> Companies were paying for more than 30 million Copilot subscriptions as of the end of June. And while individual users are free to use limited versions of Copilot’s chat experience, its most powerful tools are reserved for subscribers to the company’s M365 apps bundle, which has about 90 million paying users.
It needs to be more widely known that if you cancel a (home) 365 subscription, you'll be offered a cheaper version without AI integration.
Now if I could just get them to start paying me cash for training their AI on my open source code and selling it rather than being force fed a Copilot grant.
Microsoft allows you to subscribe to Microsoft 365 personal or family without AI for a small discount, but they hid the option on a page you can only get to if you click cancel on your subscription.
Iirc Microsoft increased the pricing on existing subscriptions on next renewal citing AI as the reason for the increase, they then added a no AI version at the original price, effectively opting all users into the 'upgrade' rather than requiring them to explicitly upgrade their licence.
Some of us have internalized it! It's like walking around in the twilight zone or the bodysnatchers movie sometimes. In a dream society we could agree on some values and argue about how to direct our shared efforts. In this one we can't agree on values or even facts. It's equal parts maddening and terrifying.
We need help in avoiding the worst futures that could come from this.
It’s a documented fact that a lot of subscription services will offer you a discount if you ask to cancel. This has been a thing in the US with cable tv for about a decade, as far as I’m aware. I think some cell providers might do the same. Not everything, but some.
My blood pressure rises everytime i use Enterprise Copilot at work. It seems they truncate message history heavily (to save input token costs?) and Copilot forgets the last thing you told it in the chat. It is like a real-life version of Momento (https://www.imdb.com/title/tt0209144/) mixed with Waiting for Gadot (https://en.wikipedia.org/wiki/Waiting_for_Godot), pretending to be in league with Anthropic.
When people ask me why I don't have any Windows machine at home, I say "for health reasons". It really raises my blood pressure and at some point I realised it's not worth it risking my health over it. You could use the same to get away from Copilot.
I only have one Windows install left and I only have it to (1) play some games with Windows-only anti-cheat and (2) play my flight simulator, where the hardware is finnicky to get working on Linux. And honestly, (2) seems to be becoming less of a hassle every month, especially now you can throw agents at the driver issues.
Doing research work with it that outputs any kind of artifact is.. really painful. It constantly crashes, has to regenerate.. and then the output is often very difficult to carry to other Microsoft products, like Sharepoint. I end up creating PDFs that I import.
I'll go ahead and burn working hours on it because I'm getting paid, and people like the things I create.. but it's a very crippled and painful version of what you get with the frontier products and most OSS options.
It boggles my mind that Microsoft can fail to deliver even bare minimum working software so consistently, with all the engineering talent they house. What are the product teams doing?
I feel it is even worse than bare minimum -- it is so bad it infuriates the user. It is hard to get away from that reaction when one uses better alternatives in their personal life.
If you prompted an AI to return answers "as though you habitually ate soap and crayons while mouth breathing" you'd still get a better response than copilot
That used to be the case, but the current Enterprise version, with the poorly-named "GPT 5.6 think deeper" option, is very decent indeed. Can't tell if it's Sol, Terra or a mixture of the two, but it's certainly not Luna.
(Stick with the default "Auto" and you probably get Phi-4 -- it's lobotomised AF.)
I use Copilot a lot now. Unlimited, free, high-quality, enterprise-protected inference is not to be sniffed at. It also writes some of the nicest code I've ever seen from an LLM, that's almost entirely free of comments.
Are you referring to GitHub Copilot? I believe this thread is referring to Windows Copilot, such as the AI feature embedded into every Office progam, which is not the same experience.
Fair point, that happens when i'm trying to be on MS Teams for a conf call, while watching my claude code terminal and then browsing HN on another monitor :-/
Maybe one problem is the time between prompts, much like going to get a coffee when you start a big build, now it's filling time between prompts (plus the build), which can be any number of distracting rabbit holes (the internet is a big place) the split context feels exhausting to me.
You're right using Microsoft products is like mixing Momento with Waiting for Gadot :)
I recently switched to New Outlook and it's full of weird subtle bugs. It takes them years to fix things and make them barely useable. Given their track record I'm not optimistic for Copilot given both the vision and the execution are questionable at best.
It's such a broken POS. Code snippets, especially for html, frequently don't render, or if they do, they render as rendered html which will frequently be an empty box, and you have to manually (for each snippet box) switch to show code mode.
I can tolerate Teams being a bit mid because the core functionality nearly always works. I can tolerate Outlook being meh because it's doing a lot (nothing I, a lowly IC really need though I can imagine the top hatted boss who does) and I think it does okay with the scope it's tackling.
But the M365 copilot app is inexcusable for being as core to Microslop's strategy.
Huh... not my experience with the "premium" license. The interface isn't as nice and definitely less powerful than something like codex/claude code/whatever, but I don't seem to have any issues with truncation or things like that. I'm usually using 5.6 thinking (not clear which 5.6 model exactly) when I'm using copilot at work. It seems... fine? I always see people here always complaining that it's nerfed, truncates things, etc etc. The only difference I see is that interface isn't as nice nor as powerful, but I don't notice the answers being noticeably worse or anything like that. Occasionally it messes up code block formatting, but that's the only thing I can think of that annoys me.
Now, the "copilots" integrated in certain Microsoft products can be really horrible. I assume those are some sort of custom models that are not very capable, most of them I consider effectively worthless. But the web chat interface where I can pick chatGPT 5.6 seems fine...
Given you're comparing to codex/claude code you probably arent talking about MS Copilot (e.g. the thing inside MS Teams but also available standalone as MS Copilot 365). Try it out and be convinced :-)
I don't think there is a misunderstanding. I don't recall ever having an issue with it either. Now I don't do anything extreme with it. Just questions, suggestions, or some basic task which are either Microsoft or basic business activity related. I haven't tried anything like coding, deep analysis, or anything complicated. It feels like it is for business tasks so I haven't really tried to push it to do anything outside of that.
I think I am! I use it in the browser instead of teams/the desktop app, but my understanding is that it should be the same thing. I do know most people don’t have the premium license for it at my company, not 100% sure what difference that makes but they say it’s better somehow and has access to better models, or at least did at one point.
But yeah, comparing to codex/claude code was a bit silly since those can do way more. It’s closer to the chatGPT. I’m pretty sure 5.6 Luna, not sure exactly since it doesn’t specify Luna/Terra/Sol
One difference is that the premium copilot has access to company files and those on your onedrive, whereas the non premium version requires you to upload files.
Does anyone even want to try this? I'm a Windows user, M365 subscriber, VS Code user (when I'm not in Codex), had a Github Copilot subscription, etc... and a heavy AI user. Basically, the prime target for Microsoft AI. And every attempt Microsoft has made to put AI into a product has been unusable garbage, using the same models I find productive in other tools.
So many AI doubters I know hate using AI tools because their employer only gives them the Microsoft offerings. Which is a totally understandable reaction to those tools.
I believe it’s checkbox-filling in the acquisition process for enterprise contracts, like a lot of other half-baked capabilities in their flagship products.
It doesn’t matter how they work, it matters you can tell leadership “yes, they have that too” to maintain existing contracts and vendor relationships.
It was foolish of them to think they had any consumer clout left. Shoving 1000 copies of an inaccurate, poorly functioning, and inconsistent product down everyone's throat will someday be a case study in how to immediately destroy a brand.
Crazy how if they had waited about 2-3 years, and kept windows AI-free and released it today using the latest models, it would absolutely be well received (see: muse, new siri)
Instead their incompetent managers were chasing promotions and pushed a technology which was not ready yet
These kinds of features absolutely come from the top. VP level, at the very least, if that's what you mean by incompetent managers. The middle managers are just trying to do their jobs and figure out how to get it done. If they say no, they're just going to lose their job faster than they'd lose it to AI.
No, the very top. Nadella has been driving this directly, he was going to get there first, shock & awe, etc. Maybe this would have worked if Microsoft cared about customer experience as much as it cares about customer lock-in.
>Crazy how if they had waited about 2-3 years, and kept windows AI-free and released it today using the latest models, it would absolutely be well received (see: muse, new siri)
But that would make John Shareholder sad! He wants it done now, this quarter!
I guess most users of Microsoft products are not the buyers of said products.
When these tools are getting forced upon you in the workplace, there's often not so much you can do. I can raise the quality issues in the 1-to-1 with my manager, but it won't have any effect.
"at all" is probably a stretch, but agreed that the number one factor for a mile is cost. users will accept a surprising level of trash, if the price is cheap.
engineers tend to think as engineers do, solving all problems with features, quality improvements, etc. but those all drive costs up, and don't always provide the return we think they will.
Microsoft inviting IT to nice events and giving them perks is the reason they keep a grip so tight on enterprise. For unknown reasons, every garbage new microsoft add-on will be enabled silently without any clear way to manage. At least for copilot pro plus premium they sent an email that "i was invited" so i could promptly reject before they activated, but here is the trick, 99% of the employees won't bother to opt-out.
Just going to use this thread to also add a nit: I hate that Microsoft made all the OEMs add a Copilot key to all their laptop keyboards. It's been Ctrl forever, and suddently they throw this net-negative key onto the keyboard. I have decades of habit of using the right side Ctrl key for varying shortcut strokes, and then when I use it now it loads up an unproductive Copilot experience that is awful. <end rant>
Not at all that surprising. There enterprise offering is barely acceptable and they have continued to drop the ball on having half decent Cowork solution.
They have all rights to win with their existing enterprise agreement but wow it’s amazing to see how bad their product cadence and decision making is. All they needed to do was copy Claude Cowork like OpenAI did.
Yeah the difference with Microsoft though is they actually hold the enterprise data and have been entrusted with doing so. So if Microsoft’s AI bypasses the governance guards a company puts in place on data that lives in O365, they are liable. With Cowork and OpenAI the liability is on the employee who is misusing the data because the company did not go into any governance agreement with Anthropic or OpenAI.
Nah it makes sense. They know their AI is second rate at best. Cortana implies 26th century excellence. If the reality is a step back (and we all know it’s several), then “haha Cortana dumb” memes would sweep the internet. They would have to rebrand anyway. If anything it was smart to rip off the band aid early.
The low quality of the voice assistant meant they doomed the Cortana name. It got associated with garbage voice assistance, so they had to rebrand.
If there had been consumer good will for voice assistant performance before poisoning the name, it might have survived... but when you first product in that realm is garbage and it lingers for a long time, you're dooming the name.
Also, synergy with Halo is meaningless to Microsoft. Most folks do not care or even know about Halo.
It kind of has, although I would argue that Siri is more successful than Cortana.
If I recall correctly, the thing that made Cortana particularly frustrating and worthless is that it did all searches using Bing, and opened any websites using Edge, regardless of your default settings. You got forced to use all the Microsoft version of every tool, even if you didn't want to.
It has. Literally nobody outside of tech bloggers like Siri AI or want Siri AI. I'm pretty positive on AI in general and having used Siri AI now since iOS 27 I can say it's fucking useless and I'm about to delete it from my devices to free up the storage.
Because Apple doesn't care. Siri is just something that is bolted on when you buy the hardware. Cortana was an actual product that Microsoft is trying to sell.
I don't know anybody who uses Siri for more than just setting an occasional timer or something. Usually it's a nuisance when she is accidentally activated and takes over the interface.
Considering how little knowledge you showed of Microsoft’s history in a previous comment, it would be better if you just sat out branding conversations.
I think it makes much more sense for Microsoft to provide a good agent orchestration layer for their operating system and productivity software (are we still calling it that?) than try to compete with the Frontier providers.
I agree, “Think Deeper” seems good for chatbot-style use. However, “Quick Response” is awful, even for simple, one-step queries and very often makes things up or contradicts its earlier replies.
This has happened to me, but only a couple of times. I think it’s an issue with the UI, not with the actual LLM output, and a reload of the page fixes it.
I think it’s related to how they display the LLM output, through some obfuscation layer in the HTML presented to the browser. I tried to save a thread as pdf once and that got me nowhere because of this.
Copilot was fantastic, it's usage was insane because it was by request not tokens for the longest time. You could make a long markdown file with lots of planned edits, and just let it rip.
It also had some really good models for basically free while it was in preview instead of being tied to a single provider.
And finally, it's vscode integration was really good. I use codex now, and changes to a thread are in this awful diff view that shows all changes in a single tab. And even that wasn't working for a while there. No clue if cursor is better never tried it because it takes a lot for me to swtich tools because of all the small things like shorcuts and configurations that I've set over a long time.
It _was_ fantastic. Then it transitioned to abysmal over night when they switched to usage based billing. Now, if you put the same 50 USD towards GitHub copilot compared to Codex, I’m pretty sure Codex gives you 20x the usage or something. Really sucks for us who are stuck with GitHub copilot at work for compliance reasons.
someone else mentioned how their blood pressure spikes from using microsoft products...
The sheer anger that I feel over how fuckingly stupid it is that they named everything copilot that basic conversation is difficult... I tried going to the article and got swarmed by paywall, finally checked the archive link and I think you're right.
This is crazy to me. I work in local government and for a state in Australia and most LG orgs are battling with the confusing of the current Enterprise Copilot conveluted-ness. This is just another change to deal with. What happens with Agents, Studio, Cowork and Scout? Is this a rebrand again of those?
This is how it looks to me now: we have "The New Copilot”, which is Home, Code and Autopilot. Meanwhile an organisation is already trying to explain usage of Microsoft 365 Copilot, Copilot Chat, Copilot in Teams, Word, Excel, Outlook etc, Copilot Studio, Agents, Agent Builder, Researcher / Analyst and other agents, Cowork, Copilot Notebooks, Copilot Actions, Workflows and connectors.
I know frontier AI labs change a lot, but this just seems like throwing a new mask on the product to cover the mess underneath.
microsoft Product remains an unmitigated dumpster fire.
i think github copilot has actually gotten to a pretty good place, and of course the two copilot products will both have features called "autopilot" that do entirely different things
Microsoft will forever be in my personal dumpsterfire. Absolute trash. How the fuck they can allow admins to disable copy and expect anyone to get anything done. I work in healthcare I’m expected to understand what is confidential and what is not as part of my job. 99% of my email has nothing confidential in it but yet I cannot copy an email address or drag a block of text anywhere. It is so infuriating…. That guess what…. Everyone sends their emails to their personal address. Good work…. I realise this is the sysadmins washing their hands of accountability. But Microsoft should allow copying an email address from their own profile pages. I do other admin that requires lists of people emails and I have to resort to all sorts of crap to get their email address out of Microsoft. It’s a joke. They can burn in hell.
It’s really interesting to watch unfold because from the outside it seems so obvious. While I know change at scale can be hard it’s surprising nobody has called for a culling or shakeup to work it out.
> Jacob Andreou, who leads product development of the unified Copilot app, told the audience of corporate customers on Wednesday. Importing some user experience features from the consumer app “made Copilot so much better in just the last few months,” he said.
I was about to say that it feels so much earlier than that. But then, 2016 was a whole decade ago.
To think, the main impetus was mostly to see if it could be done. And to show Microsoft why you don't expose things like that to the public. Microsoft said "She'll learn from your conversations" and the internet said "Bet"
It needs to be more widely known that if you cancel a (home) 365 subscription, you'll be offered a cheaper version without AI integration.
It’s really surprising to me that after the US public hasn’t internalized how corrupt the country has become. It’s all coming from the top
We need help in avoiding the worst futures that could come from this.
I'll go ahead and burn working hours on it because I'm getting paid, and people like the things I create.. but it's a very crippled and painful version of what you get with the frontier products and most OSS options.
If you prompted an AI to return answers "as though you habitually ate soap and crayons while mouth breathing" you'd still get a better response than copilot
(Stick with the default "Auto" and you probably get Phi-4 -- it's lobotomised AF.)
I use Copilot a lot now. Unlimited, free, high-quality, enterprise-protected inference is not to be sniffed at. It also writes some of the nicest code I've ever seen from an LLM, that's almost entirely free of comments.
I don't think many people are aware of just how much it can do; it's certainly not advertised, and very likely not intended.
My experience was great because I had premium license and only used thinking models
Most colleagues didn’t have premium and used default/auto model
People stick to their very first bad experience unfortunately
I recently switched to New Outlook and it's full of weird subtle bugs. It takes them years to fix things and make them barely useable. Given their track record I'm not optimistic for Copilot given both the vision and the execution are questionable at best.
I had to edit Windows Registry to avoid getting prompted every time I open old outlook. I didn’t like the looks or the bugs in the new version.
Trillion dollar industry btw.
But the M365 copilot app is inexcusable for being as core to Microslop's strategy.
Now, the "copilots" integrated in certain Microsoft products can be really horrible. I assume those are some sort of custom models that are not very capable, most of them I consider effectively worthless. But the web chat interface where I can pick chatGPT 5.6 seems fine...
But yeah, comparing to codex/claude code was a bit silly since those can do way more. It’s closer to the chatGPT. I’m pretty sure 5.6 Luna, not sure exactly since it doesn’t specify Luna/Terra/Sol
The interface is shockingly terrible, but with electron and a few hours with devtools, you can improve on it substantially.
I hope this new release does not change too much.
So many AI doubters I know hate using AI tools because their employer only gives them the Microsoft offerings. Which is a totally understandable reaction to those tools.
It doesn’t matter how they work, it matters you can tell leadership “yes, they have that too” to maintain existing contracts and vendor relationships.
By the numbers we just call them “most people”… you’re in the minority by quite a growing stretch.
(Not the be confused with the Claude Opus 5. That's a different lord that didn't turn out to be a saviour.)
https://www.pewresearch.org/short-reads/2026/08/18/young-adu...
The numbers and trend lines are clear regardless of your layman relative.
Instead their incompetent managers were chasing promotions and pushed a technology which was not ready yet
But that would make John Shareholder sad! He wants it done now, this quarter!
The reality is that most users and consumers don’t care about quality, accuracy or functionality at all.
When these tools are getting forced upon you in the workplace, there's often not so much you can do. I can raise the quality issues in the 1-to-1 with my manager, but it won't have any effect.
engineers tend to think as engineers do, solving all problems with features, quality improvements, etc. but those all drive costs up, and don't always provide the return we think they will.
Powertoys even offers keyboard remapping.
That's in relation to an estimated 450 million or so M365 subscriptions for businesses.
Not exactly a huge success.
They have all rights to win with their existing enterprise agreement but wow it’s amazing to see how bad their product cadence and decision making is. All they needed to do was copy Claude Cowork like OpenAI did.
The low quality of the voice assistant was unfortunate, but the naming was perfect.
If there had been consumer good will for voice assistant performance before poisoning the name, it might have survived... but when you first product in that realm is garbage and it lingers for a long time, you're dooming the name.
Also, synergy with Halo is meaningless to Microsoft. Most folks do not care or even know about Halo.
If I recall correctly, the thing that made Cortana particularly frustrating and worthless is that it did all searches using Bing, and opened any websites using Edge, regardless of your default settings. You got forced to use all the Microsoft version of every tool, even if you didn't want to.
Man effing hell Apple
I don't know anybody who uses Siri for more than just setting an occasional timer or something. Usually it's a nuisance when she is accidentally activated and takes over the interface.
And to Apple, Siri is to Siri AI as MacBook is to MacBook Pro.
Completely different product.
Asking for Siri is like an automatic reflex to a lot of people. You can’t change that name.
Select "GPT 5.6 Think deeper" (ugh, so unrevealing, MS) and you've got free, unlimited Sol or Terra (likely a mix depending on load) on tap.
It writes very nice, clean code with almost no comments.
I use it a lot.
I think it’s related to how they display the LLM output, through some obfuscation layer in the HTML presented to the browser. I tried to save a thread as pdf once and that got me nowhere because of this.
It also had some really good models for basically free while it was in preview instead of being tied to a single provider.
And finally, it's vscode integration was really good. I use codex now, and changes to a thread are in this awful diff view that shows all changes in a single tab. And even that wasn't working for a while there. No clue if cursor is better never tried it because it takes a lot for me to swtich tools because of all the small things like shorcuts and configurations that I've set over a long time.
The sheer anger that I feel over how fuckingly stupid it is that they named everything copilot that basic conversation is difficult... I tried going to the article and got swarmed by paywall, finally checked the archive link and I think you're right.
As someone who works with Microsoft intune autopilot regularly... ugh.
This is how it looks to me now: we have "The New Copilot”, which is Home, Code and Autopilot. Meanwhile an organisation is already trying to explain usage of Microsoft 365 Copilot, Copilot Chat, Copilot in Teams, Word, Excel, Outlook etc, Copilot Studio, Agents, Agent Builder, Researcher / Analyst and other agents, Cowork, Copilot Notebooks, Copilot Actions, Workflows and connectors.
I know frontier AI labs change a lot, but this just seems like throwing a new mask on the product to cover the mess underneath.
i think github copilot has actually gotten to a pretty good place, and of course the two copilot products will both have features called "autopilot" that do entirely different things
What fresh hell is that?
Mindboggling.
Yeah, it’s shit lol.
https://www.cbsnews.com/news/microsoft-shuts-down-ai-chatbot...
https://www.nytimes.com/2023/02/16/technology/bing-chatbot-m...
To think, the main impetus was mostly to see if it could be done. And to show Microsoft why you don't expose things like that to the public. Microsoft said "She'll learn from your conversations" and the internet said "Bet"
...well, it did exactly that, no bet here.
It's acknowledgement
What do you mean by that? In what way?