes that ugly wall look pretty again.
> I haven't found much compelling value in these use cases in my day-to-day work.
If my experience with Microsoft Office Copilot is any indication, these features produce very confusing, low-quality content if they are not completely wrong and useless. Used it once and never touched them again. (My company is still paying for this and rolling this out widely despite many reports of how unhelpful they are.) I doubt Google Workspace can do any better.
I don’t use it often either, but sometimes it is very useful. When I caught Covid last fall my wife incorrectly thought I had it three times. I was using a beta Google Gemini, and paying for it, and I asked “read my @gmail and tell me the date ranges when I have had Covid.”
That worked, but to be honest I have tried similar things more recently that didn’t work. Perhaps there is a routing model up front that decides whether or not to use a lot of compute for any given query?
Google also plans on charging more money for APIs for code completion plugins for IntelliJ IDs, etc. this year.
I would like to see AI pricing models be sustainable, not give things away for free, and have lots of control over when I use a lot of compute. I actually have this right now because I usually use LLM APIs and write my own agents for specific tasks.
I'm kind of a cynic, so I'd say that the Workspace customer isn't you, the person who's using Workspace. It's your big company's SVP of IT or whoever who wants to spend money to adopt cool AI stuff so that he can say that he did AI stuff.
> it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
I tend to agree, except these two things are kind of the same thing. It can make going through the noise easier by intelligently filtering out the noise or finding you the signal. Search. It doesn't necessarily need to eliminate the noise.
Maybe AI would be better if it prevented the noise, and its definitely going to add noise (expanding a few basic thoughts into an email with lots of fluff), but it can also solve it.
I’m getting a lot of value out of NotebookLM drafting documents. If I’ve got a bunch of notes that need to be in a coherent design doc, it can give me a good enough first draft for me to edit into shape. Alternatively when I’ve got a design doc for something, but need to submit, say, a work request to another org, NotebookLM can take my doc and turn it into another format based on a doc template pretty nicely.
These outputs still require editing for sure, but each one can easily save me half the time to write these things.
I agree. I don't want all my existing work apps to take on LLM features I don't need.
At the same time, I tried the Gemini Research feature last night, via the Gemini webapp, and was resoundingly impressed. From a vague description, it can find the open source project I was looking for, provided ample links, and a pretty good summary of the project.
I totally agree. I upgraded to the AI-enabled version of Google One because they gave a couple week free trial. I found it totally useless, and it reeked of "Some PM said we had to stuff AI in everywhere".
Note I do use ChatGPT pretty frequently, but I've found it much more useful to have a separate space for the kinds of conversations I have with ChatGPT.
I’m not a native english speaker, but working at US subsidiary I must produces reports in english etc - and having an LLM proofread my texts for me is great.
LLM:s are new modality to computing. If you need it, they are great. But just like excel/sheet have limited applications a LLM with data has limiited use as well.
I also find that summarising content helps me digest it better. I have to fully understand the source in order to write the summary. The process of writing a summary is of immense value. Sometimes the summary itself is of minimal value.
What if there was something that communicated the company’s top priorities and helped everyone align and stay organized without so many meetings, and give concise drafts for your to-dos? Would that be something you’d try?
Yeah I’m tired of workspace getting more expensive and me getting zero additional value from it. I don’t want this, didn’t ask for it, and it actively annoys me.
2. Email gets summarized at the receiver end in the daily AI email "summary" which might be something like
You have a scheduled cake tasting this weekend. Did you know there's a bakery near your office that makes wedding cakes too. By the way your friend Joe can't make it to the wedding, do you want me to send a reply?
3. Reply email gets summarized by AI.
"Your friend acknowledges that you cannot rsvp. Do you want to schedule a wedding gift delivery on their wedding day ? XYZ neighborhood/online store has a sale next week".
If you really care about this issue, I think we've brought it on ourselves.
Regarding teaching kids, we've set messaging templates for occasions that are at the center of our lives. We have Hallmark greeting cards to express feelings to people close to our hearts. If there's a template for expressing someone you're sorry their mother died, or happy they have a baby, I'm not sure throwing the stone at AI use is warranted.
In a way, I wonder if it will be the wake up call that will make simple and genuine communication acceptable again, without all the boilerplate we've built to feign care and emotions.
Gemini's marketing is so bad. This isn't the first time they ran an ad that makes you wonder what's going on there. It really says a lot that an advertising company understands what makes for good advertising so poorly these days.
> A friend of mine is a teacher and kids are already delegating their learning to ChatGPT and their learning isn't sticking.
I’m not going to defend AI here because I seldom use it myself. But it should be noted that the way we learn has already undergone multiple different shifts due to changes in technology.
Search engine were a big one. No longer did we have to learn to memorise stuff nor learn how to research properly. Now we could just type a phrase into Google / whatever and get results. So people learned how to search rather than learning the facts itself.
Second law of thermodynamics says these models will all eventually collapse (due to overtraining on their own output) to yelling gibberish at us, and biology will continue to remain the only force in the universe capable of maintaining order despite increasing entropy. I think we'll be OK.
I pay for 3 Workspace orgs, and I have Gemini disabled (or still not enabled) on all 3 of them. I'm angry that I'll have to pay more for features I don't use. Gemini should be an add-on cost, not included in the base cost
If they're raising the price of personal GMail, I don't have a problem. But Workspace with hundreds of users, now that's a problem, because it actually hurts my wallet significantly. When this increase comes, I'll have to move elsewhere.
It’s an anti competitive strategy, which in an ideal world would see them facing a crushing antitrust lawsuit from the FTC and DOJ. What they’re doing is forcing everyone to pay for their AI product. This makes it so that no other company can charge for their alternative AI products. After all, if your company’s spending goes up because of this Google price increase, your executives will not want to see double spending on AI products. So all those deserving smaller companies will miss out on these customers. Google is essentially using this forced price increase to kill their AI competitors by stealing their revenue, through illegal bundling. Just like Microsoft did with Teams to attack Slack illegally.
With Amazon as an example for CxOs of the world, sadly, this likely won't happen.
Look at Prime. So much crap involved, and quite literally all I use it for is lower cost shipping. It's almost on the edge of not worth it for me. But I bet from Amazon's perspective, they make more with the higher price, even if they lose the bottom 10% not willing to spend.
Huh.
Just made me realise, a startup that subscribes to Prime as a virtual being, and then splits off each sub-thing for full use by separate individuals would be incredibly profitable.
If any form of AI is eventually granted legal personhood, Prime's model will collapse.
Right now looking for an alternative for the same reason. Even if it cost me more on labor short term. They have been increasing prices regularly and I’m sure it will continue.
We are doing a Gemini POC and this nugget dropped in my lap today. We were not entirely unprepared as a result. The default level of access is just the interactive chatbot thing. However if you enable the Google Workspace extension it will be able to search and process all the information stored in your workspace account and also any Google Drive files that are shared with you. This includes stuff you didn't know you had access to in Shared Drives so folks better make sure their permissions are locked down. Workspace admins might be advised to turn it off at the org level until they understand the ramifications.
Reminds me of an entertaining story about Microsoft Copilot last year, where companies were turning it off because it turned out it was TOO good at its job - if any accountant anywhere in the company had messed up their SharePoint permissions asking "what does everyone at this company earn?" would spit out all of the salaries: https://simonwillison.net/2024/Aug/23/microsoft-copilot-data...
I am surprised the Workspace extension isn’t controlled by the same setting that limits general workspace search results, where you can set things up so only documents you’ve seen or are linked to from documents you have explicit access to are returned in results: https://support.google.com/a/answer/12732365?hl=en
Shid. I made the mistake of getting my entire family onto my google apps 15+ years ago. Now I am paying for about 8 people every month and this will just make it worse.
It was selling well enough. It just was not getting enough traction. By bundling AI, they are giving exposure to everyone who didn’t want to use it or didn’t see the need for it. If they pulled it away in 2 years, and then lowered the price and charged separately for AI, I think more people would see it as necessary.
AI is a better search engine. And a better grammar check for your emails. And a better writer for your reporting.
I expect take up was in the low single-digit percentage points. So charging every single subscriber $2/user (even if they don't want it) probably yields significantly more revenue.
Cool, great, fun. I have all of the “generative AI” features disabled in Workspace, and now I get to pay more for the privilege of keeping them disabled. Thanks, Google!
Do you know how to actually disable these new features (i.e. the elements that were added within Gmail, Docs, etc.)? I'm not seeing where they can be disabled and Google Workspace support was not able to point me in the right direction either...
I recently got Gemini Advanced as an additional benefit by virtue of having Google One paid storage. I'm shocked this is being given away for free, because it is now a seriously major part of my work. I literally have an Open window all day long interacting with it. It does make me wonder how much they are losing (investing) on giving all this inference away for free. Also makes me wonder what they are getting back aside from loyalty/data/?
Give it a quarter and we’ll see breathless articles about how Google saw “AI adoption increase 150%” and “Google workspace users say they can’t go without AI” (because they physically can’t remove it from their workspace).
This in turn, will be used as post-hoc justification of the value of AI and why ever more power, water and data should be funnelled into it.
I tested Gemini today, asking it to extract key pieces of data from a large report (72 slide) PDF deck which includes various visualisations, and present it as structured data. It failed miserably. Two of the key stats that are the backbone of the report, it simply made up. When I queried it, it gave an explanation, which further compounded its error. When I queried that, extracted the specific slide, and provided it, it repeated the same error.
I asked Claude to do the same thing, it got every data point, and created a little react dashboard and a relatively detailed text summary.
Maybe the prompt you used was more Claude-friendly than Gemini-friendly?
I'm only half-joking. Different models process their prompts differently, sometimes markedly so; vendors document this, but hardly anyone pays any attention to it - everyone seems to be writing prompts for an idealized model (or for whichever one they use the most), and then rate different LLMs on how well they respond.
Example: Anthropic documents both the huge impact of giving the LLM a role in its system prompt, and of structuring your prompt with XML tags. The latter is, AFAIK, Anthropic-specific. Using it improves response quality (I've tested this myself), and yet as far I've seen, no BYOK tool offering multiple vendor support respects or leverages that.
Maybe Gemini has some magic prompt features, too? I don't know, I'm in the EU, and Google hates us.
That matches with my experience, Claude is clearly ahead of its competitors in anything logic- or reasoning-based.
I find Gemini is better at queries that involve more kind of intuitive judgment over things where there isn't a clear "correct" answer. E.g. if I want a podcast recommendation, or advice on the best place to learn about a given problem, I find Gemini better than Claude.
Unfortunately for Gemini, 90% of the things I want an LLM for are better with stronger logic and reasoning.
I got a 1-year trial of Gemini Advanced with my Pixel 9 and I've had similar experiences. It makes up stuff far more often than any other models and it's just not very smart. I used the free version and thought the paid Advanced version would be better but I could hardly notice any difference, they both fail at the same prompts I've tried.
This is not to mention the poor app experience where some of the features are just missing or broken. For example it's able to "remember" stuff I ask it to remember, but when I ask it to forget something it says I have to manage it at this webpage (they didn't bother to implement this menu within the mobile app) that asks me to sign in again because it's opened in my web browser where I'm not signed into Google, and then it shows me an empty list and "Something went wrong". It's now calling me a name I told it as a joke and there's no way to make it forget
Ugh. The "vanity domains for gmail" product i've been buying for a long time is really metastasizing into something that's both too expensive and actually worse than the free experience, wonderful.
It’s weird that prediction 8, "Someday [you] will voluntarily pay Google for one of their services" has come around full circle to "and then you won't anymore, because they've dropped the ball to an extent usually associated with the private equity buyout -> loot into bankruptcy process"
This spurred me to go back and read the predictions:
>But I can tell you this: Google has changed my life. If I can't find what I'm looking for in Google in 3 tries, looking no further than the first 10 search results on each try, then it probably doesn't exist.
Plus now I'm noticing it doesn't work for more and more things. Youtube TV family sharing doesn't work, Android Auto had some problems, the news feed on my Pixel.
Wish Google would just fix the Drive search rather than lard it up with AI nonsense. Often it’s easier to ask someone to resend the link to a document than find it by searching.
There are so many bugs and sub-par implementations in workspace that Google could fix. My cynical guess is that the source code to workspace apps is probably a mystery to the current generation of 23 year olds who are tasked in maintaining them, so they change little.
I wish they fixed search in general. It is difficult to find emails if you don't know exact keywords that might have been used etc. often even if you type in the right keyword it still won't find the email, even though email contains it.
Worst thing is people sharing files tbh, if someone has a folder and shares you a multiple documents from it you don't get the folder in your drive structure so you have n free files floating around in your drive that you cannot organize yourself.
I'm in the middle of a free trial for the Workspace Gemini add-on.
It really, really sucks. I've played around with having it make tables for Sheets and it frequently gets confused or responds with ~"I can't do that, I'm just a LLM", even when feeding it one of their suggested examples word for word. Sometimes it's willing to iterate, sometimes it refuses. Once it gets confused,the only way I've been able to get it working again is by clearing the session and starting fresh.
I guess this is why it is being bundled, Google can keep working on it with someone else's money, so their profits aren't hit. It's telling that the increase is regardless you use Gemini or not.
There's a weird attitude on this site towards AI: if it's for coding or science, people generally recognize AI tooling as effective, although imperfect, and rapidly improving.
But any suggestion of using AI for business and it seems like disdain and dismissal is the majority response.
Don't you think many areas of business - maybe not all areas, but a significant amount - will just as much benefit from AI tooling as software developers and scientists?
They kinda already do it with YT Premium/YT Music. I don't have anything against YT Music, it's a perfectly fine music service from the amount I've used it. But I already have a Spotify with my preferred playlists, and I don't really have incentive to swap it over aside from maybe saving a handful of dollars a month.
Yet if you want ad-free YouTube the proper way, you can't just have Music as an option, it's rolled into your cost regardless of if you actually plan to use it or not.
Isn't that always the case with bundles and suites? Google Workspace has always been a bundle of products, and few actually used every product in the bundle.
The beauty of having a monopoly or oligopoly in a dozen major markets is that you don't have to care about customers. As much as I hate this move and don't think it will help the company, I think Google is powerful and entrenched enough that it will make out just fine. Their users will bear most of the costs.
Does this apply to the legacy free edition? I suspect not, since that edition is now only available for personal use and they mostly focus on Business and Enterprise use cases, but their public guidance isn't very clear. If it does apply, would we legacy free edition users be receiving Gemini under the Google Workspace Terms of Service preventing them from using our data for general AI training, or under the regular Google Terms of Service which might allow this?
(Tangent: I say "might allow this" because I don't know to what degree EU law requires some additional level of consent beyond accepting the Terms of Service for EU-based accounts like mine currently is, or requires them to give me an AI-specific opt-out despite having a free account. But this announcement doesn't change whatever EU law does or doesn't require, so that is unrelated to my main questions about which Gemini features will apply to the legacy free edition under which Terms of Service once this change rolls out.)
The legacy free edition includes the features of the 'Business Starter' plan, the most affordable option. In this table [1], you can find the features available for each plan.
Here are the details for the Business Starter plan specifically:
Gmail: Help me write, Side panel, Contextual smart replies (Coming soon).
Gemini app: Enterprise-grade security & privacy, Google Workspace extensions.
NotebookLM: Upload sources, create summaries and Audio Overviews, and Q&A.
There's no legacy free edition for personal use any more. That ship sailed in 2022. I do not believe there is a way to have it free after 2022. Free plans were converted to Business Starter.
Exactly. It's easy to make something free when nobody is remotely interested in paying for it. If Google wanted to make money they'd have to let people pay to remove AI from gmail and docs.
Investors are the customer, so they are pandering to them by shoving AI into everything regardless of resulting enshittification. Foie gras AI for the stock price.
I can't be alone in not wanting these features? I don't mind them being available, but I do fear a nearterm future where they are active whether I ask them to be or not.
I remember all of the scorn clippy got years ago. How is this any different? I think Inbox was probably more useful, and they didn't push it near this hard. :(
It's worse, because Clippy had no editorial control of what was being produced.
I think there's a group of people who really really want this, and they are probably the last people who should get access to an AI/LLM. Some people will just love this, because they're already bullshitting their way through life and this will just make it easier, it even looks company approved if it's in the tools provided to you.
I’m on paid Google Workspace for my one-man business : I paid for a month of the separate AI add-on but I stupidly agreed to an “annual commitment” which means that, even though I don’t use the AI stuff (it’s not particularly useful) I have to keep paying for it every month for a whole year! :-(
Anybody know if this means they’ll let me off my annual commitment now that it’s included in the base price?
Stuck on GSuite Legacy (with my own domain) and Google won't let me give them money to upgrade my storage. Workspace too expensive for family purposes.
Recently got a new phone and can't use Gemini with my old GSuite Legacy account.
No migration path back to personal @gmail.com accounts for my family.
When I moved from an @fastmail.fm email to my own domain years and years ago I just gave them money and added my domain to my account. No fuss.
Google are hopeless. They have all this consumer brand recognition and just squander it on garbage.
Google One + your own personal domain name would be great but presumably they're afraid it'll dismantle Workspace for small businesses.
> Google won't let me give them money to upgrade my storage
While not explicitly documented anywhere, they automatically increase your storage limits once you approach a certain margin of remaining free space. That happens around Tuesday-Wednesday, they just add extra 5Gb to your limit.
We have Workspace with Gemini and I haven't yet found a case where it did something useful for me.
The times that I had it try to find information in my gDrive folders it didn't find what I wanted, and I ended up using search as usual. It was also slower than me searching and looking through the docs.
Free, nothing worth writing home about. Technically you can do this for free with an Apple account but its a total mess of a system and incredibly buggy, not to mention essentially no spam filtering.
Paid you've got ProtonMail and FastMail, both decent options.
Sadly there isn't a single good alternative. I'm switching email to Proton and Drive to BackBlaze for backups and S3 for sharing. Then using Google Sheets shared with me but not sharing back.
it feels like google in panic mode, the only thing it can think of is to put a chatbot everywhere, just b/c it can. I don't see a value proposition at all.
My understanding is that every manager at Google has had one of their quarterly goals be to integrate genAI into their team's product (regardless of whether it makes sense to) for the past several years already, so you're not wrong.
I don't like AI being used in anything remotely creative.
I don't draw, not well, but I write, slightly better. I occasionally ask WordPress to have its AI generate a little blurb for me, and always wind up deleting it. It takes something I can't really describe, my voice I guess, and sucks it out. It homogenizes my writing to try to make it fit some bland ideal. I imagine to those more keen on art than I, AI art is similarly off.
And yes, stories are not the primary use of Gmail. But in business, words matter, and two seemingly synonymous words can be quite different, and two words that seem opposite may not be. I have a friend who teaches law, and they mentioned it was quite easy to tell which students cheated on one particular assignment discussing contracts. If I recall right, material contracts are a type of contract, and AI made up immaterial contracts.
While this mistake would hopefully be obvious, other mixups might not be, with potentially serious consequences.
I agree, though in my job I send many many near-pointless communications. Many jobs you are just required to pump out garbage, and AI is real good at pumping out garbage.
> If I recall right, material contracts are a type of contract, and AI made up immaterial contracts.
Fun fact, a rogue LLM impersonating a human in the 17th century is also how we got the term "imaginary" numbers. It also wrote some truly terrible philosophy but it started with a pithy sentence so everyone remembers it.
In google economics, there is no KPI incentive for fixing bugs and huge KPI incentives for monetising a new moonshot product nobody wants.
I'd like this bug fixed too. The quickest path would be to make a bounty hunt website for Googlers to fix things in their free time and push through monrepo approvals legally.
I'm honestly getting a bit sick of subscription pricing, especially for things like "productivity apps." and G-Suite (although sadly the MS alternative isn't any better).
At the end of the day, we just do the same ol' simple word processing we've done for the last 20 years.
Are there good corporate email alternatives that just do email/calendars and do them well with business-type SLAs? Zoho? FastMail?
Note that they treat any service degradation as downtime when computing 9’s. For instance, they had one imap server down today, which meant some requests were failing, and that counts against their reported numbers.
By this metric: “One machine is failing requests”, most of the hyperscalers are down all the time.
Regarding actual SLAs with money and stuff: How much is the refund worth vs. the cost of downtime?
Edit 2: Take github for example. They have unreported outages all the freaking time. Down detector says push/pull has been flaky for the last 24 hours, but the official status page says all systems operational, with a minor codespace outage yesterday.
Zoho’s SLA I can’t speak to but it’s hard to argue against free forever, including custom domains. For personal use it’s perfect and the paid packages are much better value for money than Google/Microsoft.
FastMail is wonderfully competent at being an email provider, has human support (or advanced enough an AI to fool me) and wildcard domains.
One thing I really loved is automated transcripts on youtube. I love watching youtube videos, but sometimes I want to remember where I heard some statement, so I can just copy paste the entire transcript and do ctrl+f on it.
So sad that they removed this feature. There is third party websites offering it, but I'd prefer it on the main site.
This feature had been added years ago, way before the AI hype was as big as it is now (but it's always been using deep learning models).
We've got a massive communications problem in our society, people do not know how to express efficiently, they under and over document, but rarely document "just so" one can use whatever, understand whatever, without a majority of the information exchange being throw away "wrapping" information, framing information.
LLM AIs are forcing this issue to an apex, if and only if you and your peers realize this working with LLMs is also a communications issue, also one of framing information so both the correct information is delivered and a minimum of wrapping information that needs to be filtered through to understand is not delivered. The same reason you cannot explain to your boss, or coworker, or spouse some troublesome issue preventing a goal is also why you cannot get the quality replies you want from an LLM. You cannot express you request, your information effectively so the audience can understand what you meant.
> [Billing and Service Notice] Google Workspace service and pricing updates
> Dear administrator,
> Starting today, your Google Workspace subscription includes new AI features designed to help your users improve their productivity and innovation. With these changes, we will also be updating subscription pricing starting March 17, 2025.
> ...
> These features were previously available only to users with a Gemini for Google Workspace Add-on, but now will be included with Google Workspace Business Standard plans. You will see these features added to your subscription in the coming weeks. Soon, you'll get access to even more Gemini features in your Google Workspace apps.
> Review the Google Workspace blog announcement to learn more about these changes.
> Starting March 17, 2025*, your Google Workspace Business Standard subscription price will be automatically updated to $14.00 per user, per month with an Annual/Fixed-Term Plan (or $16.80 if you have a monthly Flexible Plan).
I'm not convinced the "Gemini all the things" strategy was the right move with Workspace, they rolled it out so fast which indicates that UX research was likely rushed or bypassed completely. Had they conducted their typical extensive UXR process they would have discovered that the features are not very useful being baked into the suite.
Now where I do think there is opportunity is in building out the standalone Gemini app, as ChatGPT has proven with their Teams product that there is business value in having a dedicated chat UI for your business. We are currently subscribed to ChatGPT for Teams and use it every day across product and engineering, there isn't a need for it to be integrated directly into our productivity suite UX, but pulling data out from the suite (e.g Google Drive) into the chat UI is helpful. Organizing project folders, custom GPTs etc also hold value for us.
Okay, I have a lot of projects with a couple of email addresses attached to their domain name
In the US and European market, this was seen as a bare minimum level of professionalism and validation (other markets are more advanced on this front and have been on chat apps for the entire business for at least a decade)
regardless for email, I had been using Google workplaces for this
What’s a cheaper alternative? last time I tried something else I found I was vendor locked to google even when trying to accept calendar invites from people in other organizations that sent google calendar. That was 5 years ago though
some sectors like web3 let you do the whole project with just a username on discord/telegram/x but I do want to consider migrating my emails away from google workspace now. Its difficult to manage even changing the credit cards on file with so many projects like if one expires
Anyone have a recommendation for moving from Google workspace that has email, docs, calendar, and contacts capability? I don't really need anything else aside from those capabilities. Fastmail isn't sufficient
I accidentally started the Gemini-the-product free trial in Workspace while trying to find how I could test Gemini-the-model in AI studio.
The first task that I asked for it’s assistance with, was how to disable, cancel or unsubscribe from Gemini-the-product. It repeatedly and confidently made up instructions to adjust settings that didn’t exist in menus that weren’t where it said they were and provided links to irrelevant documentation.
It was either useless, actively misleading or extremely motivated to not be turned off.
Any of those was reason enough not to use it ever again.
It's not good enough. The place I most want it is in Sheets and it doesn't know how to fill in random cells with stuff. It can populate new tables and stuff with example data which is fine as a start but I usually want it to copilot my formulae.
It's such an obvious use case and Claude 3.5 Sonnet can give me the answer if I paste the header and a value row but Gemini is utterly useless.
You're in-app. How is a textual copy-paste better in Claude? Useless Google PM. The Oracle Java of AI.
We signed an annual pricing deal in fall and now our price is going up materially during the annual term because of this change, because we didn't buy Gemini.
I clicked the button to turn the forced AI crap off today in Gsuite. It has been on for a day (unasked, forced my google). The engineering team responded with a unanimous thumbs up when I told them. Winner of best response was:
"oh god! now i have to type complete senten..... zzzzzzzz"
Already had to migrate all my domain's off Squarespace and remove Google Workspace entirely when they randomly decided it was now not free and started billing me monthly more than Google Domains did (80% more) Some MBA got a bonus for this surely.
I would buy $GOOG stock blindly but being a paid user of theirs blows
"We've invested all the money in AI that no one wants, how do we make some of it back?"
"Why not raise the price for everyone, whether they use it or not?"
Ugh, we got hit by this. We're a non-profit with business plus licenses, so we get a decent discount. I opened a support chat as soon as I got the email about the license changes and price increase. Our contract is due next month, so I need to make sure we're OK budget wise. They can't even tell me how much our renewal will be yet. They don't know.
This morning, I logged on to find that the AI features have been turned on domain-wide for us. I couldn't find any admin controls, so I opened up a support case. The off buttons are locked behind an enterprise subscription. Our end-users need to turn off smart features to disable Gemini. There's no domain-wide / admin level control unless you purchase their most expensive licenses. It's absolutely disgusting. I'm so disappointed with how this was rolled out. We should've been given an opportunity to make an informed and intentional decision about how or if we were going to use these features.
Business are not willing to pay for Google and Microsoft artificial intelligence so they ramp up the price to everyone and everyone get it 'for free'.
Untill they eventually get hooked on that and then google and Microsoft will once again put that behind a paywall, except now everyone pays more. At least that's the plan.
Now even if employee don't see the benefit of the new deep integrated a.i. and business refuse to pay more for a.i., they aren't going to leave anyway because Microsoft is doing just the same as google.
That's either a win for Google and Microsoft, or at least a neutral outcome.
Plus Google gets to use your data for training.
That has interesting implications. What goes in as training data often comes out later as replies to questions.
If I remember correctly, the Office 365 Copilot thing is more of an upsell rather than an upcharge. Basically, if you didn't want to pay more, you would initiate the cancellation process, and during that you could "downgrade" to the plan that you already had (without copilot as part of it).
Personally, I find that to be especially scummy because it essentially sounds like they are betting on people either not understanding that nuance, or not bothering to deal with it (and subsequently, not using AI, making that venture seem vaguely more profitable)
Pretty simple formula, there simply isn't a market for an $x upcharge on mail and docs, but we have to be part of this latest grift so we'll charge everyone $x - y which is a rounding error. Except it's not. At some point someone is going to admit they have bet the farm on improved auto complete.
Is it just me or is this title a clear contradiction? Free but increases price? Do they mean now included in your cost…which is increasing? Doesn’t have the magic marketing power I suppose.
I just saw this. This is probably most useless feature probably nobody asked for except for marketing, to be able to claim they have AI email blah blah.
>Workspace AI includes things like email summaries in Gmail, generated designs for spreadsheets and videos, an automated note-taker for meetings, the powerful NotebookLM research assistant, and writing tools across apps.
Maybe I'm just an old curmudgeon stuck in my ways, but I haven't found much compelling value in these use cases in my day-to-day work. For summaries and note-taking specifically, I feel they're solving the wrong problem: it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
The real solution to too much email is fewer and higher-priority emails. The real solution to too many meetings is fewer and more-focused meetings. These tools paper over the root cause of the problem, which is that people/organizations cannot (or are unwilling to) be clear about communication priorities and say "maybe this email/meeting isn't a good use of time after all."
How is AI in email a good thing?!
There's a cartoon going around where in the first frame, one character points to their screen and says to another: "AI turns this single bullet point list into a long email I can pretend I wrote".
And in the other frame, there are two different characters, one of them presumably the receiver of the email sent in the first frame, who says to their colleague: "AI makes a single bullet point out of this long email I can pretend I read".
Right now at Amazon we are going through the annual feedback cycle where you have to write strengths and growth areas for your colleagues. You will usually have to do ~12 of those.
I don't use ChatGPT for those, but it is the epitome of what you are describing, people will take a single sentence, ask some LLM to blow it up the correct length and in the process make it a complete waste of time for everyone.
My guess is that with long-form text losing value due to LLMs, we will see a return of very succint 1-2 lines employee feedback.
I just exited the toilet following 2.5 hours of back-to-back meetings, and was looking forward to actually getting some work done when the product owner grabbed me for a conversation about priorities for the sprint planning session that's scheduled in a couple of hours.
In this week so far (first week back from Christmas / New Year leave) I've spent maybe half a day total on work that could be classified as "progress". The rest of the time has been meetings and the required meeting follow-up work.
There's no point in Sprint Planning or considering adding priorities to the current plate. It's full. But nobody has time to eat things off the plate because we're always in meetings to work out how we can eat off the plate more efficiently.
/rant
I've come back from holidays angry. Things gotta change.
I had a few use cases with searching and organizing emails I would have used. For example, I wanted a table of all my Lyft rides from a certain year with distances driven, start/end locations, cost, etc. All that info is available in the email you get after riding, so I figured Gemini could read my emails and organize the info.
Turns out it doesn't work at all. It gave me a random selection of rides, was missing info in some of them, and worst didn't realize it was giving me bad info. Pretty disappointing.
I find AI meeting transcripts and summaries to be one of the most genuinely useful things to come out of this era of LLM tools. Being able to see a quick summary of what was decided or who was supposed to do what next is just so helpful, either for refreshing your memory after the weekend or just because people aren’t all that great at taking and sharing notes.
These LLMs are excel at making more. More emails with more words. More blog posts with more fluff. Making it open to more people means more usage means more numbers being more which means more money for the people building these systems.
I don't see what I get out of 80% of these products. It's just more noise.
I can hardly wait to use it as an excuse. "Oh sorry I didn't do that because it wasn't in the AI summary" ;)
> For summaries and note-taking specifically, I feel they're solving the wrong problem: it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
I think this really encapsulates something that I hadn't been able to put my finger on in regards to LLM summarization. What it seems to indicate is that, if you need a computer to summarize a large amount of text that someone has sent to you, there are two likely possibilities:
1) The information is incredibly dense/important/technical/complex. This necessitates the extreme length of the message - (think: technical documents, research papers, a rough draft of a legal notice, or your will.) For these sorts of things, you should not rely on an LLM to summarize it, because it may miss key details of the message.
2) The person sending it to you is bad at communicating, in which case the solution is help them learn better communication, rather than "de-noising" their clumsy wording into something comprehensible.
"But what if its number 2, but it's coming from your boss?"
Then I see two obvious points to consider:
First, you should absolutely be telling them about the problem, regardless of the position that they hold. You can phrase it in a way that isn't rude. "Hey boss, I saw (message) but I'm not 100% the intent. I've actually noticed that with (other time)...I usually try to front-load the action items up front, and put the specifics lower down. Anyway, to make sure I'm tracking, you're talking about (action) on (thing), right?"
Second, until (or unless) their communication style is de-noised, then part of your job is being able to "translate" their instructions. Using an AI to do that for you is a bad idea because, at some point or another they're going to be trying to speak to you in-person, or by phone.
Not having dealt with their mannerisms in an unfiltered way might lead to you being "out of practice" and struggling much harder to figure out what they're trying to convey.
Well, that’s because you’re thinking as someone who likely has a stake in quality/specific outcomes actually happening. Or was raised/grew up in an environment where that was important.
Notably, in my experience there is a high correlation with that background and being curmudgeonly. Mainly because that means someone has been responsible for outcomes, regardless of feelings. And something often has to give, and it’s usually feelings. It’s also hard to not be cranky or even angry if someone has to constantly be the one ‘not having fun’ or cleaning up messes so the whole thing doesn’t fall apart.
There is huge market demand exactly for what you’re complaining about, which is faking things happening as convincingly as possible, precisely because being clear/concise, etc. helps with seeing the root cause of problems, and if someone is worried (or is legitimately) a root cause of the problem, of course they’ll consider that bad.
For example, a good sign of a badly led organization is that it’s always busy, but never seems to get anything done. Everything is an emergency, so nothing really gets fixed, etc.
Or there are constant meetings and emails, but nothing gets decided.
People will pay good money for the right kind of wallpaper that makes that ugly wall look pretty again.
> The real solution to too much email is fewer and higher-priority emails.
Sure, and that's an actionable solution if you can control the actions of everyone else who emails you.
> I haven't found much compelling value in these use cases in my day-to-day work.
If my experience with Microsoft Office Copilot is any indication, these features produce very confusing, low-quality content if they are not completely wrong and useless. Used it once and never touched them again. (My company is still paying for this and rolling this out widely despite many reports of how unhelpful they are.) I doubt Google Workspace can do any better.
I don’t use it often either, but sometimes it is very useful. When I caught Covid last fall my wife incorrectly thought I had it three times. I was using a beta Google Gemini, and paying for it, and I asked “read my @gmail and tell me the date ranges when I have had Covid.”
That worked, but to be honest I have tried similar things more recently that didn’t work. Perhaps there is a routing model up front that decides whether or not to use a lot of compute for any given query?
Google also plans on charging more money for APIs for code completion plugins for IntelliJ IDs, etc. this year.
I would like to see AI pricing models be sustainable, not give things away for free, and have lots of control over when I use a lot of compute. I actually have this right now because I usually use LLM APIs and write my own agents for specific tasks.
I'm kind of a cynic, so I'd say that the Workspace customer isn't you, the person who's using Workspace. It's your big company's SVP of IT or whoever who wants to spend money to adopt cool AI stuff so that he can say that he did AI stuff.
> it's not that I have all this information that I really want to go through, but it's that I have too much information and it's become all noise.
I tend to agree, except these two things are kind of the same thing. It can make going through the noise easier by intelligently filtering out the noise or finding you the signal. Search. It doesn't necessarily need to eliminate the noise.
Maybe AI would be better if it prevented the noise, and its definitely going to add noise (expanding a few basic thoughts into an email with lots of fluff), but it can also solve it.
I’m getting a lot of value out of NotebookLM drafting documents. If I’ve got a bunch of notes that need to be in a coherent design doc, it can give me a good enough first draft for me to edit into shape. Alternatively when I’ve got a design doc for something, but need to submit, say, a work request to another org, NotebookLM can take my doc and turn it into another format based on a doc template pretty nicely.
These outputs still require editing for sure, but each one can easily save me half the time to write these things.
I agree. I don't want all my existing work apps to take on LLM features I don't need.
At the same time, I tried the Gemini Research feature last night, via the Gemini webapp, and was resoundingly impressed. From a vague description, it can find the open source project I was looking for, provided ample links, and a pretty good summary of the project.
deets: https://news.ycombinator.com/item?id=42706997
I totally agree. I upgraded to the AI-enabled version of Google One because they gave a couple week free trial. I found it totally useless, and it reeked of "Some PM said we had to stuff AI in everywhere".
Note I do use ChatGPT pretty frequently, but I've found it much more useful to have a separate space for the kinds of conversations I have with ChatGPT.
I feel quite the opposite.
I’m not a native english speaker, but working at US subsidiary I must produces reports in english etc - and having an LLM proofread my texts for me is great.
LLM:s are new modality to computing. If you need it, they are great. But just like excel/sheet have limited applications a LLM with data has limiited use as well.
I also find that summarising content helps me digest it better. I have to fully understand the source in order to write the summary. The process of writing a summary is of immense value. Sometimes the summary itself is of minimal value.
What if there was something that communicated the company’s top priorities and helped everyone align and stay organized without so many meetings, and give concise drafts for your to-dos? Would that be something you’d try?
Yeah I’m tired of workspace getting more expensive and me getting zero additional value from it. I don’t want this, didn’t ask for it, and it actively annoys me.
Enshittification #353: solving cuStomers problems has poor ROI
management uses them to fluff up their emails and I use them to boil the emails down to actionable bullet points.
It's decent at search
I saw a Google AI advert that said:
"Hey Gemini, write an apology email for my friend. I can't make their wedding."
That's not a future I want to live in, and I love making machines work for me.
Thats not what I want my children to think is OK.
A friend of mine is a teacher and kids are already delegating their learning to ChatGPT and their learning isn't sticking.
What happens when social skills are delegated too?
I guess the future is
1. Friend sends an apology email drafted by LLM.
2. Email gets summarized at the receiver end in the daily AI email "summary" which might be something like
You have a scheduled cake tasting this weekend. Did you know there's a bakery near your office that makes wedding cakes too. By the way your friend Joe can't make it to the wedding, do you want me to send a reply?
3. Reply email gets summarized by AI.
"Your friend acknowledges that you cannot rsvp. Do you want to schedule a wedding gift delivery on their wedding day ? XYZ neighborhood/online store has a sale next week".
If you really care about this issue, I think we've brought it on ourselves.
Regarding teaching kids, we've set messaging templates for occasions that are at the center of our lives. We have Hallmark greeting cards to express feelings to people close to our hearts. If there's a template for expressing someone you're sorry their mother died, or happy they have a baby, I'm not sure throwing the stone at AI use is warranted.
In a way, I wonder if it will be the wake up call that will make simple and genuine communication acceptable again, without all the boilerplate we've built to feign care and emotions.
Gemini's marketing is so bad. This isn't the first time they ran an ad that makes you wonder what's going on there. It really says a lot that an advertising company understands what makes for good advertising so poorly these days.
> A friend of mine is a teacher and kids are already delegating their learning to ChatGPT and their learning isn't sticking.
I’m not going to defend AI here because I seldom use it myself. But it should be noted that the way we learn has already undergone multiple different shifts due to changes in technology.
Search engine were a big one. No longer did we have to learn to memorise stuff nor learn how to research properly. Now we could just type a phrase into Google / whatever and get results. So people learned how to search rather than learning the facts itself.
"Hey Gemini, maintain my friendships"
... back to Fortnite / Minecraft / pr0n / alcohol / drugs ...
"My AI has more friends than your AI!"
you’ll just have your ai email reader read the apology emails for you
> "A friend of mine is a teacher and kids are already delegating their learning to ChatGPT and their learning isn't sticking."
What about this:
https://blogs.worldbank.org/en/education/From-chalkboards-to...
Second law of thermodynamics says these models will all eventually collapse (due to overtraining on their own output) to yelling gibberish at us, and biology will continue to remain the only force in the universe capable of maintaining order despite increasing entropy. I think we'll be OK.
I pay for 3 Workspace orgs, and I have Gemini disabled (or still not enabled) on all 3 of them. I'm angry that I'll have to pay more for features I don't use. Gemini should be an add-on cost, not included in the base cost
If they're raising the price of personal GMail, I don't have a problem. But Workspace with hundreds of users, now that's a problem, because it actually hurts my wallet significantly. When this increase comes, I'll have to move elsewhere.
It’s an anti competitive strategy, which in an ideal world would see them facing a crushing antitrust lawsuit from the FTC and DOJ. What they’re doing is forcing everyone to pay for their AI product. This makes it so that no other company can charge for their alternative AI products. After all, if your company’s spending goes up because of this Google price increase, your executives will not want to see double spending on AI products. So all those deserving smaller companies will miss out on these customers. Google is essentially using this forced price increase to kill their AI competitors by stealing their revenue, through illegal bundling. Just like Microsoft did with Teams to attack Slack illegally.
With Amazon as an example for CxOs of the world, sadly, this likely won't happen.
Look at Prime. So much crap involved, and quite literally all I use it for is lower cost shipping. It's almost on the edge of not worth it for me. But I bet from Amazon's perspective, they make more with the higher price, even if they lose the bottom 10% not willing to spend.
Huh.
Just made me realise, a startup that subscribes to Prime as a virtual being, and then splits off each sub-thing for full use by separate individuals would be incredibly profitable.
If any form of AI is eventually granted legal personhood, Prime's model will collapse.
I agree.
It feels like Google are shoving AI down our throats and making Workspace customers pay for it's development.
I don't want your half-baked LLM features.
Right now looking for an alternative for the same reason. Even if it cost me more on labor short term. They have been increasing prices regularly and I’m sure it will continue.
Any idea where you'll move? I have a nonprofit I want to migrate away.
Why do you have Gemini turned off?
We are doing a Gemini POC and this nugget dropped in my lap today. We were not entirely unprepared as a result. The default level of access is just the interactive chatbot thing. However if you enable the Google Workspace extension it will be able to search and process all the information stored in your workspace account and also any Google Drive files that are shared with you. This includes stuff you didn't know you had access to in Shared Drives so folks better make sure their permissions are locked down. Workspace admins might be advised to turn it off at the org level until they understand the ramifications.
Reminds me of an entertaining story about Microsoft Copilot last year, where companies were turning it off because it turned out it was TOO good at its job - if any accountant anywhere in the company had messed up their SharePoint permissions asking "what does everyone at this company earn?" would spit out all of the salaries: https://simonwillison.net/2024/Aug/23/microsoft-copilot-data...
I am surprised the Workspace extension isn’t controlled by the same setting that limits general workspace search results, where you can set things up so only documents you’ve seen or are linked to from documents you have explicit access to are returned in results: https://support.google.com/a/answer/12732365?hl=en
Workspace was $12/month, now it will be $14 with AI included. AI was $20/month.
Looks like AI as an add-on wasn't selling too well.
Users will continue to be beaten with the AI cudgel until morale improves.
Pretty much. A small set of customers weren't willing to pay for AI? Now everyone has to pay for AI.
Shid. I made the mistake of getting my entire family onto my google apps 15+ years ago. Now I am paying for about 8 people every month and this will just make it worse.
It was selling well enough. It just was not getting enough traction. By bundling AI, they are giving exposure to everyone who didn’t want to use it or didn’t see the need for it. If they pulled it away in 2 years, and then lowered the price and charged separately for AI, I think more people would see it as necessary.
AI is a better search engine. And a better grammar check for your emails. And a better writer for your reporting.
I expect take up was in the low single-digit percentage points. So charging every single subscriber $2/user (even if they don't want it) probably yields significantly more revenue.
Can I pay $20 to keep the version without AI?
Cool, great, fun. I have all of the “generative AI” features disabled in Workspace, and now I get to pay more for the privilege of keeping them disabled. Thanks, Google!
Same. This is bullshit.
Nobody on my team uses these features. They're actually quite distracting.
Google gets to raise prices under the guise that these are improving productivity.
I wish there was a fast and easy alternative. Google has its claws in deep.
Do you know how to actually disable these new features (i.e. the elements that were added within Gmail, Docs, etc.)? I'm not seeing where they can be disabled and Google Workspace support was not able to point me in the right direction either...
I recently got Gemini Advanced as an additional benefit by virtue of having Google One paid storage. I'm shocked this is being given away for free, because it is now a seriously major part of my work. I literally have an Open window all day long interacting with it. It does make me wonder how much they are losing (investing) on giving all this inference away for free. Also makes me wonder what they are getting back aside from loyalty/data/?
I always felt ripped off by the 5TB/10TB plans (https://one.google.com/about/plans?hl=en&g1_landing_page=0) but now I find it to be a bargain with Gemini bundled in.
What are you using it for? It has been completely subpar compared to any other LLM for me.
It's so bad at understanding your intentions.
Is Google One the same as just having extra storage for my Google Photos? I have that but just went onto Gemini and Advanced will cost me $33 pm.
What are the odds that they will tally that extra $2/user/month up as "AI revenue" regardless of how many subscribers actually use those features?
100%
Give it a quarter and we’ll see breathless articles about how Google saw “AI adoption increase 150%” and “Google workspace users say they can’t go without AI” (because they physically can’t remove it from their workspace).
This in turn, will be used as post-hoc justification of the value of AI and why ever more power, water and data should be funnelled into it.
I tested Gemini today, asking it to extract key pieces of data from a large report (72 slide) PDF deck which includes various visualisations, and present it as structured data. It failed miserably. Two of the key stats that are the backbone of the report, it simply made up. When I queried it, it gave an explanation, which further compounded its error. When I queried that, extracted the specific slide, and provided it, it repeated the same error.
I asked Claude to do the same thing, it got every data point, and created a little react dashboard and a relatively detailed text summary.
I used exactly the same prompt with each.
Maybe the prompt you used was more Claude-friendly than Gemini-friendly?
I'm only half-joking. Different models process their prompts differently, sometimes markedly so; vendors document this, but hardly anyone pays any attention to it - everyone seems to be writing prompts for an idealized model (or for whichever one they use the most), and then rate different LLMs on how well they respond.
Example: Anthropic documents both the huge impact of giving the LLM a role in its system prompt, and of structuring your prompt with XML tags. The latter is, AFAIK, Anthropic-specific. Using it improves response quality (I've tested this myself), and yet as far I've seen, no BYOK tool offering multiple vendor support respects or leverages that.
Maybe Gemini has some magic prompt features, too? I don't know, I'm in the EU, and Google hates us.
That matches with my experience, Claude is clearly ahead of its competitors in anything logic- or reasoning-based.
I find Gemini is better at queries that involve more kind of intuitive judgment over things where there isn't a clear "correct" answer. E.g. if I want a podcast recommendation, or advice on the best place to learn about a given problem, I find Gemini better than Claude.
Unfortunately for Gemini, 90% of the things I want an LLM for are better with stronger logic and reasoning.
I got a 1-year trial of Gemini Advanced with my Pixel 9 and I've had similar experiences. It makes up stuff far more often than any other models and it's just not very smart. I used the free version and thought the paid Advanced version would be better but I could hardly notice any difference, they both fail at the same prompts I've tried.
This is not to mention the poor app experience where some of the features are just missing or broken. For example it's able to "remember" stuff I ask it to remember, but when I ask it to forget something it says I have to manage it at this webpage (they didn't bother to implement this menu within the mobile app) that asks me to sign in again because it's opened in my web browser where I'm not signed into Google, and then it shows me an empty list and "Something went wrong". It's now calling me a name I told it as a joke and there's no way to make it forget
Ugh. The "vanity domains for gmail" product i've been buying for a long time is really metastasizing into something that's both too expensive and actually worse than the free experience, wonderful.
It’s weird that prediction 8, "Someday [you] will voluntarily pay Google for one of their services" has come around full circle to "and then you won't anymore, because they've dropped the ball to an extent usually associated with the private equity buyout -> loot into bankruptcy process"
You'll have to rip that band-aid off eventually, may as well get it over with. It's only going to get worse.
I switched from G Suite to Fastmail for my custom domain and I've been very happy with it.
This spurred me to go back and read the predictions:
>But I can tell you this: Google has changed my life. If I can't find what I'm looking for in Google in 3 tries, looking no further than the first 10 search results on each try, then it probably doesn't exist.
What a sad future we're in.
Plus now I'm noticing it doesn't work for more and more things. Youtube TV family sharing doesn't work, Android Auto had some problems, the news feed on my Pixel.
Wish Google would just fix the Drive search rather than lard it up with AI nonsense. Often it’s easier to ask someone to resend the link to a document than find it by searching.
There are so many bugs and sub-par implementations in workspace that Google could fix. My cynical guess is that the source code to workspace apps is probably a mystery to the current generation of 23 year olds who are tasked in maintaining them, so they change little.
I wish they fixed search in general. It is difficult to find emails if you don't know exact keywords that might have been used etc. often even if you type in the right keyword it still won't find the email, even though email contains it.
Worst thing is people sharing files tbh, if someone has a folder and shares you a multiple documents from it you don't get the folder in your drive structure so you have n free files floating around in your drive that you cannot organize yourself.
I'm in the middle of a free trial for the Workspace Gemini add-on.
It really, really sucks. I've played around with having it make tables for Sheets and it frequently gets confused or responds with ~"I can't do that, I'm just a LLM", even when feeding it one of their suggested examples word for word. Sometimes it's willing to iterate, sometimes it refuses. Once it gets confused,the only way I've been able to get it working again is by clearing the session and starting fresh.
And it's sloooow.
None of this saves me any time or frustration.
I guess this is why it is being bundled, Google can keep working on it with someone else's money, so their profits aren't hit. It's telling that the increase is regardless you use Gemini or not.
FYI If you want to turn this off in workspace you'll need to go here https://admin.google.com/ac/managedsettings/47208553126 and here https://admin.google.com/ac/managedsettings/793154499678.
Forcing you to pay for features you never asked for and won't use. I'm sure this will work out great for google in the long term.
There's a weird attitude on this site towards AI: if it's for coding or science, people generally recognize AI tooling as effective, although imperfect, and rapidly improving.
But any suggestion of using AI for business and it seems like disdain and dismissal is the majority response.
Don't you think many areas of business - maybe not all areas, but a significant amount - will just as much benefit from AI tooling as software developers and scientists?
They kinda already do it with YT Premium/YT Music. I don't have anything against YT Music, it's a perfectly fine music service from the amount I've used it. But I already have a Spotify with my preferred playlists, and I don't really have incentive to swap it over aside from maybe saving a handful of dollars a month.
Yet if you want ad-free YouTube the proper way, you can't just have Music as an option, it's rolled into your cost regardless of if you actually plan to use it or not.
Isn't that always the case with bundles and suites? Google Workspace has always been a bundle of products, and few actually used every product in the bundle.
They have managed to vendor lock me. The price just keeps going up and I can't get out.
The beauty of having a monopoly or oligopoly in a dozen major markets is that you don't have to care about customers. As much as I hate this move and don't think it will help the company, I think Google is powerful and entrenched enough that it will make out just fine. Their users will bear most of the costs.
Does this apply to the legacy free edition? I suspect not, since that edition is now only available for personal use and they mostly focus on Business and Enterprise use cases, but their public guidance isn't very clear. If it does apply, would we legacy free edition users be receiving Gemini under the Google Workspace Terms of Service preventing them from using our data for general AI training, or under the regular Google Terms of Service which might allow this?
(Tangent: I say "might allow this" because I don't know to what degree EU law requires some additional level of consent beyond accepting the Terms of Service for EU-based accounts like mine currently is, or requires them to give me an AI-specific opt-out despite having a free account. But this announcement doesn't change whatever EU law does or doesn't require, so that is unrelated to my main questions about which Gemini features will apply to the legacy free edition under which Terms of Service once this change rolls out.)
The legacy free edition includes the features of the 'Business Starter' plan, the most affordable option. In this table [1], you can find the features available for each plan.
Here are the details for the Business Starter plan specifically:
Gmail: Help me write, Side panel, Contextual smart replies (Coming soon).
Gemini app: Enterprise-grade security & privacy, Google Workspace extensions.
NotebookLM: Upload sources, create summaries and Audio Overviews, and Q&A.
I'm also milking Google with this.
[1] https://support.google.com/a/answer/15400543
There's no legacy free edition for personal use any more. That ship sailed in 2022. I do not believe there is a way to have it free after 2022. Free plans were converted to Business Starter.
Oh no nobody’s buying your ai vaporware, let’s make everyone suffer!
Exactly. It's easy to make something free when nobody is remotely interested in paying for it. If Google wanted to make money they'd have to let people pay to remove AI from gmail and docs.
Investors are the customer, so they are pandering to them by shoving AI into everything regardless of resulting enshittification. Foie gras AI for the stock price.
I can't be alone in not wanting these features? I don't mind them being available, but I do fear a nearterm future where they are active whether I ask them to be or not.
I remember all of the scorn clippy got years ago. How is this any different? I think Inbox was probably more useful, and they didn't push it near this hard. :(
> How is this any different?
It's worse, because Clippy had no editorial control of what was being produced.
I think there's a group of people who really really want this, and they are probably the last people who should get access to an AI/LLM. Some people will just love this, because they're already bullshitting their way through life and this will just make it easier, it even looks company approved if it's in the tools provided to you.
Clippy was better.
I’m on paid Google Workspace for my one-man business : I paid for a month of the separate AI add-on but I stupidly agreed to an “annual commitment” which means that, even though I don’t use the AI stuff (it’s not particularly useful) I have to keep paying for it every month for a whole year! :-(
Anybody know if this means they’ll let me off my annual commitment now that it’s included in the base price?
> If you previously purchased Gemini for Google Workspace, you won't be charged for it after January 31, 2025.
See https://support.google.com/a/answer/15400543#zippy=%2Cwhats-...
Ask support.
What’s the worst they can do? Say no?
Stuck on GSuite Legacy (with my own domain) and Google won't let me give them money to upgrade my storage. Workspace too expensive for family purposes.
Recently got a new phone and can't use Gemini with my old GSuite Legacy account.
No migration path back to personal @gmail.com accounts for my family.
When I moved from an @fastmail.fm email to my own domain years and years ago I just gave them money and added my domain to my account. No fuss.
Google are hopeless. They have all this consumer brand recognition and just squander it on garbage.
Google One + your own personal domain name would be great but presumably they're afraid it'll dismantle Workspace for small businesses.
> Google won't let me give them money to upgrade my storage
While not explicitly documented anywhere, they automatically increase your storage limits once you approach a certain margin of remaining free space. That happens around Tuesday-Wednesday, they just add extra 5Gb to your limit.
We have Workspace with Gemini and I haven't yet found a case where it did something useful for me.
The times that I had it try to find information in my gDrive folders it didn't find what I wanted, and I ended up using search as usual. It was also slower than me searching and looking through the docs.
For small startups, what are some good alternatives to Workspace?
I use workspace due to familiarity with Gmail, and no other reason. Would love to know some cheap/easy alternatives.
Free, nothing worth writing home about. Technically you can do this for free with an Apple account but its a total mess of a system and incredibly buggy, not to mention essentially no spam filtering.
Paid you've got ProtonMail and FastMail, both decent options.
Sadly there isn't a single good alternative. I'm switching email to Proton and Drive to BackBlaze for backups and S3 for sharing. Then using Google Sheets shared with me but not sharing back.
I use Migadu for email. Great service that doesn't get in my way of using my email the way I need to (which isn't anything crazy honestly).
Are they any free email services that allow you to use your own domain?
Proton has a pretty good suite of tools for a similar price.
I just gave up my domain and went back to gmail.
I used amazon workmail for a while.
it feels like google in panic mode, the only thing it can think of is to put a chatbot everywhere, just b/c it can. I don't see a value proposition at all.
My understanding is that every manager at Google has had one of their quarterly goals be to integrate genAI into their team's product (regardless of whether it makes sense to) for the past several years already, so you're not wrong.
Isn't that the whole industry right now?
It’s time saas apps realise that they can’t make 2.5x normal license money by just sticking AI to it.
In our SaaS we added it for free. We realised that there is no way to sustainably make money off of this in long term.
It’s a great feature but not 2.5x price worth feature.
So, a 16% price increase and AI is included?
Sounds like a terrible tradeoff
I can't wait for the LLM hype train to die
I don't like AI being used in anything remotely creative.
I don't draw, not well, but I write, slightly better. I occasionally ask WordPress to have its AI generate a little blurb for me, and always wind up deleting it. It takes something I can't really describe, my voice I guess, and sucks it out. It homogenizes my writing to try to make it fit some bland ideal. I imagine to those more keen on art than I, AI art is similarly off.
And yes, stories are not the primary use of Gmail. But in business, words matter, and two seemingly synonymous words can be quite different, and two words that seem opposite may not be. I have a friend who teaches law, and they mentioned it was quite easy to tell which students cheated on one particular assignment discussing contracts. If I recall right, material contracts are a type of contract, and AI made up immaterial contracts.
While this mistake would hopefully be obvious, other mixups might not be, with potentially serious consequences.
I agree, though in my job I send many many near-pointless communications. Many jobs you are just required to pump out garbage, and AI is real good at pumping out garbage.
> I don't like AI being used in anything remotely creative.
Should be good for the workplace then.
> If I recall right, material contracts are a type of contract, and AI made up immaterial contracts.
Fun fact, a rogue LLM impersonating a human in the 17th century is also how we got the term "imaginary" numbers. It also wrote some truly terrible philosophy but it started with a pithy sentence so everyone remembers it.
This is annoying and badly done IMO. I also don't like or need the features and wouldn't pay for them if I didn't have to.
That said, what you get from Google for a few dollars / month is so far over and above any other SaaS that I'm happy to keep paying (and paying more).
I'd rather Google fix the calender integration in Gmail first.
I used to get automatically created calender events from Gmail for hotels, flights, etc. This was really nice.
But somehow it stopped working well recently. Some emails were not regonized at all (booking.com). Some flight emails are missing return flight.
In google economics, there is no KPI incentive for fixing bugs and huge KPI incentives for monetising a new moonshot product nobody wants.
I'd like this bug fixed too. The quickest path would be to make a bounty hunt website for Googlers to fix things in their free time and push through monrepo approvals legally.
Or, get hired, fix it, and resign.
I'm honestly getting a bit sick of subscription pricing, especially for things like "productivity apps." and G-Suite (although sadly the MS alternative isn't any better).
At the end of the day, we just do the same ol' simple word processing we've done for the last 20 years.
Are there good corporate email alternatives that just do email/calendars and do them well with business-type SLAs? Zoho? FastMail?
In my experience, FastMail has better uptime than any of the stuff work relies on. (It feels like one more nine, but I haven’t checked.)
Maybe you could have them randomly suspend your accounts for a few hours here and there to match the public cloud experience. :-)
Edit: Here’s their outage page, which reports > 2-3 nines for most subsystems most months:
https://fastmailstatus.com/
Note that they treat any service degradation as downtime when computing 9’s. For instance, they had one imap server down today, which meant some requests were failing, and that counts against their reported numbers.
By this metric: “One machine is failing requests”, most of the hyperscalers are down all the time.
Regarding actual SLAs with money and stuff: How much is the refund worth vs. the cost of downtime?
Edit 2: Take github for example. They have unreported outages all the freaking time. Down detector says push/pull has been flaky for the last 24 hours, but the official status page says all systems operational, with a minor codespace outage yesterday.
Compare:
https://www.githubstatus.com/
To:
https://downdetector.com/status/github/
To prove those aren’t all false reports, next time they go offline for you, go bask in the green light their status page.
Zoho’s SLA I can’t speak to but it’s hard to argue against free forever, including custom domains. For personal use it’s perfect and the paid packages are much better value for money than Google/Microsoft.
FastMail is wonderfully competent at being an email provider, has human support (or advanced enough an AI to fool me) and wildcard domains.
> Are there good corporate email alternatives that just do email/calendars and do them well with business-type SLAs? Zoho? FastMail?
There are literally tons of them.
One thing I really loved is automated transcripts on youtube. I love watching youtube videos, but sometimes I want to remember where I heard some statement, so I can just copy paste the entire transcript and do ctrl+f on it.
So sad that they removed this feature. There is third party websites offering it, but I'd prefer it on the main site.
This feature had been added years ago, way before the AI hype was as big as it is now (but it's always been using deep learning models).
Did they recently hire some Microsoft PMs?
"Our shiny new product isn't selling. How do we pump up the numbers?"
"Bundle it into another popular product, of course."
We've got a massive communications problem in our society, people do not know how to express efficiently, they under and over document, but rarely document "just so" one can use whatever, understand whatever, without a majority of the information exchange being throw away "wrapping" information, framing information.
LLM AIs are forcing this issue to an apex, if and only if you and your peers realize this working with LLMs is also a communications issue, also one of framing information so both the correct information is delivered and a minimum of wrapping information that needs to be filtered through to understand is not delivered. The same reason you cannot explain to your boss, or coworker, or spouse some troublesome issue preventing a goal is also why you cannot get the quality replies you want from an LLM. You cannot express you request, your information effectively so the audience can understand what you meant.
> [Billing and Service Notice] Google Workspace service and pricing updates
> Dear administrator,
> Starting today, your Google Workspace subscription includes new AI features designed to help your users improve their productivity and innovation. With these changes, we will also be updating subscription pricing starting March 17, 2025.
> ...
> These features were previously available only to users with a Gemini for Google Workspace Add-on, but now will be included with Google Workspace Business Standard plans. You will see these features added to your subscription in the coming weeks. Soon, you'll get access to even more Gemini features in your Google Workspace apps.
> Review the Google Workspace blog announcement to learn more about these changes.
> Starting March 17, 2025*, your Google Workspace Business Standard subscription price will be automatically updated to $14.00 per user, per month with an Annual/Fixed-Term Plan (or $16.80 if you have a monthly Flexible Plan).
I'm not convinced the "Gemini all the things" strategy was the right move with Workspace, they rolled it out so fast which indicates that UX research was likely rushed or bypassed completely. Had they conducted their typical extensive UXR process they would have discovered that the features are not very useful being baked into the suite.
Now where I do think there is opportunity is in building out the standalone Gemini app, as ChatGPT has proven with their Teams product that there is business value in having a dedicated chat UI for your business. We are currently subscribed to ChatGPT for Teams and use it every day across product and engineering, there isn't a need for it to be integrated directly into our productivity suite UX, but pulling data out from the suite (e.g Google Drive) into the chat UI is helpful. Organizing project folders, custom GPTs etc also hold value for us.
Okay, I have a lot of projects with a couple of email addresses attached to their domain name
In the US and European market, this was seen as a bare minimum level of professionalism and validation (other markets are more advanced on this front and have been on chat apps for the entire business for at least a decade)
regardless for email, I had been using Google workplaces for this
What’s a cheaper alternative? last time I tried something else I found I was vendor locked to google even when trying to accept calendar invites from people in other organizations that sent google calendar. That was 5 years ago though
some sectors like web3 let you do the whole project with just a username on discord/telegram/x but I do want to consider migrating my emails away from google workspace now. Its difficult to manage even changing the credit cards on file with so many projects like if one expires
Anyone have a recommendation for moving from Google workspace that has email, docs, calendar, and contacts capability? I don't really need anything else aside from those capabilities. Fastmail isn't sufficient
So they're raising prices just because they're now including some AI nonsense?
We’ll surely see a note is their next earnings release about the uptick in Gemini usage
If I shut down my Google Workspace account, will all the videos my users have posted on YouTube be taken down?
Also, long ago, it was possible to set up an individual Gmail account with a non-gmail.com domain. Is that still possible?
I accidentally started the Gemini-the-product free trial in Workspace while trying to find how I could test Gemini-the-model in AI studio.
The first task that I asked for it’s assistance with, was how to disable, cancel or unsubscribe from Gemini-the-product. It repeatedly and confidently made up instructions to adjust settings that didn’t exist in menus that weren’t where it said they were and provided links to irrelevant documentation.
It was either useless, actively misleading or extremely motivated to not be turned off.
Any of those was reason enough not to use it ever again.
Do you get the same as with the $20/month onegoogle subscription?
It's not good enough. The place I most want it is in Sheets and it doesn't know how to fill in random cells with stuff. It can populate new tables and stuff with example data which is fine as a start but I usually want it to copilot my formulae.
It's such an obvious use case and Claude 3.5 Sonnet can give me the answer if I paste the header and a value row but Gemini is utterly useless.
You're in-app. How is a textual copy-paste better in Claude? Useless Google PM. The Oracle Java of AI.
We signed an annual pricing deal in fall and now our price is going up materially during the annual term because of this change, because we didn't buy Gemini.
can i get a version of gmail and docs without ai? I had to stop using google keep because they added a flashy AI button that couldn't be removed.
Translation: it wasn't meeting the daily active users engagement targets, so we need to artificially juice them by forcing it on.
It’s not just that I don’t really need the AI features - it’s also that I actively don’t want to participate in adding to Google’s AI training data.
We’ve been happy customers of Workspace for around 16 years - this feels like the straw to break the camel’s back.
Strongly hoping there’ll be enough pushback from nervous corporates about data security that they’ll reconsider.
I clicked the button to turn the forced AI crap off today in Gsuite. It has been on for a day (unasked, forced my google). The engineering team responded with a unanimous thumbs up when I told them. Winner of best response was:
"oh god! now i have to type complete senten..... zzzzzzzz"
Already had to migrate all my domain's off Squarespace and remove Google Workspace entirely when they randomly decided it was now not free and started billing me monthly more than Google Domains did (80% more) Some MBA got a bonus for this surely.
I would buy $GOOG stock blindly but being a paid user of theirs blows
Google Workspace still seems like an amazing deal compared to Slack (for example) which is $15/month.
I also built a chrome extension that integrates AI into Gmail and Docs but users are not restricted to just one AI model.
https://chatgptwriter.ai
Yea, but we dont have to use their LLM. It can live where it likes but will never have a room in my walled garden.
"We've invested all the money in AI that no one wants, how do we make some of it back?" "Why not raise the price for everyone, whether they use it or not?"
Ugh, we got hit by this. We're a non-profit with business plus licenses, so we get a decent discount. I opened a support chat as soon as I got the email about the license changes and price increase. Our contract is due next month, so I need to make sure we're OK budget wise. They can't even tell me how much our renewal will be yet. They don't know.
This morning, I logged on to find that the AI features have been turned on domain-wide for us. I couldn't find any admin controls, so I opened up a support case. The off buttons are locked behind an enterprise subscription. Our end-users need to turn off smart features to disable Gemini. There's no domain-wide / admin level control unless you purchase their most expensive licenses. It's absolutely disgusting. I'm so disappointed with how this was rolled out. We should've been given an opportunity to make an informed and intentional decision about how or if we were going to use these features.
Do we know if there's a way to opt out of the AI stuff in Gmail and docs? I really don't want all this LLM garbage ruining more products.
Business are not willing to pay for Google and Microsoft artificial intelligence so they ramp up the price to everyone and everyone get it 'for free'.
Untill they eventually get hooked on that and then google and Microsoft will once again put that behind a paywall, except now everyone pays more. At least that's the plan.
Now even if employee don't see the benefit of the new deep integrated a.i. and business refuse to pay more for a.i., they aren't going to leave anyway because Microsoft is doing just the same as google.
That's either a win for Google and Microsoft, or at least a neutral outcome.
AI aside, I would pay $20 per month to have Gmail's compose UI work correctly in desktop Safari.
Makes sense that this is the only way to compete with Microsoft.
(See also how MS attacked Slack by including Teams for “free”.)
I still won't enable Gemini
What's a good alternative if you just want company emails for a new startup? Zoho?
We're shutting down workspace accounts this quarter for this exact reason.
Integration will likely be the biggest growth driver for Gemini, I guess.
That gives a clue of where the money is and what people use more :D
Google can’t charge a premium for a noticeably worse AI product.
i dont' want any of this garbage in my workspace account
Email is the killer app, and their trying to kill email.
+1 for hosting my own e-mail server. Haven’t looked back.
So, not free and I’m not allowed to say no.
Free, for only $2 per month!
Plus Google gets to use your data for training. That has interesting implications. What goes in as training data often comes out later as replies to questions.
What if I don’t want AI in either of those places?
Somebody got a promotion for this change
there's so much ai everywhere
so i dont wanna pay for it. especially not google, because.. well, im their product.
If I remember correctly, the Office 365 Copilot thing is more of an upsell rather than an upcharge. Basically, if you didn't want to pay more, you would initiate the cancellation process, and during that you could "downgrade" to the plan that you already had (without copilot as part of it).
Personally, I find that to be especially scummy because it essentially sounds like they are betting on people either not understanding that nuance, or not bothering to deal with it (and subsequently, not using AI, making that venture seem vaguely more profitable)
Pretty simple formula, there simply isn't a market for an $x upcharge on mail and docs, but we have to be part of this latest grift so we'll charge everyone $x - y which is a rounding error. Except it's not. At some point someone is going to admit they have bet the farm on improved auto complete.
Yay now I have to pay more to get bundled features I didn't want in the first place. Yay.
Is it just me or is this title a clear contradiction? Free but increases price? Do they mean now included in your cost…which is increasing? Doesn’t have the magic marketing power I suppose.
Perhaps now is the time for me to switch my personal email off Workspace. I don't use it for Docs or Drive or anything, only email.
Does anyone have experience with Amazon WorkMail or similar, cheaper services for email?
I just saw this. This is probably most useless feature probably nobody asked for except for marketing, to be able to claim they have AI email blah blah.
This is getting tiresome.
[dead]