“…with you as the user always in control.”
What a joke.
Should we (developers) start building websites for robots?
I'm not talking about nefarious motives for doing such either, for our part it is just tedious task where humans spend too much time filling in forms / clicking on UI components, and doing the download manually.
So while letting agents run wild on problems like this can (and surely will) lead to abuse, it will likely also free up so much time for the people doing such tasks for actual work.
Wouldn't it be like half the cost to your organisation if you do it for them? If govt agency xyz doesn't have the resources to build this, offer to make it, get access to the source, plug it into a dead simple API, get your data and everybody's happy?
I've never held a data analysis position so I have no idea if such an org would be open to it. If not, it sounds more like the former issue (gatekeeping and unwillingness) than the latter (inability or resource lack)
They have 50 different offers, all with their own pricing, and they all suck, and Microsoft themselves don't like their results in that field.
Let me solve it for you Microsoft: the money maker you're sitting on today, is "hey excel, make a summary of those 8 sheets to identify our 5 most profitables products and their evolution in sales the past 3 years, add that on a new sheet at the end with a visual graph". None of their product does that right now, instead they tell you the step on how to do it.
Instead of making a bazillion different weird thing, Excel and Office already have their API, just make your "AI" bridge natural talk to excel common task and see every company register it for their employee. I'm not even exagerating, I would in an instant.
I tried many AI tool for excel and none of them come anywhere close to that. It must be much harder than I first thought, but then again they spent BILLIONS on this.
For a company that own business logic as a basic idea, they're really terrible at exploiting it with new ideas. Even just natural talk to power query steps would be worth it.
Sure the api link and permissions and yada yada plays a part, but thats exactly how they can trap us into office365, with already use azure permissions and everything.
Again I'm sure it's harder than it looks, but it's not an AI problem anymore, and they're throwing billions at it.
do you mean users?
In very few use cases it is acceptable to have non-deterministic result for computation tasks. It does not matter if you are a normie or an advanced user.
I would argue that they implicitly do, as any user expects the same action performed on a computer or similar system to provide the same outcome.
> For another, if they give the same task to three different people they're going to get three different results
Give the same three tasks to a single user to be executed three times separately, and he will get supremely annoyed if his actions do not give him the same results.
They would quickly adapt to “reload it a few times if it doesn’t work right away”.
Isn’t this what people are already doing when they browse the web? I know I am.
Aren't they just pattern recognition and regurgitation machines? How would it analyze and interpret unique data with a pattern it's never seen?
Some of the AI stuff is very useful, but it's been massively oversold. In my experience, it's great for working on documentation or creating a one-time-use script to query an API, but it's not good at the "big picture" tasks they want for a sales pitch.
Of course this is a can of worms for a product because we can still not guarantee accuracy.
> How would it analyze and interpret unique data with a pattern it's never seen?
That's called unsupervised learning, and it was field of study in machine learning long before LLMs were anything near viable. Surprisingly, LLMs are good at it too.Which is why MS and every other company keeps pushing half-baked garbage AI products onto the public at every opportunity. So much money has been sunk into AI that they need something to show for it. They also hope that as people use it they can improve the AI they have so that maybe one day it can become something worthwhile. As a bonus it's one more way for them to collect your data and get you used to asking an AI for what you need instead of using your brain.
Thank you for the laughs :D
OT: I wonder if we'll see some "core personalities" that are innate to each mega-LLM (Claude, O3, Copilot, Gemini)... methinks this has comedic potential.
1. https://support.google.com/docs/answer/15820999
Disclaimer: Googler, opinions my own.
A company with an actual rationale names products PlayStation, PlayStation 2, PS3, PS4, PS5
Microsoft called their rival products XBox, XBox 360, XBox One, XBox Series
Asked to put six animals in order, some might figure Anteater, Bear, Elephant, Fox, Goose, Pigeon makes sense - alphabetical order, English names. Others might try to rank them by size, the Elephant is definitely bigger than a Bear, but is the Fox bigger than a Goose? Not sure. You might give them Latin names, there are several reasonable things you might do or at least attempt
But Microsoft are like that's easy: Elephant, Dog, Squirrel, Another Squirrel, Sparrow, Bear, Elephant. And like, that's not even six animals, and it's the wrong animals, and your order makes no sense, what is wrong with you?
Anyway, now it's Office 365 Copilot! What does that mean? It means it's Office, but with an AI which you didn't ask for, which doesn't really do much for you practically, and also which costs 50% more, and you can only opt-out by trying to cancel your subscription entirely.
You can tell AI is a grift because it's all dark patterns and lies with these people.
The things that aren’t shit are harder than they thought.
(Gemini is proving excellent for coding, but it shocks me how poor the integration is for other uses.)
The "AI" integrations I've tried in other tools (Figma, Gamma) are pretty much garbage too.
Anything they can think of. Honestly, anywhere they think that they could put AI to try to convince people that it's a real thing to care about, they put it there. Google is doing the same thing; I get popups any time I try to access any Google thing, like mail, docs, or the Google Cloud console. I don't care, Gemini, and I don't trust you.
Even Apple is jumping in on the AI train, presumably just to avoid getting beaten up in the press for being "behind the times", but thankfully they seem to be trying to start slow and make a good product out of it (which they have yet to do) rather than telling everyone that it can revolutionize everything that anyone ever does and forcing people to hear about it at every opportunity.
They also randomly time out mid-task, e.g. in image generation, without an obvious way to pay for more compute time.
The amount of repetitive data manipulation at companies is literally insane, just ask anyone in accounting, operations, hr, etc… just being able to automate repetitive task in excel for people that can’t code or aren’t excel/Vba super users would be incredibly valuable.
> The amount of repetitive data manipulation at companies is literally insane
Serious question: why isn't coffee being thrown at the low-hanging-fruit takes? Not only would it make them quicker, it would make them more reliable as well.You pretend it’s AI, but in reality you’re just building literally power automate flows which you can call through an LLM lmao
Thanks Microsoft, for giving me yet another reason not to use Edge beyond browser compatibility matrixes in project delivery.
Unfortunally this is also coming to all their developer tools, VS, VSCode, PowerShell, Windows Terminal, .NET Aspire, Azure,... everything is getting AI there somehow, as each team struggles to meet their AI OKRs, and reason of existence among endless layoffs.
Isn't that going to poison the context with low quality, SEO blog spam? How long is it going to be before content mills start filling the web with defensible lies?
For example, "exploding cell phone charger X is built with a high quality transformer." The AI will mindlessly regurgitate that without considering the nuance of source. When I evaluate that stuff the source plays a major role in the weight I give to the statement.
None of the AI I've seen does a proper job of considering the source and I'm guessing it's because they can't attribute sources due to the massive IP theft that was required to build the systems.
My working theory is they see Google's browser revenue and think, "we're the OS, we deserve a bigger cut." It's an incredible level of arrogance that's burning through decades of user trust for short-term gain.
Instead of all this, why not actually compete by being better? They could make Windows a bastion of privacy that protects users from app snooping. They could foster a real native app ecosystem again, like on macOS, so the OS itself adds unique value. That's how you build loyalty, not this constant squeezing of your user base.
Even with the Valve store cut as high as it is, and some of their sketchy terms... they've been far better stewards to gaming than MS has been to Computing/OS.
https://gs.statcounter.com/os-market-share/desktop/worldwide
For the love of $diety, complete a UI transition across the OS, get things in order, stop shoveling things onto people, don't return ads and internet search results from start-menu searches and just make great products. I was an Edge fan until they added all the garbage, starting with coupons, etc.
VS Code has been good, but they gimp their own .Net support... they do great with .Net/C# but turn around and dump their only SQL for ARM option as ARM is taking off... They rebrand RDP despite it being considered the best UI remoting option around. They keep releasing CoPilot everything that doesn't do the one thing that would give their users the most value, and are sub-par with open offerings.
At this point, I cannot support MS-SQL over PostgreSQL for any new projects given a choice. I definitely don't support developing solutions to run on Windows Server, and I'm questioning my use of VS Code despite it being by far my favorite editor since shortly after release.
A link that I found on a different HN thread that is somewhat related: https://www.osnews.com/story/19921/full-text-an-epic-bill-ga...
(Ironically, I was trying to remove an @ key that popped up in an app which reduced the length of my space key by 1/3rd)
Edit: Disabling the copilot options in settings is necessary as they are on by default, disabling does not remove the button. Thanks to this post for reminding me to review settings after an update.. *unhappyface
Edit2: So I just discovered that that you can obtain several types of debug info by triple tapping on the version number in the about screen... That's nifty!
In general, this seems to be less about empowering users and more about shoving AI in their faces. You don't need to prompt anything, it just is there.
Is it linear though? I think most people are using tabs the way everyone use documents. Open one, find the passage you need and keep it there. Lots of current web app makes that hard (lazy loading, memory leaks,...).
A better augmentation would be annotations. Create a new notetaking session which appears in a side bar, then either screenshot or highlight a section, then it is saved alongside the comment (and tags) and the context (link, date,..). No need for a lurking agent.
ADDENDUM
A somewhat simpler version can be found in Orgmode and Emacs. You can store link to almost anything and then the capture feature in Org mode can use them in templates.
I like nice browser syncing between devices. I don't even mind what that means in terms of some slack on privacy... What I don't like are in your face marketing efforts. I absolutely abhor commercials, and have been willing to pay for services that don't have them. I switched to Linux literally the first time I saw an ad in the start menu search results.
I don't need Microsoft products. I'm a computer user first, and a consumer a distant second. MS really needs to learn this lesson. Just saw that US Gov is reporting 6% Linux desktop usage, which is higher than the 5% StatsCounter recently posted. Linux has gained 20-30% in just this past year, even Gaming on Linux while less than perfect/ideal is still relatively smooth for most games thanks to Valve/Steam, they definitely use their 30% cut well and while expensive seem to be at least a good steward of their ecosystem.
I feel like Microsoft has completely lost he plot in a lot of ways. Azure and .Net have had every opportunity to become darlings in the developer community... the tooling and options are pretty good. VS Code has become just about the most popular developer editor (not to mention forks). Why they choose to abuse their common users is beyond me... if they just stuck to the great parts without trying to wrench every penny of value from every user they would be much more well regarded.
As it stands, any company selling commercial software should seriously be working on packaging for Linux... at least appImage/Flatpak. There's especially good opportunity for competing software vendors to Adobe in this space. It's one of the few gaps, and the tide of software for Linux is definitely growing and there's plenty of room for commercial software here.
https://www.bloomberg.com/news/articles/2025-06-24/chatgpt-v...
And HN discussion:
https://news.ycombinator.com/item?id=44367638
Basically, nobody wants to use Copilot, because it sucks.
So Copilot isn't using GPT?
I tried googling it and one thread said they use GPT4. Maybe it's the wrapper they built that doesn't work well.
> Maybe it's the wrapper they built that doesn't work well.
That's the issue. It's almost difficult to imagine you can build a bad product on top of GPT models, but Microsoft showed us how.
The main reason for me not to install this, honestly, is that I've heard that OpenAI is working on something similar and I already have their subscription. I might just wait a bit longer, but I really can see the appeal of this product.
I had to double-check that I was correctly opted into this new feature because it seemed like a terrible implementation of very established technology. Every task I attempted was a frustrating failure. Embarrassingly bad stuff.
If you really want to know my specific gripes: I navigated to a major hotel chain's site that I have discounted rates at, subject to availability. I asked Copilot to find me a hotel within driving distance to my home with those discounted rates available for certain dates. It claimed that it checked that the rates were available (something I didn't expect it to be able to do), and gave me a few hotels. The hotels exist, but every one of the "View hotel details" links were broken. Although it said it checked to see if rooms were available under the discounted rates, it definitely didn't. I tried to get it to book a reservation anyway, and it said that the "page isn't loading properly right now - either it's been moved or there's a glitch on the hotel's site." The hotel's site was working just fine.
I navigated to a recipe website I often use. I asked Copilot to find me some recipes fitting certain basic criteria ("vegan," "quick and easy").
The links it provided to each recipe were all to the main page - not actually to the recipes themselves, even though the labeled link text suggested it would be to the individual recipe.
Although the site has a plethora of vegan options, 2/4 of its recommendations were non-vegan recipes that it gave tips on how to make vegan. Recommending that I make quesadillas by "swapping nutritional yeast instead of dairy cheese" is a terrible, awful idea. Especially in the context of all the other great, already vegan recipes on the site it ignored to make this recommendation.
For the other converted-to-vegan recipe, I manually searched for the original recipe (since it couldn't provide recipe-specific links) to see that the author already had a vegan version of the recipe linked in the original instructions. Copilot's veganization was unnecessary and lower quality than what the author had already provided on the site.
I agree about the nutritional yeast. Tastes great but has only a single texture that can't really stand in for most of cheese types. Apparently there was a recent breakthrough with E coli producing milk proteins that sounds very promising.
Maybe it doesn’t need to change if it's working well? Have you considered that?
At this point most of us have a ChatGPT tab (or 20) open at all times- it's more convenient just having it on the page. It's also genuinely revolutionary when integrated into things like Colab notebooks.
crinkly•17h ago
"CoPilot, please install Firefox for me"
pinewurst•17h ago