Well, the major problem Microsoft is facing is that its AI products are not only shoddier than average, which is nothing new for them in many categories, but that this time the competition can actually easily leapfrog them.
Like, I have a 'Copilot' button prominently displayed in my New Outlook on MacOS (the only platform where the app-with-that-designation is sort-of usable), and it's a dropdown menu, and it has... zero items when expanded.
I asked my 'Microsoft 365 Bing Chat AI Bot Powered By ChatGPT<tm>' about that, and it wasn't able to tell me how to make that button actually do something, ending the conversation with "yeah, that's sort-of a tease, isn't it?"...
Oh, well, and I actually also have a dedicated Copilot button on my new Lenovo laptop powered-by-Windows-11. And, guess what, it does exactly nothing! I can elect to either assign this button to 'Search', which opens a WebView2 to bing.com (ehhm, yeah, sure, thanks!) or to 'Custom', in which case it informs me that 'nothing' meets the hardware requirements to actually enable that.
So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling? Because if you would have, the failings would have been obvious, right? Right??
The other day I've clicked on one of Outlook calendar's copilot prefilled questions: "who are the main attendees of this meeting". It started a long winding speech that went nowhere, so I typed in "but WHO are the attendees" and finally it admitted "I don't know, I can't see that".
Absolutely! There are so many scenarios where they could actually add some value, and they're fulfilling, like, exactly none of those?
Even in Visual Studio Enterprise, their flagship developer product, the GPT integration mostly just destroys code regardless of model output. I truly cannot fathom how any of that made it past even a cursory review. Or how that situation would last for over 6 months, but, yet, here we are.
And, again, it's fine with me: I'll just use Claude Code, but if I were a Microsoft VP-or-above, the lack of execution would sort-of, well concern me? But maybe I'm just focused on the wrong things. I mean, Cloudflare brought down, like, half the Internet twice in the past two weeks, and they're still a tech darling, so possibly incompetence is the new hotness now?
Imagine a circumstance where Windows Search was as good as Apple's Spotlight, and could integrate with cloud services to index documents, browser bookmarks, web history, maybe podcasts, etc.)
Hey Copilot, where is that document I was reading about the new network diagramming software Jacob is testing out?
Or Hey Copilot, my disk is getting pretty full. What software is taking up a lot of space that I haven't used for a while? Or are there any files I can move to cloud storage to free up space?
But no, instead it's just 'we're going to take screenshots of all your windows, OCR it, and index it, so that when someone infects your machine they can see your credit card numbers and pornography habits.'
It's so broken. I have a suspicion the AI teams are in some kind of an internal standoff with privacy/compliance teams, and Copilot simply never gets access to any tools or data.
Because if not that, then I don't know what. I can half-ass a better product with ChatGPT API and a PowerShell script, and I could've since GPT-4 was released; in fact the product can actually write itself, it's that simple to do a better job.
Being Microsoft, I'd guess it's less that and more Product Team A doesn't want to spend time building enabling features for Product Team B to look good.
I have Copilot at work, it feels so useless sometimes. As an example, I had a report which I needed to make some batch edits to. I figured why not let the robot take a crack at it, so I clicked the Copilot button and spent a couple minutes describing what I needed changed.
Copilot tells me it can't edit my current document, but it can create a new one. I figured okay, Microsoft doesn't want to set it loose on the original, guess it makes sense that it requires a copy. So I said yes.
Nope. Instead of creating a copy of my document and editing it, it created an entirely new document which excised basically everything in the original report and replaced it with a very short summary - I'm talking 5000 words down to 500. All my tables and figures were gone, as was the standard report template my employer uses.
What utter garbage. Office productivity is a major use case for LLMs, and here the largest vendor of productivity software on the planet is happy to fuck it up.
The chatbot doesn't see the document structure; it sees tokens. When it generates a response, it's not editing; it's rewriting from scratch. Without a mechanism capable of precise diff/patch operations (like Cursor attempts with code), an LLM will always be a destructive tool for complex documents. Microsoft marketed this as an editor, but in reality, it's a summary generator
Compare this with Cursor or Supermaven. There AI acts as an extension of the cursor, understanding the code's AST. VS Copilot often feels like smart copy-paste that doesn't understand where it is
Microsoft, owning VS Code and LSP, should have done this better than anyone, but they lost to startups on UX
Since DECADES Outlook is _not_ capable of putting the first(!) just(!) useful datetime-stamp found in an email which says "calendar" or something in the text when pressing CTRL+R:
This function creates a calendar entry from the current email with current receiver list, but default-date/time is alwsys set to the current even if recipients last ansewer was simply "lets do: 11am" - why does even this simple shit not work today?
You dont need AI for this simple idea even, a robus / probability parser will do fine in 99% cases
I would just be happy to never have a ICS file every saved to my drive again. Whoever created the concept of a calendar invite saving a ICS file has created me hours of frustration.
It's so easy to ship completely broken AI features because you can't really unit test them and unit tests have been the main standard for whether code is working for a long time now.
The most successful AI companies (OpenAI, Anthropic, Cursor) are all dogfooding their products as far as I can tell, and I don't really see any other reliable way to make sure the AI feature you ship actually works.
Tests are called "evals" (evaluations) in the AI product development world. Basically you let humans review LLM output or feed it to another LLM with instructions how to evaluate it.
Interesting, never really thought of it outside of this comment chain but I'm guessing approaches like this hurt the typical automated testing devs would do but seeing how this is MSFT (who already stopped having dedicated testing roles for a good while now, rip SDET roles) I can only imagine the quality culture is even worse for "AI" teams.
Traditional Microsoft devs are used to deterministic tests: assert result == expected, whereas AI requires probabilistic evals and quality monitoring in prod. I think Microsoft simply lacks the LLM Ops culture right now to build a quality evaluation pipeline before release; they are testing everything on users
50% of our code is being written by AI! Or at least, autocompleted by AI. And then our developers have to fix 50% of THAT code so that it does what they actually wanted to do in the first place. But boy, it sure produces a lot of words!
I've fooled around with some vibe coding on several LLM's like Claude, Gemini, and ChatGPT with some pretty decent results.
Since I have a full Copilot license at my corporate day gig, I figured I would try using Copilot for a basic static site. Nothing too hard, and something that's been handled easily with the other LLM's.
The prompt was pretty basic just to get something to start working with. "Build a four page template. With a home or index page, two pages of content and a contact page with a responsive slide out menu from the left hand side of the page."
It ran and put everything in a folder. I open the home page and everything was broken. I opened the files in VS Code and saw this:
I mean, if you can't even this right, I don't have much hope it can do anything more complicated. To say this was pretty sad is an understatement and clarified how far Microsoft is behind other LLM's.
Kudos to you. I never use new buttons out of fear of something irreversible happening, like sending a random email or deleting something. I still feel uncomfortable with the Gmail UX, I would _never_ use a ”hello iz magic ai”-button.
Me: Please find any items in my inbox or sent items indicating (a) that I have agreed to take on a task or (b) identifying me as the person responsible for a task, removing duplicates and any items that I have unambiguously replied to via email or Teams. Time window is preceding 7 days.
Copilot: Prints a list with, at best, 5% accuracy
I know some folks have the peculiar idea that search is dead in favor of AI, but if AI can't accurately find information, it is useless. As near as I can tell, Copilot finds 3-4 items (but rarely the SAME 3-4 items across runs) and calls it a day. It just seems like nobody is actually testing any of this stuff. Microsoft is actively destroying its credibility because it's offering a tool with a party trick but is utterly unreliable. I will, therefore, not rely on it.
It's a generalization problem. We can train LLMs that 'know' a lot of stuff in the global sense but the tasks that are interesting to people require the LLM to know a lot about you and your world in a very specific sense. The technical problem is that it's all corner cases and that's impossible to scale right now. No amount of context window is going to get you there either.
I respectfully disagree. The scenario posted by the GP is very clearly an "easy" task of basically sentiment analysis. The only thing it requires is a way to query your message history, then it just goes through them 1-by-1. I would expect any of the current SOTA agentic models to handle this with ease.
And that’s ok. Those are core features of the phone that absolutely must work reliably and consistently. Far better to do a few important things really well than a hundred things execrably.
However the other day I asked the Gemini assistant on my phone to check the birthdays in my calendar, get all their dates, then make a graph of how many fall in each period with a 15-day moving average. It did everything as instructed including writing a python script to generate the graph, then discussed the results with me :)
I would expect Siri to be able to do anything on the iPhone that I can - change settings, report stats, kill/launch apps, etc.
It would be nice if it could control 3rd party apps too, like GMail, but being able to control the stuff that Apple themselves have built doesn't seem a lot to ask.
Apple is somewhat privacy focused, allegedly. I would imagine that letting an unpredictable LLM read whatever system data it wants is not something they want to partake in.
Maybe there are different concerns now that Apple are apparently going to be using Google LLMs for Siri, although I'd have thought that any access to system settings or other iOS functionality would be via tool calls with user opt-in.
However, the original Siri, obtained from SRI ("Siri" = SRI) was pre-LLMs, and there should have been no more concern over accessing system settings or controlling apps than things like reading/sending messages. For some reason Apply completely dropped the ball with Siri; initially it was expected that they would expand it into revenue generating areas like restaurant reservations etc, but then nothing. I'm not sure what documentation exists, but even today it's not clear what Siri is actually capable of.
This is what move fast and break things looks like in a enterprise the size of microsoft.
It's mostly break things and little moving fast.
But the idea is that it's AI or death, so some broken buttons seems of less importances than the buttons itself being there, because the button working is a problem involving several teams, so no one is actually responsible, but the button being there is some team problem, and hell yeah they solved in the first sprint.
"Move fast and break things" is fine if you're a social networking site and breaking things means people can't get their racist memes or browse marketplace for twenty minutes until you push a change.
It's less fine if the things you're breaking are your core operating systems and the office suite that makes you most of your money and it takes you months to get the relevant teams aligned to push out a fix for the bad idea your execs pushed.
This is a problem at a wide array of tech companies. Everyone wants to Meta, everyone wants to be Google.
Guys, we're building B2B enterprise software. The most important things our clients care about is this hunk of junk working. Changing it is probably bad, actually, because the users are using it 8 hours a day and they don't want to deal with annoying popups about new features and UI churn for the sake of churn.
It should be noted that the things you are supposed to be allowed to break are YOUR things. When you start breaking MY things then we're going to have a real problem.
> Oh, well, and I actually also have a dedicated Copilot button on my new Lenovo laptop powered-by-Windows-11. And, guess what, it does exactly nothing! I can elect to either assign this button to 'Search', which opens a WebView2 to bing.com (ehhm, yeah, sure, thanks!) or to 'Custom', in which case it informs me that 'nothing' meets the hardware requirements to actually enable that.
How did you manage this? Probably some company-wide group policy saves you. It keeps starting copilot for me, drives me crazy.
I did absolutely nothing special, other than running the latest-and-greatest Windows 11 Enterprise, which is what we put on most of our laptops without any customizations other than "require 2FA and some antivirus and firewalling" via Intune.
And I just went into our Azure admin portal, looking for any AI goodies to enable, and... there just doesn't seem to be anything there? And we have an Enterprise P2 subscription, which is usually where all the good stuff is, but, yeah...
Well, I happened to think many companies woild configured copilot to be disabled: claims to prevent corporate secrets from leaking out or refusal to install internal closed-net Microsoft CoPilot server.
> So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling?
Satya Nadella insists that Bing365Pilot has supercharged his productivity, but determining if he's high on his own supply or lying through his teeth is an exercise for the reader.
> Copilot consumes Nadella’s life outside the office as well. He likes podcasts, but instead of listening to them, he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond. At the office, he relies on Copilot to deliver summaries of messages he receives in Outlook and Teams and toggles among at least 10 custom agents from Copilot Studio. He views them as his AI chiefs of staff, delegating meeting prep, research and other tasks to the bots. “I’m an email typist,” Nadella jokes of his job, noting that Copilot is thankfully very good at triaging his messages.
This is just him aping every other AI CEO. Every single one has to act like the agents are super-geniuses mere moments away from achieving the singularity like people can't try them out themselves and be disappointed. Some of it is "we think this will work soon so it's ok if we pretend like it's working now", but I think a lot is just needing to constantly shove hot air into the balloon before it pops.
> he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond.
What a dorky thing to do. Does the CEO have some concept he's living a life that precisely _zero_ of his customers do? Who would even think to do this?
> “I’m an email typist,” Nadella jokes of his job
Yea, I have actual work to do, perhaps you should familiarize yourself with this?
> He likes podcasts, but instead of listening to them, he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond.
I remember reading that when it first came out and all I can think is: No, he doesn't like podcasts, if you like podcasts you listen to them.
That's like saying "He loves food, but instead of eating it he feeds it to an analyzer that tells him what elements were detected in it".
I have to assume it's all BS/lies because if that's a truthful statement (about podcasts and the other things) then I really question wtf they are doing over there. None of that sounds like "the future", it sounds like hell. I cannot imagine how shitty it would be to have all my emails/messages to the CEO being filtered through an AI and getting AI slop back in return.
The podcast thing reads like something he made up that sounds cool and futuristic on the surface, but doesn't actually make any sense. Maybe what's actually happening is he's just having the LLM give him the cliff's notes of the transcript, but that isn't interesting enough so he's making up some BS about having a conversation with the AI.
Nadella has to have his own custom agents. It isn't even possible for an enterprise like MSFT to not have custom agents that are still remotely useful.
So, his experience with Copilot agents != Average Customer's experience
Look, if the CEO of a five trillion dollar company investing hundreds of billions into AI can come up with some custom agents to handle every aspect of the CEO's work, surely your average Microsoft 365-subscribing corporation can do the same.
It's just another example of the rich being wildly out of touch. Yes, Beyonce has the same 24 hours in a day that the rest of us have, but she also has enough money to pay people to do every aspect of her life that isn't bringing her joy or wealth. Yes, AI can be used to streamline workflows or help you find signal in the noise so you can focus on the important things better, but if every company has to build that themselves then no company is going to see the value of spending a bunch of extra money on something that they can only get benefit from if they spend even more money.
If the 'AI agents' that Nadella is talking about were part of Copilot then sure, okay, I could see a benefit, but when people in this thread are saying that Outlook can't even tell you who is in a meeting then it certainly explains why Nadella doesn't understand the lack of value.
The only good feature about copilot in Microsoft 365 is that if you ask it to delete itself enough times eventually some other app will eventually be promoted to the most prominent when you first login.
Competition could as well in the past, I only switched from Borland compilers into Microsoft ones, after Borland decided to focus on enterprise customers.
And we all know the OEM deals preventing them to ship OS/2, BeOS, or whatever else they would like to.
The problem is that in modern computing landscape they cannot play that strategy, even when owning plenty of FOSS projects, there are plenty alternatives.
There's the Copilot cli, then we have the Copilot built in to VSCode, then we have the Copilot that's in Github, then we have th M365 Copilot and the Windows Copilot and maybe a few more.
All work very differently and have different skillsets.
> Like, I have a 'Copilot' button prominently displayed in my New Outlook on MacOS (the only platform where the app-with-that-designation is sort-of usable), and it's a dropdown menu, and it has... zero items when expanded.
I guess that's worse than the Gemini button in Google Sheets that asks me to subscribe to AI services. I have multiple times been in a sheet and thought "asking an LLM how to do this thing I want to do right here in this product would actually be great if it works", remembered there was an AI-looking button in the top right, clicked it, and nope'd out of the subscription.
I just want to know if it works or not before I buy it.
Honestly, I can't tell if you're being sarcastic and it actually does disable the button, or if you're being jaded and cynical and it just hides it for six hours then it comes back. Either option seems equally likely at this point.
The other thing is that they're FORCING it on the users. Its invasive and creepy and assaulty and it has multiplied my loathing of Microsoft infinitely.
Someone in Microsoft needs to watch a lecture on affirmative consent.
Don't get me wrong, I enjoy playing with AI, both local and online. If you tell me its available if I want it, I might come dabble a bit.
But cram AI into every facet of my machine? That's a 1 way ticket to never being installed on any system I ever own in the future ever again.
Companies should treat AI like a madam treats the workers in a whorehouse. You don't make them go door to door. You let the johns come to you.
> So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling? Because if you would have, the failings would have been obvious, right? Right??
At what point does Charlie Brown not kick at the ball Lucy offers?
It's been 19 years since "Microsoft Re-Designs the iPod Packaging". [0] Is the disconnect displayed in this message thread that there's always 10,000 new people discovering a fact? [1]
Hard disagree: there is a whole universe of Windows-based developers. But even for them, the best offer seems to be a frequently-updated but still-entirely-underwhelming Visual Studio Enterprise plug-in that after 6 months (or so) can't even show proposed changes in response to a prompt without destroying surrounding code...
As a .NET developer who actually likes some Microsoft products, I can say this: the Copilot series is the worst thing they've shipped since Internet Explorer—and honestly, it might overtake it. The sad part is they had a huge head start before competitors gained access to powerful models, yet this is what we got.
If you haven’t seen how bad it is, here’s one example: Copilot Terminal. In theory, it should help you with terminal commands. Sounds great. In practice, it installs a chat panel on the right side of your terminal that has zero integration with the terminal itself.
It can’t read what’s written, it can’t send commands, it has no context, and the model response time is awful. What’s the point of a “terminal assistant” that can’t actually assist the terminal?
This lack of real integration is basically the core design of most Copilot products. If you’ve been lucky enough to avoid them, good for you. If your company forces you to use them because they’re bundled with a Microsoft license, I genuinely feel your pain.
Yes, but...does it? How do you know it isn't missing key points raised in the meeting? Not collecting actionable items?
Summarising seems like the absolute lamest thing an LLM could do for me. Like I want a Reader's Digest of my life, written but a word guessing machine.
This is arguably the only use case where a chat interface (or just text) actually works, because it's an information compression task. But building a $50 billion business strategy just on the fact that AI can recap a meeting is risky. If this is the only feature that works well, then the business ROI is questionable
> lack of real integration is basically the core design of most Copilot products
I would wager a month's wages that this is the doing of some internal Security Review, wherein a bunch of out-of-touchers decided that the customers will want to prefer to be Safe and Secure instead of getting some actual value from integrating copilot into shell workflows.
Meanwhile people are yolo'ing it with various janky DIY wires and duct-tape githobbits that mash together whatever open weights model and user-level access to the system (or worse).
> a bunch of out-of-touchers decided that the customers will want to prefer to be Safe
You mean the other way around, right? Because what could possibly go wrong when we let a language model hallucinate its way through which terminal command rhymes best with your prompt according to that SO comment from training data.
i mentioned this upthread but an LLM with enough access to be fully integrated into all apps/services/files in an enterprise managed workstation sounds like privilege escalation attacks just waiting to happen.
You'd be hearing a few more hair raising failure stories if they hadn't done that. And possibly a few big customers institute Win11 or Copilot site wide bans. Or just straight up going out of business. In businesses other than software, it's possible to be legally liable for mistakes.
> This lack of real integration is basically the core design of most Copilot products.
I think they're scared of the very real security issues with LLMs that may be unsolvable. It's not wise to give an llm free reign, at best maybe across your local computer but to be fully integrated into every application and every file it needs root. That would be the front door to many privilege escalation incidents in an enterprise managed laptop/desktop.
When was the last time Microsoft had a unified vision that was focused on building an amazing line of products that integrated well with each other?
I can only think of short snippets in history where they moved in that direction for maybe a year or two & then went scatterbrain.
Microsoft has benefited from a monopoly in the enterprise and has never been forced to innovate from a product perspective. See Slack/Teams as a case study of how they have operated when even slightly pushed.
* Edit - .NET, C#, TypeScript teams are an exception to the above. Highly underrated. Amazing talent there. Not sure who all gets credit. Anders & Mads for sure though.
I'd say the late '80s - early '90s when Microsoft was building the early versions of Microsoft Office. Integration among productivity apps was one of the key points of competition among all of the office suites of that era.
There were other huge coordinated efforts like the TwC initiative and the Windows 10 refactoring but those were invisible to end users.
You are right if talking about the efforts improving C# and CLR, taking the lessons out of Midori, Blazor and Aspire.
Regarding F#, VB, and C++/CLI, the poor souls are on a lifeline that makes CLR nowadays stand out for C# Language Runtime, instead of the original Common.
And the chaos that reigns on Windows Forms, WPF, WinUI, MAUI certainly isn't helping.
Finally they are constrained in what they can put into C# DevKit as means to not canibalize Visual Studio and Windows sales, it can only be good enough, not great and feature parity.
Also regardless of the reasoning behind it, having TypeScript rewriten into C# instead of Go would have been a great opportunity to make C# more relevant outside Microsoft shops, instead anyone looking to contribute to TypeScript compiler will be learning Go instead.
Try to think about it like a bunch of merchants running to line up in front of you as fast as possible with anything that resembles a product you might buy. The only thing they were successful at thus far was recognizing that something needs to be in your terminal, and they ran as fast as possible to your terminal with anything to beat others to the front of the line. I suppose they are doing this with all their efforts. A universe where a C-exec said in no uncertain terms, "get anything out there", is the very universe we're in.
Apple is the only merchant not running to line up with anything at the moment.
IMHO, one company needs to make the bold move and make a fork of their OS that is AI native with AI native apps/workflow and phase out the old paradigm. It'll have to be two product lines, but I think the new OS will have uptake like we've never seen before.
AI works. It's actually useful. Since GPT-4, tool calling capability is good enough. It's trivial to do a better job than Copilot on any task using any current model of the major LLM providers. I'm not talking API, even with basic chat frontend, regular users easily beat Copilot by simply copy-pasting between Word/Excel and the chat frontend.
If a twelve year old can one-shot a better product for any given use case than Microsoft Copilot, then it's not just "merchants running to line up in front of you", something more basic must be broken.
The crazy thing is Microsoft was so early with an AI product but was burned pretty badly when it instantly turned into a Nazi. Funny how the constant complaint on Twitter is how the modern AI agents are too "woke" and how Elon has to keep fighting his own AI model to conform with his viewpoints.
I think the biggest revelation of the last 3 years or so is that Microsoft does not have either the will or the talent (or both) to effectively execute anymore. Everything it currently stands on is a legacy product with roots in the Ballmer or Gates eras. They owe their Azure footprint and "success" today to Ballmer.
Their inability to produce anything useful with Copilot is the largest example of this, but there are others. They are getting lapped by a ~300 person software company in the race to consumer-ize an x86 PC a into turnkey gaming platform, even with $100 billion in game studios and owning the API that every major game is developed against. Their footprint in education is gone, completely replaced by Google who not only produced an operating system that could be effectively run and managed on commodity hardware, but also developed the centralized functions for school administrations to use to manage classrooms at scale.
The consumer situation for Microsoft right now might be even worse than it was when Nadella took over.
Another example: the complete surrender of the Xbox platform. They spent the last 10 years buying up game studios and then closing them, failing to develop any good exclusive titles, and letting Sony eat their lunch. Now they are officially giving up and will just let Sony own the console space altogether.
Xbox was THE gaming console 20 years ago. Playstation was always a contender but Xbox Live was synonymous with the online console gaming experience. Halo was an untouchable juggernaut of a series for the first 3 titles.
It's mind boggling that Microsoft just let all that die without a fight. Worse, they seem to have actively shot themselves in the foot and then given up.
Playstation/Sony always a contender? They were the clear leader. The PS2 is still the best selling console of all time. 160M units to Xbox's 24M. Xbox360 caught up but the PS3 still sold more units. PS4 sold twice as many as Xbox One. The console wars has been mostly Sony vs Nintendo for a long time.
I don't have data to support it and I know AD was a gamechanger but I feel like Microsoft pumping free software to educational institutions was the main driver for it's adoption in the corporate world. When every person who kind of knows how to use the computer, can use MS windows and office and has to be retrained to use anything else, it's a no brainer to just give users what they know.
Don't you think this is why alarm bells should be going off right now about Google's complete domination of primary school systems? Google Classroom and Chromebooks are everywhere. Google Docs is the lingua franca, not Microsoft Word
The main benefit of Google Docs is that, at least, Google is a better steward. Not good, but better.
And also, their software is exceedingly simple, which is great for breaking vendor lock-in. Microsoft got away with murder because none of their shit works like it should and every single product they make is the "black sheep" in the category it's in.
These predictions about the decline of Microsoft are like the Year of Linux on the Desktop; neither is going to happen anytime soon. Y'all can start predicting doom when there's a multi-year trend of declining revenue for MSFT and then maybe there's something to discuss.
By that time, it will already be too late. I'd argue it's probably already too late... dead man walking kind of situation. Definitely true for consumers, it might survive out of inertia in the corporate environment for decades.
Yes, I miss the Windows developer experience from Balmer era, "Developers, Developers, Developers" certainly doesn't apply to Windows native development experience.
The only thing good from Satya era on Windows development experience, is the improved terminal, and that now I don't have to install VMWare Workstation any longer.
There was a honeymoon period after Nadella took the reigns where it looked like they were putting out good products - Edge before enshittification, VS Code, buying GitHub.
Microsoft's entire business model for decades has been to shove shoddy products down people's throats. And somehow, they have figured out how to do it too, because otherwise Teams wouldn't be used by anyone.
Microsoft’s best pitch (and Google benefits from this too) is that contracts are annoying and take forever to execute. If you can sign a deal for Outlook and Teams it’s so much easier than separate contracts for Outlook and Slack. You’ll get very far with that logic alone.
Most companies I’ve been at that use Teams over Slack is not “We can’t get contract for Slack” but “We have Teams included, why would we pay for Slack?” - Accountant
I guess Microsoft lost this battle, at least at some companies, because I'm now at one that uses Slack and Google, with no dependency on Microsoft Office.
Microsoft won this battle if you check the numbers. Last I saw it was 85% Microsoft vs 15% Google which seems right with my experience. Current company is Google Worksapce while last 3 were Office365.
Teams was clearly never meant to be used as a standalone product. It was a pretty effective defense play. Block other competitors from gaining a foothold.
Thats the whole point. The only people using Teams are the ones who are already committed to Microsoft 365. Companies on GSuite mostly use Slack, I doubt there is a single one using Teams.
Its all about Excel. It really is the best spreadsheet, and everyone knows how to use it. But that comes in an Office bundle that includes Teams. And that is why we must suffer.
Was certainly the case in the early years of Google Sheets. For me, the gap is entirely closed. I'm willing to believe that Excel still has the better platform for extreme power users but I've done some pretty slick stuff with Google Sheets and that was four or five years ago. It must be even better today (though I'm not currenlty doing much with spreadsheets).
Has google sheets filled out the lambda helper functions yet? If so that could narrow the gap.
Passing around vba based xlsm is really awful, so if google sheets has lambdas they can probably get a lead with google sheets queey language over filter.
Groupby and pivotby are the new excel alternative, but if they filled out lambdas, then does that keave VBA and power query as the only reasons for Excel?
Not sure what lambda helpers are. But you can write Google AppScript which is just JS and do some pretty cool stuff, including define custom functions and fully integrate with Google APIs. I've used to to send emails, or create calendar events for example.
I agree with it, but it's a wild world we live in when the best spreadsheet has default behaviors which will fuck your data pasted into it when you're not paying attention.
Yeah, its also awesome when you open a CSV with long numbers like tracking numbers or IMEIs and they get converted to scientific notation and lose "precision" when saved in Excel format...
I think the expectation of software quality in consumers has completely evaporated, and so people expect appallingly bad software and think it's normal. This isn't just the case for Microsoft - look at the baby-ification of macOS UI (let's make it like a children's iPad), any Google software offering (very poor performance) and the majority acceptance of web-browser-in-a-window-is-a-desktop-application software development. They accept things like Slack and Teams (awful software) but also put up with Skype getting worse in the decades before.
Office 2003 and prior were quite good, but then people think Google Docs is somehow equivalent to the functionality of Word.
Admittedly, Active Directory was afflicted by impossibly tiny windows for the management tabs, but the functionality worked (and you could write your own extensions to the LDAP tree and COM-based UI interface for them) as proven by the rehashing of part of the functionality into Google's "organisational" offerings (sign into Chrome and receive restrictions from your organisational (company) overlord, the new Group Policy).
It's a real pity. If we showed software today via a time machine to ourselves 25+ years ago, we'd be shocked at how slow and ineffectual it was and deeply distressed that this was the norm worldwide.
I'm guessing the broader demographic of users simply don't think the software is crap. My buddy working in water transportation was just raving about Teams to me the other day. His praise basically boiled down to being integrated with his organization, providing him easy access to his department-resources. I suppose it does serve my buddy well.
Could be because I use a Mac, and I do most of my work chatting in Slack, but I don't hate Teams. It has my calendar, it has chat that is shitty but good enough during meetings (although finding which room you have unread messages in can be confusing), and it has video meetings with transcripts and great Exchange integration.
Ditto. The more interesting part is how many people will defend it. Presumably some mix of post-purchase rationalisation and inherited assumptions about what's "standard" even when those assumptions stopped being true ages ago.
I find it infuriating, but that's how the system's supposed to work. It's the definition of a monopoly and they're in the extraction phase. When there's no competition (and eventually there's always going to be a winner) you don't need to make good products anymore.
They've successfully indoctrinated whole generations to use Windows/Office. Here in Brazil using a computer was (probably still is) synonymous to using Windows/Office. Everyone had their pirated version of Windows and many don't even know that alternatives exist. When those people open companies they'll use what they know.
Software companies have to build for the most popular OSes and most can't justify anything else. Which then means most software only works on Windows and people can't leave it even if there are better alternatives (see Adobe). Finally, any non-closed computer comes with Windows so the cycle continues forever.
My theory is that they deliberately make Windows so shit to filter out anyone with taste. Once you have a userbase of people who don't know better, you can sell them any old crap. Like Teams.
Reminds of an research article from Microsoft!. It detailed on why scam emails about `Nigerian prince` are so obviously dumb. The reasoning being it specifically need to target only those who can fall for it. Anything more sophisticated and they would get people who wouldn't fall for scam in subsequent communication.
Except it makes no sense because as a scammer your goal is to get as many people as possible in contact with you so that you can scam them. You can only score on the goals you attempt so cutting out any person, no matter the reason, is illogical.
You’re assuming that there’s no cost involved in moving a potential victim through the pipeline. I’m sure AI has changed the game, but the general idea was that beyond the initial blast of spam you would have someone actually responding to those who fell for it. Putting in signals that it was a scam filtered out individuals who would waste scammer time because they would eventually figure it out before falling victim. By selecting for people who literally can’t pick up on obvious signs of a scam, you save yourself a lot of time and energy.
Keep following through the logic... You manage to hook someone who absolutely knows you're a scammer, and they keep responding to you taking up precious time you could be spending with someone who is actually likely to give you money. So, what is the upside to getting a response from someone who is never ever going to give you anything?
Occam's razor says they are just bad at English grammar because it isn't their native language/dialect and their education probably wasn't that great.
This is easily demonstrable by conversing with the scammers and noting that their actual English ability is the same as exists in the initial letter. Even when they have no chance of the scam succeeding and have been outed they write the same way. You can see plenty of evidence here:
Occam's razor says the sun orbits the earth, everybody dies from Sudden Unexplained Death Syndrome, and the correct way to spell Occam's razor is Okams Raza (in all languages, because lavishes other than English are difficult).
It's literally a platitude. It's like the saying 'when the going gets tough, the tough get going': it's reallyemorable and descriptive and is maybe a good guideline in many situations.
But using it to evaluate the tensile strength of various metals according to their velocity would be wild, because it had never pretended to be anything like a rule. It's not like theory of gravity or 'I before e except after c', which are based on actual analysis and results.
Legit assuming that everything is as simple as it can be, that the most obvious idea to occur to any untrained observer is the most accurate, is literally a guaranteed way to go though life without understanding anything, at all. Using it to argue with people who appear to obvious what they're talking about (and there are so, so many undisputed studies on the exact reasons scammers do what they do: it's too filter people or. There is no debate, academically) is a pretty slippery slope to 'anybody who doesn't think and act exactly like me is lying, because no reasons or facts exist unless I personally hold or after with them', and it's definitely a thought process worth challenging.
Although to be fair, its best application might be re. online arguments that you don't really care that much about. So if you just meant that the previous poster had given a reason and you were going with that because it's easier, my bad.
>"Microsoft's entire business model for decades has been to shove shoddy products down people's throats."
I remember this one. In the 90s MS reps would come to our company and sing about how their Visual Basic was superior to Delphi. When pointed to countless features that proved the opposite all they were able to say is that the MS has bigger dick.
Their recommendation was to have 2 developers instead of one we had. One would code GUI / front end in Visual Basic and the other write DLLs that would do all the meat.
I once worked at a company that wrote their UI in Delphi but then wrote their logic in C++ using boatloads of COM and MFC (for non-gui elements; the std library with containers was "new" at the time) so they ended up with 20 years of development baggage to talk to SQL Server over ODBC whilst implementing their logic on a repeated basis in C++ (COM loves you implementing interface functions you can't remove), calling from Delphi.
They were trying to port it to a "modern" system and modern compiler so had millions of lines of code to fix, and their UI was MFC-based (so another shot in the foot).
Microsoft's entire business model has been tying. Countless millions are forced to use Copilot because their IT department has contracts with Microsoft, and those same contracts are why they use Office, Teams, and so on. Their developers use Visual Studio, deploy to Azure, and run it all against SQL Server. Their email comes from Exchange.
It has been an incredibly lucrative strategy. We all herald some CEO's prowess in growing revenue when they've been doing the same playbook for decades now, and have been running on the inertia of Windows dominance on the desktop. Every new entrant is pushed out through countless incredibly lazy IT departments that just adopt whatever Microsoft shits out.
It's actually surprising that the one and only area where this really failed was as they tried to lever tying to the mobile market. A couple of missteps along the way are the only reason every office drone isn't rocking their Lumia ExchangeLive! CoDevice.
People here are mostly too young to remember but the original Microsoft business model was this:
Find a software market currently addressed by high price products; create a reasonably good product for that market; sell it for significantly less than the incumbent. Sell much higher volume of said product than the incumbent, thereby make much more profit. Repeat/rinse.
The Windows lock-in, embrace extend etc came after this. You can't lock in customers if they didn't already willingly buy your product.
No the original Microsoft business model was to get the incumbent (IBM) to bundle your product (DOS, bought from someone else) onto their product so that you had a near-monopoly, then use that to sell your other software onto that, occasionally making technical changes to make it difficult for your competitors.
>Microsoft, which needed an operating system for the IBM Personal Computer,[9][10] hired Tim Paterson in May 1981 and bought 86-DOS 1.10 for US$25,000 that July
IMO, it's time for leadership change at Microsoft. Satya revitalized the company but now it needs a Product person that knows how to rebuild the quality of it's products.
I would think that if they actually spent the time and money fixing the core functionality of their core products (like Windows and Office) that they might have a much easier time promoting things like Copilot. Instead they leave their users wondering why they're so hell-bent on shoehorning AI into a Start menu that takes whole seconds longer to open than it should or into Windows Search that regularly fails to find installed programs or local files.
Microsoft is a public company. That means their primary product is not products or services, it's their stock. Selling products & services can be an advertisement for their stock, but there are other methods of convincing people to buy their stock, too. Currently the stock market only wants stocks that have "AI" associated with them. It doesn't matter whether users like it or not, because having a viable business is not what the stock market is currently focused on. So, Microsoft is doing what they need to do to sell their primary product: shove AI into everything.
If you're someone who owns Microsoft, what option would you prefer?
1. Stock price remains the same but revenue doubles.
2. Revenue stays the same, but stock price doubles.
Assuming all else equal, and recognizing that this is absolutely a simplification, but if these were the two choices then it seems a no brainer that you'd go with option 2. Revenue is a means of increasing stock price.
That’s like asking whether you’d rather have a pizza with the same diameter but twice the area, or a pizza with twice the diameter but the same area. It makes no sense.
Stock prices were, are, and will always be rough approximations of the NPV of future profits. It’s not perfect, of course, but it’s roughly true.
Doubling revenue in any remotely sustainable way will have way more than a 2x impact on stock price because of exponential growth. So yeah, as a stockhodler, you’d rather double revenue with flat stock price because you’d buy the crap out of the stock when you realize the market has not factored revenue growth in to pricing.
Imagining that public companies care about stock price more than revenue is literally like saying a hungry customer care more about pizza radius than area.
I guess it depends on what kind of investor you are.
If you're holding MS stock long-term, and you plan to gradually shift away from equities as you near retirement and then gradually liquidate your holdings to fund your retirement, juicing the stock in the short term does nothing for you.
If you're holding short-term, then you also need to sell the stock after it gets juiced, so that you can move your capital to not-yet-juiced stocks.
Missed a point above - that for said short-term investor... that strategy doesn't actually work, since a "sell high buy low" strategy on individual stocks is outperformed by just holding ETFs long-term.
So really, which investors does short-term stock juicing benefit? Insider traders, I guess.
A long term investor would prefer 1. Most likely option 2 would shortly be followed by halving of the stock price. Option 1 lets someone buy shares in a company that is greatly undervalued and will lead to long terms gains.
I just think you are the other person who replied to me are making the same mistake, which is mixing the means of accomplishing a goal with the goal itself. This is a mistake I see a lot (especially in software development), where people get too attached to a specific means or method that they end up confusing the method itself with the actual thing that needs to be accomplished.
The goal for an investor is captured in the stock price itself. In programming terms, a corporation is a function whose output is its stock price/market cap, and revenue is but one of a host of inputs into that corporation that determines what the stock price is. Other inputs can be operating expenses, whether dividends are issued, future prospects for the company such as entering new markets etc etc... and you can have beliefs about how those various inputs affect the output or how these inputs change the output over time (short term vs. long term), that's perfectly fine... but when push comes to shove, the goal is not the revenue, it's not entering a new market, it's not reducing operating expenses... the goal is increasing the stock price (well technically the market cap).
That's precisely the argument I'm making. A company's stock price can be undervalued, which exactly means that the stock price can increase without any change in the company's revenue or profitability. The stock price can increase strictly because the actual value of the company has not yet been fully realized without any material change necessary on the part of the company.
As an investor, that's the ultimate goal of your investment.
Sure. Decision makers are paid in stock price, not revenue. They would rather do whatever increases the stock price the most, with the least effort/expense.
Yyes, and prefer it to doubling profit or cashflow etc.
I have not looked at MS in particular, but generally that is what the remuneration of the people at the top of most public companies is most strongly linked to.
It's basically the reason this bubble not only exists but has a chance not to pop : there is so much stock value in it that the big tech all want to keep feeding it, and they're sitting on so much cashflow they can afford to do it
It's absurd, but that's where it is. And a company like OpenAI basically hangs on it, because they have obligation almost ten time their revenue and the only way this does not deflate quickly is if others keep feeding it cash.
> Boeing spent about $300,000 to help Ortberg move to Seattle. His decision comes more than two decades after Boeing leadership decided to move company headquarters out of Seattle. Ortberg received about $18 million for the months he was the CEO in 2024.
> In 2022, Calhoun received $22.5 million from Boeing. Most of his 2022 compensation was in the form of estimated value of stock and option awards. He received the same $1.4 million salary as in 2021. ... In February 2023, Boeing awarded Calhoun an incentive of about $5.29 million in restricted stock units to "induce him to stay throughout the company's recovery". In March 2023, Boeing announced Calhoun was being given shares worth $15 million that will vest in installments over three years.
Works for Tesla, the absolute poster child for this kind of behavior. They can lose double digit percentage of market share in key places and still give the CEO a trillion dollars.
I've been thinking about this recently. The centrality of the stock market, while historically a great tool to allocate resources efficiently, might actually be a big weakness for the USA today. A capable adversary, like China, can kill entire strategic sectors in the US using the stock market. If they undercut the US companies and are willing to accept low returns on their investments, then the respective USA competition will be driven out of business by their investors, because there will be other sectors to invest in, with higher RoI. Do this at various points in strategic value chains, and over a decade or so it might kill entire verticals in strategic sectors, leaving the US economy vulnerable to any kinds of shocks.
As someone who is essentially financially illiterate, what does this mean, "allocate resources efficiently?" Nobody's investing in companies that promise to cure world hunger or alleviate childhood suffering. They're investing in technologies that can extract the most wealth from the population, regardless of externalities. Is that desirable?
Then again, I can't fathom what people would be doing with their money if the stock market weren't there. I imagine they might naturally wind up with some sort of...stock market.
The operating principle here being that prices are units of information, which in aggregate reveal some combination of market demand, present supply, production costs, etc. All else being equal, an investor who's looking to put an investment into a new business will try to find the best rate of return. The existence of a relatively higher profit margin for an industry suggests an unmet market need, and then directs the flow of capital into it (if you expect that for every $1 you invest into a roofing nail plant will return $1.25 over the next year vs a $2 return from a new insulin plant, more new cash will flow into the insulin plant, more insulin gets made, and if the investor guessed right about the demand for it, they turn a profit). In a sentence, money flows towards trying to give people what we think they want more of.
The theory posited above is that you could try to manipulate these signals as a sort of economic warfare. If you expect that every dollar you put into our aforementioned roofing nail factory will get you minuscule or negative return, nobody's going to want to invest in building/expanding nail factories, and they'll put their cash somewhere it can grow instead. This is all well and good so long as you've got happy trading relationships with people who can sell you nails, but if one day the nails stop coming--you've got a supply chain shock until you either open new factories or find someone else willing to sell nails to you. The theory here being that if you had a LOT of goods that became tied up in a single point of failure--someone forcing that failure could create a great deal of internal instability to be exploited for geopolitical ends.
That's what's meant by efficiency, it's allocating it to the place that has the highest return on investment.
As you point out, in practice what's efficient is what can capture the highest return, not necessarily the highest return per se. If say investing in education had high returns society wide but those returns couldn't be captured, that's not an efficient use of private capital.
As somebody doesn't consider himself a capitalist, wouldn't it be fair to say it is "the most efficient" in precisely one thing: capital reproducing itself?
And if so, why is that necessarily a good thing? Why should that be our goal as society as opposed to things like minimizing child mortality, increasing literacy rates, making sure we don't have a ton of our fellow humans living on the street in misery etc etc - things that make the lives of our fellow humans better? Why is capital growth the metric we have chosen to optimize for? Surely there's better things to optimize for?
Excuse the polemic, but infinite growth with no regard for anything else is the ideology of a cancer cell - and to me that is increasingly what it feels like when we are wasting all these resources on a dying planet just to make numbers go up.
Ultimately that money is made by people choosing to spend their money on something, because it helps them, because they like it, because they need it for whatever reason (real or imagined). That's what grounds the financial markets: eventually someone is buying a thing because they want the thing, and all the rest of it is basically just figuring out who can make the thing, how many people want the thing and how badly, and whether the stuff used to make the thing could make a different thing that people want more. Financial markets can depart from that reality for a while, but mainly because of a collective belief in some falsehood about the above (everyone really badly wants AI, right?).
Number go up infinitely is due to inflation and that's basically just an incentive to not hoard cash indefinitely, and instead use it for something useful. But the only thing that uses up is numbers. Everything else is because people, on average, want more stuff and are willing and able to work hard to get it.
(Of course, this generally means that the markets chase the desires of those who have something valuable enough. People who don't will be marginalised by this mechanism, for sure. And of course there's lots of opportunity for people to steal or abuse powerful positions in the market to the detriment of others. Which is why a free market is not the be all and end all of organising a society, and other organisational structures exist to regulate it and to allocate resources in a less transactional manner)
But isn't that counter to the very article we're commenting on? Everyone is shoving half baked AI junk into everything because that's what makes number go up on the stock market, but I'm pretty sure that's not actually what most people would want those resources to be used for.
I'd posit that markets are completely detached from the real world and are more of a speculative/religious element than an indicator of any ground truths.
Edit: I just realized I missed a sentence of yours where you kinda spoke to this. I still believe that this is more of a rule than an exception - there is nothing inherently tying markets and reality together - they're mostly about people making bets on what the next big hype is; not on what is actually useful to anyone.
Most people or most people with money? Ultimately it's the people making investment decisions you need to convince in order to get the money from them. Then the reality check is whether you can give them that money back and more. Those investors should in theory be self-interested in making sure that this is actually useful, because if they don't they'll lose their money, but in practice they are not superhuman and prone to fads and echo-chambers (especially because it's a relatively smaller group that can move a lot of money around, and short-term investing rewards running with the crowd to some degree), but there's not a single group of people that would not fall victim to poor decisions in this regard (whether you're imagining a centrally managed economy, the electorate, or some hypothetical benevolent dictator).
Optimizing for capital returns is a simplification of the real world, where it allows for comparing whether it makes more sense to put one's money into opportunity A or B.
There's a lot that's not captured by solely looking at dollars, like the examples that you bring up, such as quality of life, human welfare, and so on.
If you care about minimizing child mortality, increasing literacy, pulling people up out of poverty, you should be a capitalist, as it's empirically the best way to meet those goals. This seems to be a hard thing for many to understand or accept because it is largely a second order effect, the capitalist primarily concerned with their own personal gain but winds up improving the lives of others as a side effect.
This is the essence of Adam Smith's often misunderstood invisible hand metaphor. Of the individual he observed: "By pursuing his own interest, he frequently promotes that of the society more effectually than when he really intends to promote it." Second order effects stack up and improve quality of life for more people better than trying to do so explicitly.
Multiplying capital creates abundance and that abundance allows for improved standards of living for and the means to spend excess resources in support of charitable endeavors. Growth is good because it means more abundance and opportunity. I would argue that pursuit of growth is not an ideology but a force of nature. Life is opportunistic and will expand to wherever there is fertile conditions, and often adapt even when they are not. We are part of nature and understand this intuitively, seeking growth opportunities. As an example, one is better off being part of a growing company (more wages and opportunities) than one that is stagnant or declining (fighting for scraps and survival).
>If you care about minimizing child mortality, increasing literacy, pulling people up out of poverty, you should be a capitalist, as it's empirically the best way to meet those goals
If you look at it empirically, the majority of people brought out of poverty (and I suspect the other metrics but am not as familiar with them) in the past few decades have been in China as the result of deliberate policies by the CPC.
We’re already there when it comes to having the industrial base necessary to fight a protracted conventional war with china. Which leaves a large ? over the US dominance over the pacific.
Y-yeah. HYPOTHETICALLY, this is something an adversary to the USA might attempt to do, and it would really kneecap the US if they were successful.
But would only happen if USA decided to totally financialize all sectors of its economy and make a small set of oligarchic corporations THE load-bearing element of its strategic capacity, leading us to chase market returns even if those returns totally kneecapped our ability to build anything at all of actual value.
While not strictly necessary, it is a great power multiplier.
It helps as it is both a gauge of the success of the strategy, and also a lever where the process can be fine tuned, eg. slowly buying stock then strategically dumping in the right time, correlated with other external shocks can have wider effect to whole industries through controlling the public opinion on specific industries.
It rewards self-destructive behavior in favor of short-term gains. Shareholders have *zero* commitment to the companies they buy shares from and will happily switch their entire portfolios on a whim. It's essentially people chasing the new shiny thing every single day.
Let's not forget it's a known fact that people with insider knowledge will profit over everyone else.
How is that efficient in any shape or form?
> If they undercut the US companies and are willing to accept low returns on their investments, then the respective USA competition will be driven out of business by their investors, because there will be other sectors to invest in, with higher RoI.
You're basically explaining one of the reasons stocks are a horrible idea for distributing resources.
It has nothing to do with whether or not it's central or distributed, it's merely the incentives they create. It's essentially Goodhart's law on steroids.
Depends what you compare it with. I grew up in the post soviet union. That system allocated resources to various monopolies who were too big to fail. Turns out allocating capital based on who can make things that people want/need to buy, and do it with a profit, multiplies said capital way faster. From this point of view, over time your initial base grows into all kinds of industries etc. That's probably why the USA won the cold war.
Until your economy congeals into various monopolies who are too big to fail, and you lose the rematch because you're so busy counting beans you stop paying attention.
A bad system is still bad even if something else is worse.
> That system allocated resources to various monopolies who were too big to fail.
Like our banks today? Like OpenAI is trying to sell themselves as so the US government bails them out when they inevitably can't pay off their debts? Like all the other monopolies that are too big to fail?
I can't see the difference.
> turns out allocating capital based on who can make things that people want/need to buy, and do it with a profit, multiplies said capital way faster.
That assumes the stock market does that. It simply doesn't and history is proof. It's also a fact that free market capitalism has consistently done worse than before based on actual numbers.
I highly suggest you read 23 things they don't tell you about capitalism. It goes into a lot more detail and helps you understand things in a broader perspective.
If we want to improve our world we need to move past the faults of the current capitalist system and the soviet union. We need better incentives. Profit has proven to be too disconnected from actual value.
The stock market weirdly enough ruins the idea of capitalism. Catering to shareholders hurts the idea that competition would create better and cheaper products.
Its not the stock market per se. The biggest problem is a lack of good regulation to ensure competition and the resulting drift of oligopolies.
The stockmarket enables that by making takeovers easier as you have a higher proportion of short termist shareholders who 1) fail to block value destroying acquisitions on one side and 2) jump at the chance to make a quick profit on the other.
that's not the idea of capitalism; the idea of capitalism is that you should be able to make money by virtue of owning stuff. it's an inherently rich-get-richer scheme, competition has little to do with it.
I don't see how free market capitalism fixes that. I looked up a definition to make sure I wasn't missing something and as per investopedia:
"The term “free market capitalism” refers to an economy that puts no or minimal barriers in the way of privately owned businesses. Matters such as worker rights, environmental protection, and product safety will be addressed by businesses as the marketplace demands."
it's basically worship of owning the means of production and not being regulated in its use, e.g. if you own a company you get to dictate all sorts of unreasonable things to your employees, and any benefits gained from automation accrue to whoever can afford the up front money to own the machines.
Competition creating better products isn't an idea that defines capitalism though: the same would apply to cottage industry.
Capitalism is defined by having the capitalist, who provides capital, and without the ability to sell their share of stock it's difficult to see what the value would be. So you kind of require stock markets.
Edit: which is why it's odd to call China communist. They have 3 stock exchanges. They're really a capitalist single-party state.
China uses Capitalism as a tool where the Party feels it would be beneficial (for the Party), and crushes it mercilessly when it gets in the way (other than this real estate problem they have right now).
In the U.S. we have mistaken Capitalism for a religion, and so it wags the dog, so to speak. Since our founding we have made some attempts at finding a balance between our use of the tools of Capitalism and socialism (in more the Democratic Socialism style, rather than the Communism style), and we had a good run in the decades after WWII. But starting with McCarthyism, and really picking up under Regan we have prided ourselves on adopting Capitalism as a religion, and it really shows up in both the income inequality as well as the increasing role of (and corrupting influence of) money in our politics/government.
That's fair, I should reframe. The incentive given to decision makers at Microsoft is company stock. That means the primary focus for everyone who makes decisions at Microsoft is the stock price, which in turn means the stock price is the primary product for the company itself.
Name a major U.S. public company in recent years that has consistently prioritized improving its product over boosting short term stock price or extracting maximum profits. If capitalism were truly a healthy system about building strong products to create healthy markets, this should be the norm (and enshitification shunned), not the exception.
What we actually see is a system of chartered extraction. Corporate executives are like Norman lords, granted their 'title' (CEO of instead of Earl of) by shareholders (rather than a king) in return for which both are/were expected to extract maximum value by any means necessary. Extractive tactics often at the expense of long-term product strength are behaviors shareholders expect if the CEO is to keep their bestowed 'title'.
Don't forget the progenitor joint stock company The East India Company, Capitalism in it's purest form without government restriction. Profit-maximizing, absentee extraction, with company executives serving as quasi-feudal lords over assets and people. Modern corporate capitalism is hard to distinguish, in its structure,history, behavior, and incentives, from the Norman extraction system, it's just dressed in a more politically palatable wrapper and forced to mellow out from it's desired East India Company style final form.
> Name a major U.S. public company in recent years that has consistently prioritized improving its product over boosting short term stock price or extracting maximum profits.
Well in some perverse sense, I'd say Meta qualifies here. Zuck isn't beholden to other shareholders and is free to burn truckloads of money on worthless projects. The big asterisk is that for Meta, "improving its product" is effectively "creating the best digital cigarettes".
Lots of companies do this (Arizona green tea immediately comes to mind), but more importantly, this has no bearing on the discussion of whether or not capitalism is "good" (really, who ascribes morality to any economic system?), because the US is primarily an exercise in regulatory capture.
Good doesn't mean moral in this case, it means valuable. Capitalism, or whatever the fuck we have right now, doesn't create much real value, but instead it creates a lot of fake value that we just proxy.
And I say "or whatever the fuck" because, as you point out, it's not really capitalism. And maybe it never was - after all, the US Golden Era was when we had the most central planning, the most social services, and the highest taxes.
But everybody right now is exploiting one teensy little flaw of the economy - humans don't live very long. There's no reason to make stuff good for later when you can take a cash advance and get rich now. Everything is always someone else's problem.
I think one fun case study is superstar CEOs. They're so highly desirable, and they work 5 years at a company and "turn it around" before moving onto the next. But if you look at the company after they leave, it's in shambles. And yet, they are the most sought after CEOs. You don't want the Arizona Green Tea guy, no, you want the guy who ran Chipotle into the ground. But hey, at least he increased their stock price for, like, a little bit.
Arizona comes to mind because it's the outlier that everyone knows about, the company that bucks the capitalist pressure. Your choice to use it basically proves my point.
Capitalism outside regulatory capture looks like the East India company. You are basically giving me 'no REAL communism has ever been tried'.
Because they so much want to be a service business than a software business. Microsoft execs are losing sleep over becoming the next IBM, not realizing they are already there and have been for a long time.
Their main problem is that they never really learned how to compete on merit, just on first-to-market and all kinds of legal (and illegal) tricks.
I’m actually somewhat stoked about generative AI from a “good enough” perspective, because at this inflection point where a lot of countries and organizations are looking for Microsoft alternatives (digital sovereignty, etc), this is the best time to be able to build and deploy alternatives with the productivity advantages (if any) AI might provide.
Big Tech thinks they have a moat, when it’s really diffuse power being made available via genAI to build software good enough to replace them.
Big tech's moat is mostly built up around perceived security (not actual security) and abuse of monopoly positions, AI is going to make it easier for them to maintain this because it serves as an addictive component for the users, or so they hope. In practice it only appeals to the lowest common denominator, which is exactly how they built their empires in the first place. AI allows non-experts to pretend they are experts with confidence, and to produce output they claim as their own to which they have no real title. As a democratizing principle that's great, but as a quality-of-service-provided item it could easily become a disaster and I think MS et-al are betting that they will be able to 'fake it until they make it' on the quality front to avert that disaster.
So far, I'm not seeing it. All I see is a massive leap forward in the first two years that still had some fundamental problems and a lot of fancy packaging of the same broken stuff since then. We're looking at band-aids here, not actual progress.
Or do things that actually work. Why, for example, can I not translate a PowerPoint using Copilot in Powerpoint? Why do I need to save it, then upload it into ChatGPT, translate it, then download it again, and open it in PowerPoint for further editing. But at the same time get all kind off nonsense I don't want pushed at me in Windows, like that MSN news clickbait crap.
Exactly, even those of us that like Windows have a hard time talking about it when Microsoft treats it so badly, I really miss Balmer era in regards to Windows.
The only good thing that came out of Satya era has been the Windows Terminal and WSL.
As long as companies, and consumers, still pick Windows and Office, then why spend the resources. Making Windows better won't move the sales number significantly, but removing the ads and the potential AI upsell is a direct hit to revenue.
The sad reality seems to be that Microsoft do not care about the majority of their products anymore. Only Azure, Microsoft 365 CoPilot, CoPilot and maybe CoPilot.
I'm not familiar with many "consumers" who still pick a Windows and Office, and in this generation, there are very few consumers picking xbox. Outside of enterprises, they seem to be losing market share everywhere, and at this rate they'll be akin to IBM or Oracle in a few years.
But "we're fixing Windows and Office" doesn't get you a fat bonus, and it doesn't get investors a double-digit growth quarter. Anyone who tries to do the right thing will be disposed of. Why don't people understand this?
I will say that with enough group policy and sysinternals turning absolutely everything off, turning all of the settings to maximum performance lowest flashiness, no web results, killing Cortana with reckless abandon my Windows installation is actually what I would consider to be snappy. I was surprised.
It doesn't make it any better that Microsoft does this, but as a piece of practical advice, it seems like it can be done. There does still exist a core of Windows under all that garbage that is fast.
All that tinkering is getting you dangerously close to daily driving linux. And the advantage there is that the maker isnt actively trying to get in your way
This is what's so amazing to me, a primary feature of cloud storage is multi-device usage. But by default this ridiculous product causes a mess if you use it on 2 or more computers, and many apps save app data to my documents so you have useless (at best) or conflicting at worst bloat being copied between computers. I want to use it but Microsoft makes it a huge pain if I do.
A huge percentage do, whether it's most or not would be a stupid metric by which to design a cloud storage product in such a way that causes issues when one does sign in on multiple devices.
I was hoping for a real look at weaknesses in Microsoft’s AI products. They ship lots of “AI features,” but only Copilot and Azure’s ChatGPT hosting see broad use. Instead, the article mostly reads as anti-MS/OpenAI without much detail.
From my experience, Microsoft’s GPT-5 integrations in Word, PowerPoint, and their ChatGPT clone struggle with basic tasks. Copy/pasting from ChatGPT still works better.
To be fair, building solid AI features is hard when model capabilities change so quickly. Reasoning and tool use only became reliable in the latest models, and when these Office features were planned, GPT-5 didn’t exist.
Side tangent: Copy/pasting from the Windows Copilot app is absolute dogshit. It makes no sense that the simple action of "copy this text" is this broken.
They bought Dragon a few years ago, and 2 years ago they debuted the Dragon Ambient Experience, then renamed to Dragon Copilot. We had dozens of doctors try it, after a handful of months most had quit, it was a bad product. We switched to a competitor at literally 1/6th the price, and we don't even have to offer it, the doctors tell each other about it and they ask for it.
Nadella has done a lot of listening through is CEO reign but it looks like MS is back in a "don't listen to customers, tell them what they'll get" phase.
Lol the Copilot app isn't even that useful on iOS for a blind person. On Android, you type something in, hit sent, and the app pipes the pure output of the AI, Markdown formatting and citation markup included, to the screen reader. That's at least something. I mean it's crumbs, yes, but we blind people are very, very used to crumbs.
On iOS, you type a message and send, and... nothing.
Big corporate AI products are all currently stupid bolt-ons that some committee decided solved a problem.
When the internet came out, did many legacy companies lead the way with online experiences, figuring out what the real killer apps now that everyone was connected were? I don’t know for sure, but I doubt it, I think it gave rise to some of the present crop of big tech, and others reinvented themselves after the use cases were discovered.
All that to say, I expect the same here. In 10 years there will be AI uses we take for granted, built by companies we haven’t heard of yet (plus the coding apps) and nobody will talk about stupid “rephrase with AI” and other mindless crap that legacy companies tried to push.
> Big corporate AI products are all currently stupid bolt-ons that some committee decided solved a problem.
Or not even .. maybe someone said all products need to be AI enabled, so now they are. Just append "AI" to the product name, add bolt-on to call an LLM to do something, and declare mission accomplished.
Big corpos have reached the stage where they can hire ex-politburo apparatchiks from the soviets or china, straight into C-suite roles, and nothing will materially change.
I moved largely away from Microsoft products in 2010 personally, and 2015 or so professionally. Just came back to them last year for a client.
What I am seeing are two really disturbing patterns: 1) really, really bad stuff that has been there for decades is still there, weighing everything down, for "backwards compatibility", and 2) a lot of horrible fluff and nonsense everywhere.
#1 is pretty self-explanatory, as I am editing docs in Word I am finding my muscle memory circa 1999 or so still works with it due to all of the formatting bugs that still exist.
#2 I saw with Windows 11 and the crazy adverts everywhere, seemingly random UI choices, half-broken or implemented tools and applications, and now random UI buttons thrown everywhere.
The 1-2 punch is devastating and makes using Microsoft products 10x more depressing than it was 10-15 years ago, and it wasn't a happy fest then either for me.
So to me, AI is just "more of the same" with Microsoft. It is more random shit thrown on the wall to see what will stick.
Backwards compatibility has always been a Microsoft staple. What used to be a huge selling point - depending on the audience - is now clearly a crutch. Right now, it seems that the tech debt has finally started making the whole stack lean like the Tower of Pisa.
I feel like the case for Microsoft inability to execute in a lot of verticals should really be studied, not saying this as a sound bite, I'd genuinely like to know how that is possible.
Their investment in OpenAI, giving them what was, at least ~1-2 year ago if not now, the best possible LLM to integrate in the office suite yet they are unable to deliver value with it.
Their ownership of Xbox and Windows should have allowed them to get a much better foothold in gaming yet their marketplace is still, to this very day, a broken experience with multiple account types. It's been 10 years.
The counter point is Azure obviously which still has great growth numbers, but that's a different org.
From the outside, it just seems like they should be doing better than they are. They have much better business integration than Google and Amazon. The fit is obvious and people are borderline hooked on excel. Why aren't they dominating completely?
What fascinates me is how someone at MSFT managed to get their own button (the Copilot) button approved. Having your own button or tab is a big deal (I remember we debated having our own shopping tab on the Instagram app while I was at Meta - and we had to abandon it when it didn't get the traction we were hoping for).
These people have a UI button AND a hardware button on actual consumer devices, and it doesn't do anything. How?
AI assistance is a gold rush — promotions are to be made and huge complex system to be over-engineered in Big Tech. The race to stake out the future empires is underway, and there is no time to think about the quality control, UX etc. But who am I kidding though, there is no time to think about those things during the chillest of times either, as any user of Power Automate can concur.
It's being treated like a gold rush but I don't think it really is. This is like dotcom 1.0 all over again. They didn't know what the best use cases for the internet were but they still poured billions into it. The gold rush didn't come until the 2000s when social media took off.
Normally you get a frontier exploration phase where fringe people experiment with the new technology and try to figure out what it's good for. It feels like we just skipped that step entirely.
It's sad that the "gold rush" on social media just means "using the Internet" is a depressing experience of being tracked incessantly and sold to, constantly. A lot of the content isn't to educate or delight, but to build a following exclusively so that you can sell to said following.
It's more like the "poo rush", where you are the one getting pooped on.
It seems to me that America is a great big nomenclature butchering machine. If they can butcher what words like crypto or AI mean, they’re willing to put their whole economy behind that. I need to rethink my investment strategies.
This is true. Just look at what selling of products with "memory" means - "the tablet has 256GB memory"; no it doesn't. It has 8GB RAM and 256GB non-volatile storage.
I suppose "AI" is easier to say than "word-guessing non-deterministic instruction transformer", but it does carry "intelligence" into the hearer, when the truth is that "AI" is not intelligent - it is a great mimicker of everything it has read.
Between the RAM shortage and the forced migration to Win11 along with a forced HW upgrade, I'd start shorting MS asap. This bit about lack of adoption of Copilot is just icing on the cake.
> FirstPageSage AI Chatbot Usage Chart (December 3, 2025)
What a bizarre way to organize the chart. Clearly Anthropic is leading -- their early bet on "programming" as the main use case is paying off.
The recent report from Openrouter [1] confirms as much: coding is the number one use case, with role playing / fantasy writing in second place.
I have a feeling the remaining use cases will never dominate, instead they will slowly mature into acceptable practices across the various industries. That will probably take longer than the investment bubble can hold though.
Claude Code usage probably isn't counted as "chatbot" use. Also, I think you're overestimating how many people program vs. how many people are using AI chatbots as the new websearch. Orders of magnitude more of the latter.
Mustafa is a side-character, he's the friend of a brother of someone who knows what they're doing. Competence isn't something that people pick up through proximity, and Microsoft is finding that out in real-time.
I personally never found Microsoft software anything but shoddy: Microsoft word, internet explorer, power shell, outlook, powerpoint... these all make me shudder... I'll give them visual studio, that one is pretty good. I think Excel is fine. Skype was okay I guess. Anyway, I think the shoddiness can't really be the actual threat to Microsoft given their track record.
When I joined my current employer, a few years ago, one of the things I thought might be good was that for the first time in a fairly long career I'd be using Visual Studio.
Previously I'm a vi or vim user for everything, for many years.
But I can say after a few years experience I'm not really impressed. It's too big and too slow. It has a few things I kinda like, a lot of half-working things I'd love if they worked consistently (e.g. some things I work on can be debugged, some can't, experts might know why but I don't) but as they are they're too unreliable to really change how I do things - but overall it's not enough to e.g. miss VS when I'm writing my own stuff, still in vim on Linux.
Actually all of the Microsoft technologies I've run into were disappointing with only two exceptions which I'll get to. Powershell felt like they hadn't really learned the right lessons from the Unix shells for example, and Entra ID (called "Azure Active Directory" when I started caring) is a confusing mess.
Two exceptions: 1. C# is a pretty good language. Mostly it's a better Java. Is that amazing? Not really, but it's still pretty good, it delivers reasonable performance, there's a large ecosystem, I don't hate writing it.
2. Azure itself has to have a way to "cut off" payments because Microsoft sold a product where students can get a limited amount of credit. The student doesn't have any money, so if they had $50 of Azure credit and Microsoft lets them spend $85 before turning off all the Azure systems that credit was funding, well, too fucking bad - Microsoft eats the $35 loss. Accordingly Microsoft are better (not perfect, but better) than AWS or Google's thing at actually turning stuff off when it exceeds what you asked to spend.
Microsoft's problem isn't bad AI, but a fundamentally wrong UX pattern. They're trying to shoehorn Copilot into linear workflows - Word, Excel, Outlook - but chat is a terrible interface for editing documents or analyzing tables. Google with Gemini is winning not because their model is better (though it is good), but because they integrate AI better as a tool, not as a conversational partner. This difference in mental model defines utility
I've used Copilot ih VSCode quite extensively, using both GPT-5(.1) and Sonnet 3.7 and Sonnet 4.5 as the LLM backend. (Before our company got the deals in place for OpenAI and Anthropic to use their tooling directly)
The results are measurably worse every time vs. doing the exact same operation with Codex Cli or Claude Code.
There's something in the Microsoft AI framework that makes the worse and more "stupid" than the competition.
My first experience of copilot was learning that my Microsoft Office subscription was increasing in price.
Then I dug a little deeper and saw that I could exclude it and go back to my old price.
For what I do with office, I have no use for copilot so you can guess what I did.
An unfortunate reality that software makers need to grasp is that a lot of people don’t use their software properly to begin with, meaning the addition of AI doesn’t push the needle - it’s just yet another “time saving feature” that the user won’t touch.
Maybe they could add a helpful paper clip to improve sales.
Edit: Or better still, convince all of their customers to throw away perfectly good hardware and upgrade to one with a single extra chip, creating a hazardous waste epidemic for landfills as a nice side effect. It's especially important to do this in the middle of a RAM and HDD shortage.
Really, I'll just never be half the great business strategist that these guys are. <sigh>
It doesn’t matter. Microsoft has a monopoly on the desktop, and they have no qualms using it to displace competing products. They did it with Teams, and they’ll keep doing it because they know there’s no appetite for anti-trust prosecution anymore (or maybe they feel comfortable arguing they’re no longer a monopoly because they have no presence in mobile).
Every procurement team is going to point to copilot, saying it’s included with the other Microsoft services a company is already paying for, so duplicate AI products won’t be approved for purchase.
Microsoft is laying claim to the desktop real estate, so in a few more generations of the technology, they’ll have the customers and competitors will already be starved out.
Not just about the products imho. I do some consulting for law firms who typically use the MSFT stack, and I was excited about the private ChatGPT services in Azure, because from my (admittedly limited) sample of law firms, nobody likes using Copilot and LLMs need to be private/secure. The amount of outdated and poor quality documentation for Azure services is amazing given how nascent these services are.
Their copilot stuff is such a mess. Over promising, dressed with marketing slop on top of an under-performing product.
However; their azure offerings are somewhat decent. AI Search is getting quite decent, access to Open AI models served through foundry is quite the differentiator (although the rate limits are an issue), their reference architectures are very helpful and their ancillary services, like document intelligence integrate well.
I'm a bit worried that their marketing dept has gotten their crappifying and obfuscating eyes set on those services though, given the recent rebranding of Foundry and what not, but the underlying services are decent.
See WinUI after Project Reunion announcement 5 years ago, unfortunately fits exactly the same description, and we are way past COVID to use that as an excuse.
And now with how things are going with the American gov, foreign companies see security flaws on two fronts: Microsoft's AI's unreliability+invasiveness and the US gov's unreliability+invasiveness relating to its stateside companies.
Microsoft is using the deep penetration of SharePoint in companies to sell Copilot license. At least in France it's well and alive and I see much more Copilot licenses than actual OpenAI uses.
I just had a sales call with someone from Microsoft who was looking for an AI tool to automate some Excel work they were doing. I doubt they'll buy our product, but it gave me a good laugh.
BTW. Feeding that AI boom, that is driving up the cost of hardware... while at the same time asking people to buy new computers only to run the next version of their OS.
Microsoft doesn't just have a shoddy AI problem. Microsoft has a direction problem. I'm no fan of Ballmer, but his Microsoft seemed like they knew what they were doing, and were actually trying to be good at it. Nadella seems extremely clueless and seems happy to just ignore and later axe consumer products that don't generate immediate revenue.
And noone should actually be shocked about his ineffectiveness. Covid was a great example of how clueless his leadership has been. Skype used to be a verb people used in common parlance, and yet they dropped the ball and let Zoom take over both consumer and enterprise segments while focusing on "restructuring" Skype into Teams for no reason whatsoever.
Prior to Covid, he was ready to let Windows run its course and axe that too. The sudden demand for sub-$500 laptops during the pandemic showed him that people still liked Windows and wanted a good OS from Microsoft. But instead of capitalizing on it to give customers what they wanted, he just gave us an ad-filled spyware with AI slop.
I have zero hope in any product with a Copilot in its name (including GitHub). At this point, unless there's a change in leadership, it's only a matter of time before XBox faces the axe.
I am honestly struggling to figure out Microsoft's licensing system. We have it here and there are some on my team that want to try "CoPilot" (not even sure what that means now) but there is no licenses option for me to enable them to do that. I am sure they are targeting the larger places than my 50 person shop but it is so confusing I don't have the time in my day to figure out how to enable copilot for my team.
I see mentions of Gemini as a fast growing alternative to ChatGPT. Isn't anyone troubled by the fact that for consumers there is no way to keep your data from being used for model training if you want to maintain history of your Gemini chats.
ChatGPT respects privacy and allows for maintaining history while also opting out of using ones data for model training
I trust Google ad monopoly to keep my data actually secure. They have a great track record of not sharing their datasets with anyone because this gives them an edge pushing ads down people throats. Google is honest about what they doing. Google also not going away anytime soon so they also not going to sell off their datasets to highest bidder.
And I don't trust Sam Altman and AI.com at all since their whole thing was built on lies. They could start regaining the trust by changing their company name.
It's not just about protecting data in the old sense - typically from other corporate entities. It would suck if your information somehow made it into a generally available model that then leaked some of that anyone asking a question
Its been a while so i had trouble finding it (but Grok obliged)
Moreover its always the edge cases that people are 'OK' with, but again if they can do it (setup the infrastructure) for one thing they can do it for anything, and it makes 'trusting' them seem naive. Since trusting was the original statement.
Microsoft’s early/rushed attempt at AI (the various things called “Copilot”) does start to feel like how they lost the mobile wars. They had a tech fairly early (Windows Mobile) but utterly failed to execute on it and ultimately, allowed a competitor to dominate with a better product.
It’ll be interesting to see how this plays out. With the sums of money involved it could end up being make or break?
Gemini is really great now. Fast is insanely fast and handles 90% of queries. Deep Research works better than OpenAI's deep research given their search expertise.
It is going to be very hard for OpenAI or Microsoft to compete with this now that Google has gotten their act together.
When you're on the internet for a while, you develop an intuition for the average level of Microsoft hate that is supposed to exist at any given time. It's a rather stable measure.
I'm seeing an unusual increase in that, and after some thought, I think the only plausible reason is that another company is afraid of them and creating anti-MS editorialized buzz.
My gut feeling tells me this is all related to long-term sustainability of AI products, and both GitHub and VSCode are major points of anxiety for other AI coding platforms.
We will eventually know if I'm right or not. If we see moves that point to other companies trying to decouple from GitHub and VSCode, then it's probably what is actually happening here.
Isn’t the problem more so that nobody wants to buy it and Microsoft is leveraging its power and influence to successfully coerce and pressure people into buying them?
You using OpenAI’s products? Aren’t Microsoft getting cuts as owners?
It doesn't (but it seems pretty much every Windows install has it installed, and if you try to open it, it just tells you that there's no Cortana anymore.)
BTW. They did the same with Maps and the PDF reader: the two apps I used the most on my Windows tablet. An "upgrade" replaced the app with a nonce.
That just made me so incredibly angry.
Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
But this also means they end up with average products. They don't have the talent to do something exceptional.
This has worked well for them when they can just come in and copy something (say AWS in Azure) and not pay the innovation cost, but AI seems different for some reason, perhaps in the same way search was. You need the top 20% in order to really be successful.
That’s exactly what happened. For the past decade, the crème de la crème went to Google and Meta, which offered nearly double what Microsoft paid new graduates. Microsoft hired the next tier, after the top talent had already been skimmed off by Google and Meta.
this is just not true. Building great products with average talent is a sign of great management, and it's been done before in both business and sports. moneyball is about this idea at some level.
Plenty of SV is building below average products with exceptional talent.
> Plenty of SV is building below average products with exceptional talent.
Yes, you can hire exceptional talent and give them poor directions, resulting in poor products.
But to hire mediocre talent and still produce competitive products you must have an unfair advantage of some sort. The Windows and Office monopolies gave Microsoft that unfair advantage. But it is becoming clear that this unfair advantage does not extend to AI.
> Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
This is such a BS attitude. A person’s ability doesn’t map directly with their salary.
The very best developers I’ve ever worked with? Small companies outside of major cities (for life style reasons). Not everyone is trying to max their income; some want a good job doing something they love.
Small companies can offer more meaningful impact, if you’re not the type who likes being a tiny cog or commanding armies.
You are coping. Overwhelmingly the quality of developers is reflected in their pay packages. Overwhelmingly most people are motivated by money/benefits/QOL/GPUs.
This is an amusing response. I personally have chased the TC (to an extent), climbed the ladder, and sit at the top (below the c-suite). This isn’t cope, it’s a genuine reflection based on many years of experience having worked for multinationals in various countries.
Perhaps compensation should be a reflection of ability, but the real world is far more nuanced.
The only time they copied something successfully and did not rely on major tie-ins with their existing monopoly was Xbox and that division lost money hand over fist for a long time.
> Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
That's only if you're only looking at the top 10% or w/e of the market in the first place. New grad software engineer salaries at MS are higher than the median software engineer salary in the US.
Microsoft has this problem with most of its products.
It's not just AI, it's a market fit and quality problem.
They don't need to solve it, however.
Their strategy has been quite clear: make it barely usable so that is passes muster to auditors, integrate it with systems that corporations need, and sell them on the integrations.
Teams and Azure suck?
So what?
Big companies will pay for that, because it's integrated with their ldap, has an audit trail, gives them the ISO-whatever stamp, and lets them worry about something else.
That the users are miserable is almost never the question for the ones signing the checks.
In a world where box-checking is paramount, this approach is a winning strategy.
I don't understand why they name everything Copilot now, it's the most confusing thing in the world. Copilot AI, Copilot 365 Copilot. GitHub Copilot makes the most sense I think because it's Copilot AI but for coding.
Well, the major problem Microsoft is facing is that its AI products are not only shoddier than average, which is nothing new for them in many categories, but that this time the competition can actually easily leapfrog them.
Like, I have a 'Copilot' button prominently displayed in my New Outlook on MacOS (the only platform where the app-with-that-designation is sort-of usable), and it's a dropdown menu, and it has... zero items when expanded.
I asked my 'Microsoft 365 Bing Chat AI Bot Powered By ChatGPT<tm>' about that, and it wasn't able to tell me how to make that button actually do something, ending the conversation with "yeah, that's sort-of a tease, isn't it?"...
Oh, well, and I actually also have a dedicated Copilot button on my new Lenovo laptop powered-by-Windows-11. And, guess what, it does exactly nothing! I can elect to either assign this button to 'Search', which opens a WebView2 to bing.com (ehhm, yeah, sure, thanks!) or to 'Custom', in which case it informs me that 'nothing' meets the hardware requirements to actually enable that.
So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling? Because if you would have, the failings would have been obvious, right? Right??
The other day I've clicked on one of Outlook calendar's copilot prefilled questions: "who are the main attendees of this meeting". It started a long winding speech that went nowhere, so I typed in "but WHO are the attendees" and finally it admitted "I don't know, I can't see that".
Absolutely! There are so many scenarios where they could actually add some value, and they're fulfilling, like, exactly none of those?
Even in Visual Studio Enterprise, their flagship developer product, the GPT integration mostly just destroys code regardless of model output. I truly cannot fathom how any of that made it past even a cursory review. Or how that situation would last for over 6 months, but, yet, here we are.
And, again, it's fine with me: I'll just use Claude Code, but if I were a Microsoft VP-or-above, the lack of execution would sort-of, well concern me? But maybe I'm just focused on the wrong things. I mean, Cloudflare brought down, like, half the Internet twice in the past two weeks, and they're still a tech darling, so possibly incompetence is the new hotness now?
Imagine a circumstance where Windows Search was as good as Apple's Spotlight, and could integrate with cloud services to index documents, browser bookmarks, web history, maybe podcasts, etc.)
Hey Copilot, where is that document I was reading about the new network diagramming software Jacob is testing out?
Or Hey Copilot, my disk is getting pretty full. What software is taking up a lot of space that I haven't used for a while? Or are there any files I can move to cloud storage to free up space?
But no, instead it's just 'we're going to take screenshots of all your windows, OCR it, and index it, so that when someone infects your machine they can see your credit card numbers and pornography habits.'
It's so broken. I have a suspicion the AI teams are in some kind of an internal standoff with privacy/compliance teams, and Copilot simply never gets access to any tools or data.
Because if not that, then I don't know what. I can half-ass a better product with ChatGPT API and a PowerShell script, and I could've since GPT-4 was released; in fact the product can actually write itself, it's that simple to do a better job.
Being Microsoft, I'd guess it's less that and more Product Team A doesn't want to spend time building enabling features for Product Team B to look good.
The org chart joke, etc: https://www.reddit.com/r/ProgrammerHumor/comments/6jw33z/int...
> the product can actually write itself,
Used to be, that was just something people said. Turns out, we live in a future where that's literally true!
I have Copilot at work, it feels so useless sometimes. As an example, I had a report which I needed to make some batch edits to. I figured why not let the robot take a crack at it, so I clicked the Copilot button and spent a couple minutes describing what I needed changed.
Copilot tells me it can't edit my current document, but it can create a new one. I figured okay, Microsoft doesn't want to set it loose on the original, guess it makes sense that it requires a copy. So I said yes.
Nope. Instead of creating a copy of my document and editing it, it created an entirely new document which excised basically everything in the original report and replaced it with a very short summary - I'm talking 5000 words down to 500. All my tables and figures were gone, as was the standard report template my employer uses.
What utter garbage. Office productivity is a major use case for LLMs, and here the largest vendor of productivity software on the planet is happy to fuck it up.
The chatbot doesn't see the document structure; it sees tokens. When it generates a response, it's not editing; it's rewriting from scratch. Without a mechanism capable of precise diff/patch operations (like Cursor attempts with code), an LLM will always be a destructive tool for complex documents. Microsoft marketed this as an editor, but in reality, it's a summary generator
I’ve found it fairly useful in Excel. The suggestions to clean up data are pretty good and it’s spat out some quite gnarly formula on request
Compare this with Cursor or Supermaven. There AI acts as an extension of the cursor, understanding the code's AST. VS Copilot often feels like smart copy-paste that doesn't understand where it is Microsoft, owning VS Code and LSP, should have done this better than anyone, but they lost to startups on UX
> I truly cannot fathom how any of that made it past even a cursory review.
Maybe it's a fifth column group working to destroy Microsoft.
LOL
Since DECADES Outlook is _not_ capable of putting the first(!) just(!) useful datetime-stamp found in an email which says "calendar" or something in the text when pressing CTRL+R: This function creates a calendar entry from the current email with current receiver list, but default-date/time is alwsys set to the current even if recipients last ansewer was simply "lets do: 11am" - why does even this simple shit not work today? You dont need AI for this simple idea even, a robus / probability parser will do fine in 99% cases
Agree 100%. Don't train a 100-billion parameter neural transformer if you can solve it with a bash one-liner.
Use your human intelligence and save our planet!
I would just be happy to never have a ICS file every saved to my drive again. Whoever created the concept of a calendar invite saving a ICS file has created me hours of frustration.
It's so easy to ship completely broken AI features because you can't really unit test them and unit tests have been the main standard for whether code is working for a long time now.
The most successful AI companies (OpenAI, Anthropic, Cursor) are all dogfooding their products as far as I can tell, and I don't really see any other reliable way to make sure the AI feature you ship actually works.
Tests are called "evals" (evaluations) in the AI product development world. Basically you let humans review LLM output or feed it to another LLM with instructions how to evaluate it.
https://www.lennysnewsletter.com/p/beyond-vibe-checks-a-pms-...
Interesting, never really thought of it outside of this comment chain but I'm guessing approaches like this hurt the typical automated testing devs would do but seeing how this is MSFT (who already stopped having dedicated testing roles for a good while now, rip SDET roles) I can only imagine the quality culture is even worse for "AI" teams.
Yes. Because why would there ever be a problem with a devqaops team objectively assessing their own work's effectiveness?
Traditional Microsoft devs are used to deterministic tests: assert result == expected, whereas AI requires probabilistic evals and quality monitoring in prod. I think Microsoft simply lacks the LLM Ops culture right now to build a quality evaluation pipeline before release; they are testing everything on users
Microsoft: What? You want us to eat this slop? Are you crazy?!
50% of our code is being written by AI! Or at least, autocompleted by AI. And then our developers have to fix 50% of THAT code so that it does what they actually wanted to do in the first place. But boy, it sure produces a lot of words!
I've fooled around with some vibe coding on several LLM's like Claude, Gemini, and ChatGPT with some pretty decent results.
Since I have a full Copilot license at my corporate day gig, I figured I would try using Copilot for a basic static site. Nothing too hard, and something that's been handled easily with the other LLM's.
The prompt was pretty basic just to get something to start working with. "Build a four page template. With a home or index page, two pages of content and a contact page with a responsive slide out menu from the left hand side of the page."
It ran and put everything in a folder. I open the home page and everything was broken. I opened the files in VS Code and saw this:
And then this: I mean, if you can't even this right, I don't have much hope it can do anything more complicated. To say this was pretty sad is an understatement and clarified how far Microsoft is behind other LLM's.I guess this is code people working at MS would commit to a useless file in order to simulate work done with AI support.
There are cheap 1b open coding models that do a far better job.
Not a problem. We can call it HTML6 and make browsers interpret it correctly.
Kudos to you. I never use new buttons out of fear of something irreversible happening, like sending a random email or deleting something. I still feel uncomfortable with the Gmail UX, I would _never_ use a ”hello iz magic ai”-button.
Me: Can you access my inbox and Teams messages?
Copilot: Yep!
Me: Please find any items in my inbox or sent items indicating (a) that I have agreed to take on a task or (b) identifying me as the person responsible for a task, removing duplicates and any items that I have unambiguously replied to via email or Teams. Time window is preceding 7 days.
Copilot: Prints a list with, at best, 5% accuracy
I know some folks have the peculiar idea that search is dead in favor of AI, but if AI can't accurately find information, it is useless. As near as I can tell, Copilot finds 3-4 items (but rarely the SAME 3-4 items across runs) and calls it a day. It just seems like nobody is actually testing any of this stuff. Microsoft is actively destroying its credibility because it's offering a tool with a party trick but is utterly unreliable. I will, therefore, not rely on it.
It's a generalization problem. We can train LLMs that 'know' a lot of stuff in the global sense but the tasks that are interesting to people require the LLM to know a lot about you and your world in a very specific sense. The technical problem is that it's all corner cases and that's impossible to scale right now. No amount of context window is going to get you there either.
I respectfully disagree. The scenario posted by the GP is very clearly an "easy" task of basically sentiment analysis. The only thing it requires is a way to query your message history, then it just goes through them 1-by-1. I would expect any of the current SOTA agentic models to handle this with ease.
Sounds like Siri - unable to control much of anything on the iPhone outside of reading/sending text messages and setting alarms.
It also can initiate phone calls and play music.
And those are really the core use cases for AirPods, HomePods, and CarPlay, the contexts where a hands-free eyes-free interface is most useful.
And that’s ok. Those are core features of the phone that absolutely must work reliably and consistently. Far better to do a few important things really well than a hundred things execrably.
However the other day I asked the Gemini assistant on my phone to check the birthdays in my calendar, get all their dates, then make a graph of how many fall in each period with a 15-day moving average. It did everything as instructed including writing a python script to generate the graph, then discussed the results with me :)
I would expect Siri to be able to do anything on the iPhone that I can - change settings, report stats, kill/launch apps, etc.
It would be nice if it could control 3rd party apps too, like GMail, but being able to control the stuff that Apple themselves have built doesn't seem a lot to ask.
Apple is somewhat privacy focused, allegedly. I would imagine that letting an unpredictable LLM read whatever system data it wants is not something they want to partake in.
Maybe there are different concerns now that Apple are apparently going to be using Google LLMs for Siri, although I'd have thought that any access to system settings or other iOS functionality would be via tool calls with user opt-in.
However, the original Siri, obtained from SRI ("Siri" = SRI) was pre-LLMs, and there should have been no more concern over accessing system settings or controlling apps than things like reading/sending messages. For some reason Apply completely dropped the ball with Siri; initially it was expected that they would expand it into revenue generating areas like restaurant reservations etc, but then nothing. I'm not sure what documentation exists, but even today it's not clear what Siri is actually capable of.
We all know they Apple lack the capability, not a choice.
I asked Microsoft 365 Copilot to create a new word document for me (since they have hidden the link on office.com) and... it refused to do that.
Edit: Just tried again. It refused to do it. I mean WTF.
This is what move fast and break things looks like in a enterprise the size of microsoft.
It's mostly break things and little moving fast.
But the idea is that it's AI or death, so some broken buttons seems of less importances than the buttons itself being there, because the button working is a problem involving several teams, so no one is actually responsible, but the button being there is some team problem, and hell yeah they solved in the first sprint.
"Move fast and break things" is fine if you're a social networking site and breaking things means people can't get their racist memes or browse marketplace for twenty minutes until you push a change.
It's less fine if the things you're breaking are your core operating systems and the office suite that makes you most of your money and it takes you months to get the relevant teams aligned to push out a fix for the bad idea your execs pushed.
This is a problem at a wide array of tech companies. Everyone wants to Meta, everyone wants to be Google.
Guys, we're building B2B enterprise software. The most important things our clients care about is this hunk of junk working. Changing it is probably bad, actually, because the users are using it 8 hours a day and they don't want to deal with annoying popups about new features and UI churn for the sake of churn.
It should be noted that the things you are supposed to be allowed to break are YOUR things. When you start breaking MY things then we're going to have a real problem.
> Oh, well, and I actually also have a dedicated Copilot button on my new Lenovo laptop powered-by-Windows-11. And, guess what, it does exactly nothing! I can elect to either assign this button to 'Search', which opens a WebView2 to bing.com (ehhm, yeah, sure, thanks!) or to 'Custom', in which case it informs me that 'nothing' meets the hardware requirements to actually enable that.
How did you manage this? Probably some company-wide group policy saves you. It keeps starting copilot for me, drives me crazy.
> How did you manage this?
I did absolutely nothing special, other than running the latest-and-greatest Windows 11 Enterprise, which is what we put on most of our laptops without any customizations other than "require 2FA and some antivirus and firewalling" via Intune.
And I just went into our Azure admin portal, looking for any AI goodies to enable, and... there just doesn't seem to be anything there? And we have an Enterprise P2 subscription, which is usually where all the good stuff is, but, yeah...
Enterprise version?
Well, I happened to think many companies woild configured copilot to be disabled: claims to prevent corporate secrets from leaking out or refusal to install internal closed-net Microsoft CoPilot server.
I've always suspected Microsoft is a front for an ineffable cosmic evil that is trying to crush the human spirit into bewildered, abject despair.
How else do you explain Teams and the Hotmail UI?
You're thinking of Larry Ellison.
He's no eldritch horror, just a garden variety Bond villain. Volcanic island and everything.
That would make an interesting SCP...
That would certainly explain the Loop UI
> So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling?
Satya Nadella insists that Bing365Pilot has supercharged his productivity, but determining if he's high on his own supply or lying through his teeth is an exercise for the reader.
> Copilot consumes Nadella’s life outside the office as well. He likes podcasts, but instead of listening to them, he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond. At the office, he relies on Copilot to deliver summaries of messages he receives in Outlook and Teams and toggles among at least 10 custom agents from Copilot Studio. He views them as his AI chiefs of staff, delegating meeting prep, research and other tasks to the bots. “I’m an email typist,” Nadella jokes of his job, noting that Copilot is thankfully very good at triaging his messages.
https://www.bloomberg.com/news/features/2025-05-15/microsoft...
This is just him aping every other AI CEO. Every single one has to act like the agents are super-geniuses mere moments away from achieving the singularity like people can't try them out themselves and be disappointed. Some of it is "we think this will work soon so it's ok if we pretend like it's working now", but I think a lot is just needing to constantly shove hot air into the balloon before it pops.
On the other hand, if it's true, it explains a LOT about Microsoft's silly AI strategy.
Maybe he's like Saddam Hussein, being gaslit by his own minions, but in this case the minions are barely functioning word regurgitators...
> he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond.
What a dorky thing to do. Does the CEO have some concept he's living a life that precisely _zero_ of his customers do? Who would even think to do this?
> “I’m an email typist,” Nadella jokes of his job
Yea, I have actual work to do, perhaps you should familiarize yourself with this?
Are you kidding? You get to be the star guest of every podcast you listen to and everything you say is amazingly witty and insightful.
I know exactly the kind of people who would think to do this. Unfortunately. :)
> He likes podcasts, but instead of listening to them, he loads transcripts into the Copilot app on his iPhone so he can chat with the voice assistant about the content of an episode in the car on his commute to Redmond.
I remember reading that when it first came out and all I can think is: No, he doesn't like podcasts, if you like podcasts you listen to them.
That's like saying "He loves food, but instead of eating it he feeds it to an analyzer that tells him what elements were detected in it".
I have to assume it's all BS/lies because if that's a truthful statement (about podcasts and the other things) then I really question wtf they are doing over there. None of that sounds like "the future", it sounds like hell. I cannot imagine how shitty it would be to have all my emails/messages to the CEO being filtered through an AI and getting AI slop back in return.
> in the car on his commute to Redmond
This was funny to me, because he lives like 8 minutes away.
Maybe he has Microsoft Copilot Full-Self-Driving installed
Does that number include hours wasted in traffic jams?
The podcast thing reads like something he made up that sounds cool and futuristic on the surface, but doesn't actually make any sense. Maybe what's actually happening is he's just having the LLM give him the cliff's notes of the transcript, but that isn't interesting enough so he's making up some BS about having a conversation with the AI.
Nadella has to have his own custom agents. It isn't even possible for an enterprise like MSFT to not have custom agents that are still remotely useful.
So, his experience with Copilot agents != Average Customer's experience
Look, if the CEO of a five trillion dollar company investing hundreds of billions into AI can come up with some custom agents to handle every aspect of the CEO's work, surely your average Microsoft 365-subscribing corporation can do the same.
It's just another example of the rich being wildly out of touch. Yes, Beyonce has the same 24 hours in a day that the rest of us have, but she also has enough money to pay people to do every aspect of her life that isn't bringing her joy or wealth. Yes, AI can be used to streamline workflows or help you find signal in the noise so you can focus on the important things better, but if every company has to build that themselves then no company is going to see the value of spending a bunch of extra money on something that they can only get benefit from if they spend even more money.
If the 'AI agents' that Nadella is talking about were part of Copilot then sure, okay, I could see a benefit, but when people in this thread are saying that Outlook can't even tell you who is in a meeting then it certainly explains why Nadella doesn't understand the lack of value.
The only good feature about copilot in Microsoft 365 is that if you ask it to delete itself enough times eventually some other app will eventually be promoted to the most prominent when you first login.
There's AI in Teams to. I wanted to use it to recolect info from my chats but apparently it's unable to do so.
So it’s like Copilot in Excel, that can’t interact with or seemingly even see the contents of the spreadsheet that’s in focus.
Competition could as well in the past, I only switched from Borland compilers into Microsoft ones, after Borland decided to focus on enterprise customers.
And we all know the OEM deals preventing them to ship OS/2, BeOS, or whatever else they would like to.
The problem is that in modern computing landscape they cannot play that strategy, even when owning plenty of FOSS projects, there are plenty alternatives.
There are SO MANY Copilots
There's the Copilot cli, then we have the Copilot built in to VSCode, then we have the Copilot that's in Github, then we have th M365 Copilot and the Windows Copilot and maybe a few more.
All work very differently and have different skillsets.
Don't forget copilot in notepad.exe
> Like, I have a 'Copilot' button prominently displayed in my New Outlook on MacOS (the only platform where the app-with-that-designation is sort-of usable), and it's a dropdown menu, and it has... zero items when expanded.
I guess that's worse than the Gemini button in Google Sheets that asks me to subscribe to AI services. I have multiple times been in a sheet and thought "asking an LLM how to do this thing I want to do right here in this product would actually be great if it works", remembered there was an AI-looking button in the top right, clicked it, and nope'd out of the subscription.
I just want to know if it works or not before I buy it.
> I actually also have a dedicated Copilot button on my new Lenovo laptop powered-by-Windows-11. And, guess what, it does exactly nothing!
A physical button? So it’s like a mechanical placebo? If you have AI FOMO you can keep pressing it to calm the anxiety.
What do you mean 0 items? When you expand it gives you options to write a message or interact with search. Seems like its a you problem.
On new-outlook-on-Windows there’s an option to disable the copilot button. Wanna take a guess what that does?
Honestly, I can't tell if you're being sarcastic and it actually does disable the button, or if you're being jaded and cynical and it just hides it for six hours then it comes back. Either option seems equally likely at this point.
The button stays in the bar, only greyed out. Its dropdown is still functional, allowing you to turn it back on.
The other thing is that they're FORCING it on the users. Its invasive and creepy and assaulty and it has multiplied my loathing of Microsoft infinitely.
Someone in Microsoft needs to watch a lecture on affirmative consent.
Don't get me wrong, I enjoy playing with AI, both local and online. If you tell me its available if I want it, I might come dabble a bit.
But cram AI into every facet of my machine? That's a 1 way ticket to never being installed on any system I ever own in the future ever again.
Companies should treat AI like a madam treats the workers in a whorehouse. You don't make them go door to door. You let the johns come to you.
> So, my question to anyone in the Microsoft C-suite: have you ever tried to, like, actually use, like anything that you're selling? Because if you would have, the failings would have been obvious, right? Right??
They all use Macs lol.
At what point does Charlie Brown not kick at the ball Lucy offers?
It's been 19 years since "Microsoft Re-Designs the iPod Packaging". [0] Is the disconnect displayed in this message thread that there's always 10,000 new people discovering a fact? [1]
0. https://www.youtube.com/watch?v=EUXnJraKM3k
1. https://xkcd.com/1053/
I bet c-suite uses Mac
I bet they use iphones and leave computers to their underlings/assistants.
"Compute? Think with own brain? Our servants do that for us."
Yeah, and I also bet their Outlook 'Copilot' button has more-than-zero options in its dropdown menu.
But I'd actually love to know how to achieve that, and so far Microsoft AI is awfully silent on the subject...
Nah, they probably just have Copilot as a bullet point on a slide, count that as "using AI", and are psyched for their next board meeting.
And everything aimed at developers assumes you're using Unix.
Hard disagree: there is a whole universe of Windows-based developers. But even for them, the best offer seems to be a frequently-updated but still-entirely-underwhelming Visual Studio Enterprise plug-in that after 6 months (or so) can't even show proposed changes in response to a prompt without destroying surrounding code...
Hardly, there is this misconception in FOSS circles that software development outside UNIX walls is an ilusion.
As a .NET developer who actually likes some Microsoft products, I can say this: the Copilot series is the worst thing they've shipped since Internet Explorer—and honestly, it might overtake it. The sad part is they had a huge head start before competitors gained access to powerful models, yet this is what we got.
If you haven’t seen how bad it is, here’s one example: Copilot Terminal. In theory, it should help you with terminal commands. Sounds great. In practice, it installs a chat panel on the right side of your terminal that has zero integration with the terminal itself. It can’t read what’s written, it can’t send commands, it has no context, and the model response time is awful. What’s the point of a “terminal assistant” that can’t actually assist the terminal?
This lack of real integration is basically the core design of most Copilot products. If you’ve been lucky enough to avoid them, good for you. If your company forces you to use them because they’re bundled with a Microsoft license, I genuinely feel your pain.
Works good in Teams. Summarises meetings, collects action items. It’s pretty great actually.
I don’t know how many forests we are burning to have a digital secretary, but surely the environment can take one more for the team?
"Summarises meetings, collects action items"
Yes, but...does it? How do you know it isn't missing key points raised in the meeting? Not collecting actionable items?
Summarising seems like the absolute lamest thing an LLM could do for me. Like I want a Reader's Digest of my life, written but a word guessing machine.
This is arguably the only use case where a chat interface (or just text) actually works, because it's an information compression task. But building a $50 billion business strategy just on the fact that AI can recap a meeting is risky. If this is the only feature that works well, then the business ROI is questionable
“It’s just like your toaster”
> lack of real integration is basically the core design of most Copilot products
I would wager a month's wages that this is the doing of some internal Security Review, wherein a bunch of out-of-touchers decided that the customers will want to prefer to be Safe and Secure instead of getting some actual value from integrating copilot into shell workflows.
Meanwhile people are yolo'ing it with various janky DIY wires and duct-tape githobbits that mash together whatever open weights model and user-level access to the system (or worse).
> a bunch of out-of-touchers decided that the customers will want to prefer to be Safe
You mean the other way around, right? Because what could possibly go wrong when we let a language model hallucinate its way through which terminal command rhymes best with your prompt according to that SO comment from training data.
i mentioned this upthread but an LLM with enough access to be fully integrated into all apps/services/files in an enterprise managed workstation sounds like privilege escalation attacks just waiting to happen.
Wow a thing I really want is an AI I didn't install to read all my terminal commands and ship them off to MS for analysis.
Thank god for this security team.
You'd be hearing a few more hair raising failure stories if they hadn't done that. And possibly a few big customers institute Win11 or Copilot site wide bans. Or just straight up going out of business. In businesses other than software, it's possible to be legally liable for mistakes.
> This lack of real integration is basically the core design of most Copilot products.
I think they're scared of the very real security issues with LLMs that may be unsolvable. It's not wise to give an llm free reign, at best maybe across your local computer but to be fully integrated into every application and every file it needs root. That would be the front door to many privilege escalation incidents in an enterprise managed laptop/desktop.
> worst thing they've shipped since Internet Explorer
IE6 was a really good browser when it shipped in 2001, especially compared to Netscape 6.
The problem was it didn't improve for the next 5 years.
When was the last time Microsoft had a unified vision that was focused on building an amazing line of products that integrated well with each other?
I can only think of short snippets in history where they moved in that direction for maybe a year or two & then went scatterbrain.
Microsoft has benefited from a monopoly in the enterprise and has never been forced to innovate from a product perspective. See Slack/Teams as a case study of how they have operated when even slightly pushed.
* Edit - .NET, C#, TypeScript teams are an exception to the above. Highly underrated. Amazing talent there. Not sure who all gets credit. Anders & Mads for sure though.
I'd say the late '80s - early '90s when Microsoft was building the early versions of Microsoft Office. Integration among productivity apps was one of the key points of competition among all of the office suites of that era.
There were other huge coordinated efforts like the TwC initiative and the Windows 10 refactoring but those were invisible to end users.
Not really.
You are right if talking about the efforts improving C# and CLR, taking the lessons out of Midori, Blazor and Aspire.
Regarding F#, VB, and C++/CLI, the poor souls are on a lifeline that makes CLR nowadays stand out for C# Language Runtime, instead of the original Common.
And the chaos that reigns on Windows Forms, WPF, WinUI, MAUI certainly isn't helping.
Finally they are constrained in what they can put into C# DevKit as means to not canibalize Visual Studio and Windows sales, it can only be good enough, not great and feature parity.
Also regardless of the reasoning behind it, having TypeScript rewriten into C# instead of Go would have been a great opportunity to make C# more relevant outside Microsoft shops, instead anyone looking to contribute to TypeScript compiler will be learning Go instead.
Useless Clippy all over again. At least Clippy stayed in his lane and could be turned off.
Try to think about it like a bunch of merchants running to line up in front of you as fast as possible with anything that resembles a product you might buy. The only thing they were successful at thus far was recognizing that something needs to be in your terminal, and they ran as fast as possible to your terminal with anything to beat others to the front of the line. I suppose they are doing this with all their efforts. A universe where a C-exec said in no uncertain terms, "get anything out there", is the very universe we're in.
Apple is the only merchant not running to line up with anything at the moment.
IMHO, one company needs to make the bold move and make a fork of their OS that is AI native with AI native apps/workflow and phase out the old paradigm. It'll have to be two product lines, but I think the new OS will have uptake like we've never seen before.
There's a problem with this theory.
AI works. It's actually useful. Since GPT-4, tool calling capability is good enough. It's trivial to do a better job than Copilot on any task using any current model of the major LLM providers. I'm not talking API, even with basic chat frontend, regular users easily beat Copilot by simply copy-pasting between Word/Excel and the chat frontend.
If a twelve year old can one-shot a better product for any given use case than Microsoft Copilot, then it's not just "merchants running to line up in front of you", something more basic must be broken.
The crazy thing is Microsoft was so early with an AI product but was burned pretty badly when it instantly turned into a Nazi. Funny how the constant complaint on Twitter is how the modern AI agents are too "woke" and how Elon has to keep fighting his own AI model to conform with his viewpoints.
WTH Are you talking about?
https://en.wikipedia.org/wiki/Tay_(chatbot)
Context “ https://www.cbsnews.com/news/microsoft-shuts-down-ai-chatbot... “
I think the biggest revelation of the last 3 years or so is that Microsoft does not have either the will or the talent (or both) to effectively execute anymore. Everything it currently stands on is a legacy product with roots in the Ballmer or Gates eras. They owe their Azure footprint and "success" today to Ballmer.
Their inability to produce anything useful with Copilot is the largest example of this, but there are others. They are getting lapped by a ~300 person software company in the race to consumer-ize an x86 PC a into turnkey gaming platform, even with $100 billion in game studios and owning the API that every major game is developed against. Their footprint in education is gone, completely replaced by Google who not only produced an operating system that could be effectively run and managed on commodity hardware, but also developed the centralized functions for school administrations to use to manage classrooms at scale.
The consumer situation for Microsoft right now might be even worse than it was when Nadella took over.
Another example: the complete surrender of the Xbox platform. They spent the last 10 years buying up game studios and then closing them, failing to develop any good exclusive titles, and letting Sony eat their lunch. Now they are officially giving up and will just let Sony own the console space altogether.
Xbox was THE gaming console 20 years ago. Playstation was always a contender but Xbox Live was synonymous with the online console gaming experience. Halo was an untouchable juggernaut of a series for the first 3 titles.
It's mind boggling that Microsoft just let all that die without a fight. Worse, they seem to have actively shot themselves in the foot and then given up.
Playstation/Sony always a contender? They were the clear leader. The PS2 is still the best selling console of all time. 160M units to Xbox's 24M. Xbox360 caught up but the PS3 still sold more units. PS4 sold twice as many as Xbox One. The console wars has been mostly Sony vs Nintendo for a long time.
> Xbox was THE gaming console 20 years ago
Maybe in the US, but not overall.
Xbox was never the leader. They were always the challenger from the begging, but it looks like things are not getting better.
In US, XBox was never first in the rest of the world.
I don't have data to support it and I know AD was a gamechanger but I feel like Microsoft pumping free software to educational institutions was the main driver for it's adoption in the corporate world. When every person who kind of knows how to use the computer, can use MS windows and office and has to be retrained to use anything else, it's a no brainer to just give users what they know.
Don't you think this is why alarm bells should be going off right now about Google's complete domination of primary school systems? Google Classroom and Chromebooks are everywhere. Google Docs is the lingua franca, not Microsoft Word
The main benefit of Google Docs is that, at least, Google is a better steward. Not good, but better.
And also, their software is exceedingly simple, which is great for breaking vendor lock-in. Microsoft got away with murder because none of their shit works like it should and every single product they make is the "black sheep" in the category it's in.
We've been hearing empty punditry like that for the last several decades all while they've been a poster child for "number go up": https://www.macrotrends.net/stocks/charts/MSFT/microsoft/rev...
These predictions about the decline of Microsoft are like the Year of Linux on the Desktop; neither is going to happen anytime soon. Y'all can start predicting doom when there's a multi-year trend of declining revenue for MSFT and then maybe there's something to discuss.
Just as Proton running Windows binaries is "Linux Gaming", WSL is the "Year of Linux on the Desktop".
By that time, it will already be too late. I'd argue it's probably already too late... dead man walking kind of situation. Definitely true for consumers, it might survive out of inertia in the corporate environment for decades.
Yes, I miss the Windows developer experience from Balmer era, "Developers, Developers, Developers" certainly doesn't apply to Windows native development experience.
The only thing good from Satya era on Windows development experience, is the improved terminal, and that now I don't have to install VMWare Workstation any longer.
Three years? Try 20.
There was a honeymoon period after Nadella took the reigns where it looked like they were putting out good products - Edge before enshittification, VS Code, buying GitHub.
Also WSL.
Microsoft's entire business model for decades has been to shove shoddy products down people's throats. And somehow, they have figured out how to do it too, because otherwise Teams wouldn't be used by anyone.
Microsoft’s best pitch (and Google benefits from this too) is that contracts are annoying and take forever to execute. If you can sign a deal for Outlook and Teams it’s so much easier than separate contracts for Outlook and Slack. You’ll get very far with that logic alone.
Most companies I’ve been at that use Teams over Slack is not “We can’t get contract for Slack” but “We have Teams included, why would we pay for Slack?” - Accountant
I guess Microsoft lost this battle, at least at some companies, because I'm now at one that uses Slack and Google, with no dependency on Microsoft Office.
Microsoft won this battle if you check the numbers. Last I saw it was 85% Microsoft vs 15% Google which seems right with my experience. Current company is Google Worksapce while last 3 were Office365.
Teams was clearly never meant to be used as a standalone product. It was a pretty effective defense play. Block other competitors from gaining a foothold.
Thats the whole point. The only people using Teams are the ones who are already committed to Microsoft 365. Companies on GSuite mostly use Slack, I doubt there is a single one using Teams.
Its all about Excel. It really is the best spreadsheet, and everyone knows how to use it. But that comes in an Office bundle that includes Teams. And that is why we must suffer.
Was certainly the case in the early years of Google Sheets. For me, the gap is entirely closed. I'm willing to believe that Excel still has the better platform for extreme power users but I've done some pretty slick stuff with Google Sheets and that was four or five years ago. It must be even better today (though I'm not currenlty doing much with spreadsheets).
Has google sheets filled out the lambda helper functions yet? If so that could narrow the gap.
Passing around vba based xlsm is really awful, so if google sheets has lambdas they can probably get a lead with google sheets queey language over filter.
Groupby and pivotby are the new excel alternative, but if they filled out lambdas, then does that keave VBA and power query as the only reasons for Excel?
Not sure what lambda helpers are. But you can write Google AppScript which is just JS and do some pretty cool stuff, including define custom functions and fully integrate with Google APIs. I've used to to send emails, or create calendar events for example.
Byrow, bycol, map, reduce, scan, makearray, and most recently groupby and pivotby.
Basically functions that take a lambda argument and apply it over some range.
I agree with it, but it's a wild world we live in when the best spreadsheet has default behaviors which will fuck your data pasted into it when you're not paying attention.
Yeah, its also awesome when you open a CSV with long numbers like tracking numbers or IMEIs and they get converted to scientific notation and lose "precision" when saved in Excel format...
Exactly what I'm talking about. I'm sure there's other gremlins in there, but this one bites me the most often, and the hardest.
The one failure mode is so fucking insidious where it zeros out your last 2 digits, but otherwise it looks like it pasted OK!
I think the expectation of software quality in consumers has completely evaporated, and so people expect appallingly bad software and think it's normal. This isn't just the case for Microsoft - look at the baby-ification of macOS UI (let's make it like a children's iPad), any Google software offering (very poor performance) and the majority acceptance of web-browser-in-a-window-is-a-desktop-application software development. They accept things like Slack and Teams (awful software) but also put up with Skype getting worse in the decades before.
Office 2003 and prior were quite good, but then people think Google Docs is somehow equivalent to the functionality of Word.
Admittedly, Active Directory was afflicted by impossibly tiny windows for the management tabs, but the functionality worked (and you could write your own extensions to the LDAP tree and COM-based UI interface for them) as proven by the rehashing of part of the functionality into Google's "organisational" offerings (sign into Chrome and receive restrictions from your organisational (company) overlord, the new Group Policy).
It's a real pity. If we showed software today via a time machine to ourselves 25+ years ago, we'd be shocked at how slow and ineffectual it was and deeply distressed that this was the norm worldwide.
I find it fascinating how they are able to sell their crap software.
It's the oldest trick in the IT book - focus on the buyer persona and ignore the user persona.
I'm guessing the broader demographic of users simply don't think the software is crap. My buddy working in water transportation was just raving about Teams to me the other day. His praise basically boiled down to being integrated with his organization, providing him easy access to his department-resources. I suppose it does serve my buddy well.
Could be because I use a Mac, and I do most of my work chatting in Slack, but I don't hate Teams. It has my calendar, it has chat that is shitty but good enough during meetings (although finding which room you have unread messages in can be confusing), and it has video meetings with transcripts and great Exchange integration.
Windows 11, on the other hand, offends me.
Ditto. The more interesting part is how many people will defend it. Presumably some mix of post-purchase rationalisation and inherited assumptions about what's "standard" even when those assumptions stopped being true ages ago.
I find it infuriating, but that's how the system's supposed to work. It's the definition of a monopoly and they're in the extraction phase. When there's no competition (and eventually there's always going to be a winner) you don't need to make good products anymore.
They've successfully indoctrinated whole generations to use Windows/Office. Here in Brazil using a computer was (probably still is) synonymous to using Windows/Office. Everyone had their pirated version of Windows and many don't even know that alternatives exist. When those people open companies they'll use what they know.
Software companies have to build for the most popular OSes and most can't justify anything else. Which then means most software only works on Windows and people can't leave it even if there are better alternatives (see Adobe). Finally, any non-closed computer comes with Windows so the cycle continues forever.
My theory is that they deliberately make Windows so shit to filter out anyone with taste. Once you have a userbase of people who don't know better, you can sell them any old crap. Like Teams.
Reminds of an research article from Microsoft!. It detailed on why scam emails about `Nigerian prince` are so obviously dumb. The reasoning being it specifically need to target only those who can fall for it. Anything more sophisticated and they would get people who wouldn't fall for scam in subsequent communication.
That's exactly what I was thinking of!
Except it makes no sense because as a scammer your goal is to get as many people as possible in contact with you so that you can scam them. You can only score on the goals you attempt so cutting out any person, no matter the reason, is illogical.
You’re assuming that there’s no cost involved in moving a potential victim through the pipeline. I’m sure AI has changed the game, but the general idea was that beyond the initial blast of spam you would have someone actually responding to those who fell for it. Putting in signals that it was a scam filtered out individuals who would waste scammer time because they would eventually figure it out before falling victim. By selecting for people who literally can’t pick up on obvious signs of a scam, you save yourself a lot of time and energy.
Keep following through the logic... You manage to hook someone who absolutely knows you're a scammer, and they keep responding to you taking up precious time you could be spending with someone who is actually likely to give you money. So, what is the upside to getting a response from someone who is never ever going to give you anything?
Occam's razor says they are just bad at English grammar because it isn't their native language/dialect and their education probably wasn't that great.
This is easily demonstrable by conversing with the scammers and noting that their actual English ability is the same as exists in the initial letter. Even when they have no chance of the scam succeeding and have been outed they write the same way. You can see plenty of evidence here:
* https://www.419eater.com/html/letters.php
Occam's razor says the sun orbits the earth, everybody dies from Sudden Unexplained Death Syndrome, and the correct way to spell Occam's razor is Okams Raza (in all languages, because lavishes other than English are difficult).
It's literally a platitude. It's like the saying 'when the going gets tough, the tough get going': it's reallyemorable and descriptive and is maybe a good guideline in many situations.
But using it to evaluate the tensile strength of various metals according to their velocity would be wild, because it had never pretended to be anything like a rule. It's not like theory of gravity or 'I before e except after c', which are based on actual analysis and results.
Legit assuming that everything is as simple as it can be, that the most obvious idea to occur to any untrained observer is the most accurate, is literally a guaranteed way to go though life without understanding anything, at all. Using it to argue with people who appear to obvious what they're talking about (and there are so, so many undisputed studies on the exact reasons scammers do what they do: it's too filter people or. There is no debate, academically) is a pretty slippery slope to 'anybody who doesn't think and act exactly like me is lying, because no reasons or facts exist unless I personally hold or after with them', and it's definitely a thought process worth challenging.
Although to be fair, its best application might be re. online arguments that you don't really care that much about. So if you just meant that the previous poster had given a reason and you were going with that because it's easier, my bad.
> And somehow, they have figured out how to do it too
You say this like it was a mystery to start with. When you own 90+% of the user base, you can create trends with any changes implemented
>"Microsoft's entire business model for decades has been to shove shoddy products down people's throats."
I remember this one. In the 90s MS reps would come to our company and sing about how their Visual Basic was superior to Delphi. When pointed to countless features that proved the opposite all they were able to say is that the MS has bigger dick.
Their recommendation was to have 2 developers instead of one we had. One would code GUI / front end in Visual Basic and the other write DLLs that would do all the meat.
I once worked at a company that wrote their UI in Delphi but then wrote their logic in C++ using boatloads of COM and MFC (for non-gui elements; the std library with containers was "new" at the time) so they ended up with 20 years of development baggage to talk to SQL Server over ODBC whilst implementing their logic on a repeated basis in C++ (COM loves you implementing interface functions you can't remove), calling from Delphi.
They were trying to port it to a "modern" system and modern compiler so had millions of lines of code to fix, and their UI was MFC-based (so another shot in the foot).
"Fun" times.
Microsoft's entire business model has been tying. Countless millions are forced to use Copilot because their IT department has contracts with Microsoft, and those same contracts are why they use Office, Teams, and so on. Their developers use Visual Studio, deploy to Azure, and run it all against SQL Server. Their email comes from Exchange.
It has been an incredibly lucrative strategy. We all herald some CEO's prowess in growing revenue when they've been doing the same playbook for decades now, and have been running on the inertia of Windows dominance on the desktop. Every new entrant is pushed out through countless incredibly lazy IT departments that just adopt whatever Microsoft shits out.
It's actually surprising that the one and only area where this really failed was as they tried to lever tying to the mobile market. A couple of missteps along the way are the only reason every office drone isn't rocking their Lumia ExchangeLive! CoDevice.
People here are mostly too young to remember but the original Microsoft business model was this:
Find a software market currently addressed by high price products; create a reasonably good product for that market; sell it for significantly less than the incumbent. Sell much higher volume of said product than the incumbent, thereby make much more profit. Repeat/rinse.
The Windows lock-in, embrace extend etc came after this. You can't lock in customers if they didn't already willingly buy your product.
No the original Microsoft business model was to get the incumbent (IBM) to bundle your product (DOS, bought from someone else) onto their product so that you had a near-monopoly, then use that to sell your other software onto that, occasionally making technical changes to make it difficult for your competitors.
>the original Microsoft business model
From 1981
>Microsoft, which needed an operating system for the IBM Personal Computer,[9][10] hired Tim Paterson in May 1981 and bought 86-DOS 1.10 for US$25,000 that July
IMO, it's time for leadership change at Microsoft. Satya revitalized the company but now it needs a Product person that knows how to rebuild the quality of it's products.
Perhaps they could replace the leader with a LLM trained exclusively on the MSDN documentation. It'd be technical-focused.
So... Clippy?
I would think that if they actually spent the time and money fixing the core functionality of their core products (like Windows and Office) that they might have a much easier time promoting things like Copilot. Instead they leave their users wondering why they're so hell-bent on shoehorning AI into a Start menu that takes whole seconds longer to open than it should or into Windows Search that regularly fails to find installed programs or local files.
Microsoft is a public company. That means their primary product is not products or services, it's their stock. Selling products & services can be an advertisement for their stock, but there are other methods of convincing people to buy their stock, too. Currently the stock market only wants stocks that have "AI" associated with them. It doesn't matter whether users like it or not, because having a viable business is not what the stock market is currently focused on. So, Microsoft is doing what they need to do to sell their primary product: shove AI into everything.
Are you saying they would rather double stock price than double revenue?
If you're someone who owns Microsoft, what option would you prefer?
1. Stock price remains the same but revenue doubles.
2. Revenue stays the same, but stock price doubles.
Assuming all else equal, and recognizing that this is absolutely a simplification, but if these were the two choices then it seems a no brainer that you'd go with option 2. Revenue is a means of increasing stock price.
That’s like asking whether you’d rather have a pizza with the same diameter but twice the area, or a pizza with twice the diameter but the same area. It makes no sense.
Stock prices were, are, and will always be rough approximations of the NPV of future profits. It’s not perfect, of course, but it’s roughly true.
Doubling revenue in any remotely sustainable way will have way more than a 2x impact on stock price because of exponential growth. So yeah, as a stockhodler, you’d rather double revenue with flat stock price because you’d buy the crap out of the stock when you realize the market has not factored revenue growth in to pricing.
Imagining that public companies care about stock price more than revenue is literally like saying a hungry customer care more about pizza radius than area.
I guess it depends on what kind of investor you are.
If you're holding MS stock long-term, and you plan to gradually shift away from equities as you near retirement and then gradually liquidate your holdings to fund your retirement, juicing the stock in the short term does nothing for you.
If you're holding short-term, then you also need to sell the stock after it gets juiced, so that you can move your capital to not-yet-juiced stocks.
Missed a point above - that for said short-term investor... that strategy doesn't actually work, since a "sell high buy low" strategy on individual stocks is outperformed by just holding ETFs long-term.
So really, which investors does short-term stock juicing benefit? Insider traders, I guess.
A long term investor would prefer 1. Most likely option 2 would shortly be followed by halving of the stock price. Option 1 lets someone buy shares in a company that is greatly undervalued and will lead to long terms gains.
I just think you are the other person who replied to me are making the same mistake, which is mixing the means of accomplishing a goal with the goal itself. This is a mistake I see a lot (especially in software development), where people get too attached to a specific means or method that they end up confusing the method itself with the actual thing that needs to be accomplished.
The goal for an investor is captured in the stock price itself. In programming terms, a corporation is a function whose output is its stock price/market cap, and revenue is but one of a host of inputs into that corporation that determines what the stock price is. Other inputs can be operating expenses, whether dividends are issued, future prospects for the company such as entering new markets etc etc... and you can have beliefs about how those various inputs affect the output or how these inputs change the output over time (short term vs. long term), that's perfectly fine... but when push comes to shove, the goal is not the revenue, it's not entering a new market, it's not reducing operating expenses... the goal is increasing the stock price (well technically the market cap).
You should look up the concept of value investing. If you can buy shares in an undervalued company do so.
That's precisely the argument I'm making. A company's stock price can be undervalued, which exactly means that the stock price can increase without any change in the company's revenue or profitability. The stock price can increase strictly because the actual value of the company has not yet been fully realized without any material change necessary on the part of the company.
As an investor, that's the ultimate goal of your investment.
Sure. Decision makers are paid in stock price, not revenue. They would rather do whatever increases the stock price the most, with the least effort/expense.
If only there was some correlation between stock price and revenue…
Yyes, and prefer it to doubling profit or cashflow etc.
I have not looked at MS in particular, but generally that is what the remuneration of the people at the top of most public companies is most strongly linked to.
Yes! Not least because doubling revenue doesn't cause the stock to double.
It's basically the reason this bubble not only exists but has a chance not to pop : there is so much stock value in it that the big tech all want to keep feeding it, and they're sitting on so much cashflow they can afford to do it
It's absurd, but that's where it is. And a company like OpenAI basically hangs on it, because they have obligation almost ten time their revenue and the only way this does not deflate quickly is if others keep feeding it cash.
> Microsoft is a public company. That means their primary product is not products or services, it's their stock.
Yeah that's a great business idea, ask Boeing how that's going
Current CEO:
> Boeing spent about $300,000 to help Ortberg move to Seattle. His decision comes more than two decades after Boeing leadership decided to move company headquarters out of Seattle. Ortberg received about $18 million for the months he was the CEO in 2024.
https://en.wikipedia.org/wiki/Robert_Ortberg
Previous CEO:
> In 2022, Calhoun received $22.5 million from Boeing. Most of his 2022 compensation was in the form of estimated value of stock and option awards. He received the same $1.4 million salary as in 2021. ... In February 2023, Boeing awarded Calhoun an incentive of about $5.29 million in restricted stock units to "induce him to stay throughout the company's recovery". In March 2023, Boeing announced Calhoun was being given shares worth $15 million that will vest in installments over three years.
https://en.wikipedia.org/wiki/Dave_Calhoun
Seems to be going all right?
Not referring to the CEO. I was referring to short term stock price and shareholders being your only customers versus the people buying your product
Oh yeah it's absolutely terrible for the business and its customers.
Works for Tesla, the absolute poster child for this kind of behavior. They can lose double digit percentage of market share in key places and still give the CEO a trillion dollars.
Maybe the stock market is not a good system to organize ones economy around then?
I've been thinking about this recently. The centrality of the stock market, while historically a great tool to allocate resources efficiently, might actually be a big weakness for the USA today. A capable adversary, like China, can kill entire strategic sectors in the US using the stock market. If they undercut the US companies and are willing to accept low returns on their investments, then the respective USA competition will be driven out of business by their investors, because there will be other sectors to invest in, with higher RoI. Do this at various points in strategic value chains, and over a decade or so it might kill entire verticals in strategic sectors, leaving the US economy vulnerable to any kinds of shocks.
As someone who is essentially financially illiterate, what does this mean, "allocate resources efficiently?" Nobody's investing in companies that promise to cure world hunger or alleviate childhood suffering. They're investing in technologies that can extract the most wealth from the population, regardless of externalities. Is that desirable?
Then again, I can't fathom what people would be doing with their money if the stock market weren't there. I imagine they might naturally wind up with some sort of...stock market.
The operating principle here being that prices are units of information, which in aggregate reveal some combination of market demand, present supply, production costs, etc. All else being equal, an investor who's looking to put an investment into a new business will try to find the best rate of return. The existence of a relatively higher profit margin for an industry suggests an unmet market need, and then directs the flow of capital into it (if you expect that for every $1 you invest into a roofing nail plant will return $1.25 over the next year vs a $2 return from a new insulin plant, more new cash will flow into the insulin plant, more insulin gets made, and if the investor guessed right about the demand for it, they turn a profit). In a sentence, money flows towards trying to give people what we think they want more of.
The theory posited above is that you could try to manipulate these signals as a sort of economic warfare. If you expect that every dollar you put into our aforementioned roofing nail factory will get you minuscule or negative return, nobody's going to want to invest in building/expanding nail factories, and they'll put their cash somewhere it can grow instead. This is all well and good so long as you've got happy trading relationships with people who can sell you nails, but if one day the nails stop coming--you've got a supply chain shock until you either open new factories or find someone else willing to sell nails to you. The theory here being that if you had a LOT of goods that became tied up in a single point of failure--someone forcing that failure could create a great deal of internal instability to be exploited for geopolitical ends.
That's what's meant by efficiency, it's allocating it to the place that has the highest return on investment.
As you point out, in practice what's efficient is what can capture the highest return, not necessarily the highest return per se. If say investing in education had high returns society wide but those returns couldn't be captured, that's not an efficient use of private capital.
As somebody doesn't consider himself a capitalist, wouldn't it be fair to say it is "the most efficient" in precisely one thing: capital reproducing itself?
And if so, why is that necessarily a good thing? Why should that be our goal as society as opposed to things like minimizing child mortality, increasing literacy rates, making sure we don't have a ton of our fellow humans living on the street in misery etc etc - things that make the lives of our fellow humans better? Why is capital growth the metric we have chosen to optimize for? Surely there's better things to optimize for?
Excuse the polemic, but infinite growth with no regard for anything else is the ideology of a cancer cell - and to me that is increasingly what it feels like when we are wasting all these resources on a dying planet just to make numbers go up.
Ultimately that money is made by people choosing to spend their money on something, because it helps them, because they like it, because they need it for whatever reason (real or imagined). That's what grounds the financial markets: eventually someone is buying a thing because they want the thing, and all the rest of it is basically just figuring out who can make the thing, how many people want the thing and how badly, and whether the stuff used to make the thing could make a different thing that people want more. Financial markets can depart from that reality for a while, but mainly because of a collective belief in some falsehood about the above (everyone really badly wants AI, right?).
Number go up infinitely is due to inflation and that's basically just an incentive to not hoard cash indefinitely, and instead use it for something useful. But the only thing that uses up is numbers. Everything else is because people, on average, want more stuff and are willing and able to work hard to get it.
(Of course, this generally means that the markets chase the desires of those who have something valuable enough. People who don't will be marginalised by this mechanism, for sure. And of course there's lots of opportunity for people to steal or abuse powerful positions in the market to the detriment of others. Which is why a free market is not the be all and end all of organising a society, and other organisational structures exist to regulate it and to allocate resources in a less transactional manner)
But isn't that counter to the very article we're commenting on? Everyone is shoving half baked AI junk into everything because that's what makes number go up on the stock market, but I'm pretty sure that's not actually what most people would want those resources to be used for.
I'd posit that markets are completely detached from the real world and are more of a speculative/religious element than an indicator of any ground truths.
Edit: I just realized I missed a sentence of yours where you kinda spoke to this. I still believe that this is more of a rule than an exception - there is nothing inherently tying markets and reality together - they're mostly about people making bets on what the next big hype is; not on what is actually useful to anyone.
Most people or most people with money? Ultimately it's the people making investment decisions you need to convince in order to get the money from them. Then the reality check is whether you can give them that money back and more. Those investors should in theory be self-interested in making sure that this is actually useful, because if they don't they'll lose their money, but in practice they are not superhuman and prone to fads and echo-chambers (especially because it's a relatively smaller group that can move a lot of money around, and short-term investing rewards running with the crowd to some degree), but there's not a single group of people that would not fall victim to poor decisions in this regard (whether you're imagining a centrally managed economy, the electorate, or some hypothetical benevolent dictator).
Optimizing for capital returns is a simplification of the real world, where it allows for comparing whether it makes more sense to put one's money into opportunity A or B.
There's a lot that's not captured by solely looking at dollars, like the examples that you bring up, such as quality of life, human welfare, and so on.
If you care about minimizing child mortality, increasing literacy, pulling people up out of poverty, you should be a capitalist, as it's empirically the best way to meet those goals. This seems to be a hard thing for many to understand or accept because it is largely a second order effect, the capitalist primarily concerned with their own personal gain but winds up improving the lives of others as a side effect.
This is the essence of Adam Smith's often misunderstood invisible hand metaphor. Of the individual he observed: "By pursuing his own interest, he frequently promotes that of the society more effectually than when he really intends to promote it." Second order effects stack up and improve quality of life for more people better than trying to do so explicitly.
Multiplying capital creates abundance and that abundance allows for improved standards of living for and the means to spend excess resources in support of charitable endeavors. Growth is good because it means more abundance and opportunity. I would argue that pursuit of growth is not an ideology but a force of nature. Life is opportunistic and will expand to wherever there is fertile conditions, and often adapt even when they are not. We are part of nature and understand this intuitively, seeking growth opportunities. As an example, one is better off being part of a growing company (more wages and opportunities) than one that is stagnant or declining (fighting for scraps and survival).
>If you care about minimizing child mortality, increasing literacy, pulling people up out of poverty, you should be a capitalist, as it's empirically the best way to meet those goals If you look at it empirically, the majority of people brought out of poverty (and I suspect the other metrics but am not as familiar with them) in the past few decades have been in China as the result of deliberate policies by the CPC.
Then why are all of these metrics worse in the US than even in notoriously poor communist Cuba?
We’re already there when it comes to having the industrial base necessary to fight a protracted conventional war with china. Which leaves a large ? over the US dominance over the pacific.
Y-yeah. HYPOTHETICALLY, this is something an adversary to the USA might attempt to do, and it would really kneecap the US if they were successful.
But would only happen if USA decided to totally financialize all sectors of its economy and make a small set of oligarchic corporations THE load-bearing element of its strategic capacity, leading us to chase market returns even if those returns totally kneecapped our ability to build anything at all of actual value.
Good thing we haven't done that!
> while historically a great tool to allocate resources efficiently
Any empirical support for that?
Do you need the stock market to undercut industries though? I'm not sure it's necessary
While not strictly necessary, it is a great power multiplier.
It helps as it is both a gauge of the success of the strategy, and also a lever where the process can be fine tuned, eg. slowly buying stock then strategically dumping in the right time, correlated with other external shocks can have wider effect to whole industries through controlling the public opinion on specific industries.
> a great tool to allocate resources efficiently
Sorry but... WTF are you talking about?
It rewards self-destructive behavior in favor of short-term gains. Shareholders have *zero* commitment to the companies they buy shares from and will happily switch their entire portfolios on a whim. It's essentially people chasing the new shiny thing every single day.
Let's not forget it's a known fact that people with insider knowledge will profit over everyone else.
How is that efficient in any shape or form?
> If they undercut the US companies and are willing to accept low returns on their investments, then the respective USA competition will be driven out of business by their investors, because there will be other sectors to invest in, with higher RoI.
You're basically explaining one of the reasons stocks are a horrible idea for distributing resources.
It has nothing to do with whether or not it's central or distributed, it's merely the incentives they create. It's essentially Goodhart's law on steroids.
Depends what you compare it with. I grew up in the post soviet union. That system allocated resources to various monopolies who were too big to fail. Turns out allocating capital based on who can make things that people want/need to buy, and do it with a profit, multiplies said capital way faster. From this point of view, over time your initial base grows into all kinds of industries etc. That's probably why the USA won the cold war.
Until your economy congeals into various monopolies who are too big to fail, and you lose the rematch because you're so busy counting beans you stop paying attention.
A bad system is still bad even if something else is worse.
> That system allocated resources to various monopolies who were too big to fail.
Like our banks today? Like OpenAI is trying to sell themselves as so the US government bails them out when they inevitably can't pay off their debts? Like all the other monopolies that are too big to fail?
I can't see the difference.
> turns out allocating capital based on who can make things that people want/need to buy, and do it with a profit, multiplies said capital way faster.
That assumes the stock market does that. It simply doesn't and history is proof. It's also a fact that free market capitalism has consistently done worse than before based on actual numbers.
I highly suggest you read 23 things they don't tell you about capitalism. It goes into a lot more detail and helps you understand things in a broader perspective.
If we want to improve our world we need to move past the faults of the current capitalist system and the soviet union. We need better incentives. Profit has proven to be too disconnected from actual value.
[1]: https://en.wikipedia.org/wiki/23_Things_They_Don%27t_Tell_Yo...
I'd argue the Cold War isn't over and the US is losing it right now with how their president grovels before the Russian one
The stock market weirdly enough ruins the idea of capitalism. Catering to shareholders hurts the idea that competition would create better and cheaper products.
Its not the stock market per se. The biggest problem is a lack of good regulation to ensure competition and the resulting drift of oligopolies.
The stockmarket enables that by making takeovers easier as you have a higher proportion of short termist shareholders who 1) fail to block value destroying acquisitions on one side and 2) jump at the chance to make a quick profit on the other.
that's not the idea of capitalism; the idea of capitalism is that you should be able to make money by virtue of owning stuff. it's an inherently rich-get-richer scheme, competition has little to do with it.
I think its obvious the GP means free market capitalism, which is what almost everyone who favours capitalism thinks is the form it should take.
The free market abhors competition. It's much more profitable to be a monopoly - and profitable enough by far to squash any competitors in infancy.
Capitalists abhor competition. Adam Smith (as in "invisible hand") pointed this out. That is a subversion of a free market.
I don't see how free market capitalism fixes that. I looked up a definition to make sure I wasn't missing something and as per investopedia:
"The term “free market capitalism” refers to an economy that puts no or minimal barriers in the way of privately owned businesses. Matters such as worker rights, environmental protection, and product safety will be addressed by businesses as the marketplace demands."
it's basically worship of owning the means of production and not being regulated in its use, e.g. if you own a company you get to dictate all sorts of unreasonable things to your employees, and any benefits gained from automation accrue to whoever can afford the up front money to own the machines.
That is a bad definition and I cannot find it on investopedia. it is essentially an extreme libertarian spin on the definition.
There are better definitions on both wikipedia and Britannica:
https://en.wikipedia.org/wiki/Free_market_capitalism
https://www.britannica.com/money/free-market
Especially this hit:
https://en.wikipedia.org/wiki/Free_market_capitalism#Concept...
And these people has audacity to proclaim that socialism is based on fantasy.
Talk about utopia huh?
Free markets never existed, don't exist and never will. Markets are defined by laws and regulations in which they exist in. They can't ever be "free".
Relevant U.S. Constitution: https://constitution.congress.gov/constitution/article-1/#ar...
> “The Congress shall have Power… To regulate Commerce with foreign Nations, and among the several States, and with the Indian Tribes;”
I agree. A free market requires regulation to ensure competition otherwise it ceases to be a free market.
I assume the downvotes are because people do not know the difference between "free market" and "entirely unregualted market"?
Please get your definitions from someone reasonable (Adam Smith might be a good start) rather than Ayn Rand.
Applies globally.
Competition creating better products isn't an idea that defines capitalism though: the same would apply to cottage industry.
Capitalism is defined by having the capitalist, who provides capital, and without the ability to sell their share of stock it's difficult to see what the value would be. So you kind of require stock markets.
Edit: which is why it's odd to call China communist. They have 3 stock exchanges. They're really a capitalist single-party state.
China uses Capitalism as a tool where the Party feels it would be beneficial (for the Party), and crushes it mercilessly when it gets in the way (other than this real estate problem they have right now).
In the U.S. we have mistaken Capitalism for a religion, and so it wags the dog, so to speak. Since our founding we have made some attempts at finding a balance between our use of the tools of Capitalism and socialism (in more the Democratic Socialism style, rather than the Communism style), and we had a good run in the decades after WWII. But starting with McCarthyism, and really picking up under Regan we have prided ourselves on adopting Capitalism as a religion, and it really shows up in both the income inequality as well as the increasing role of (and corrupting influence of) money in our politics/government.
That is not what stock market is. A company does not have to focus on stock price and stock price is not its primary product.
That's fair, I should reframe. The incentive given to decision makers at Microsoft is company stock. That means the primary focus for everyone who makes decisions at Microsoft is the stock price, which in turn means the stock price is the primary product for the company itself.
A company does need to keep its market cap above that of its assets, if it wants to avoid a hostile takeover and sell-off of those assets.
Name a major U.S. public company in recent years that has consistently prioritized improving its product over boosting short term stock price or extracting maximum profits. If capitalism were truly a healthy system about building strong products to create healthy markets, this should be the norm (and enshitification shunned), not the exception.
What we actually see is a system of chartered extraction. Corporate executives are like Norman lords, granted their 'title' (CEO of instead of Earl of) by shareholders (rather than a king) in return for which both are/were expected to extract maximum value by any means necessary. Extractive tactics often at the expense of long-term product strength are behaviors shareholders expect if the CEO is to keep their bestowed 'title'.
Don't forget the progenitor joint stock company The East India Company, Capitalism in it's purest form without government restriction. Profit-maximizing, absentee extraction, with company executives serving as quasi-feudal lords over assets and people. Modern corporate capitalism is hard to distinguish, in its structure,history, behavior, and incentives, from the Norman extraction system, it's just dressed in a more politically palatable wrapper and forced to mellow out from it's desired East India Company style final form.
> Name a major U.S. public company in recent years that has consistently prioritized improving its product over boosting short term stock price or extracting maximum profits.
Well in some perverse sense, I'd say Meta qualifies here. Zuck isn't beholden to other shareholders and is free to burn truckloads of money on worthless projects. The big asterisk is that for Meta, "improving its product" is effectively "creating the best digital cigarettes".
Lots of companies do this (Arizona green tea immediately comes to mind), but more importantly, this has no bearing on the discussion of whether or not capitalism is "good" (really, who ascribes morality to any economic system?), because the US is primarily an exercise in regulatory capture.
Good doesn't mean moral in this case, it means valuable. Capitalism, or whatever the fuck we have right now, doesn't create much real value, but instead it creates a lot of fake value that we just proxy.
And I say "or whatever the fuck" because, as you point out, it's not really capitalism. And maybe it never was - after all, the US Golden Era was when we had the most central planning, the most social services, and the highest taxes.
But everybody right now is exploiting one teensy little flaw of the economy - humans don't live very long. There's no reason to make stuff good for later when you can take a cash advance and get rich now. Everything is always someone else's problem.
I think one fun case study is superstar CEOs. They're so highly desirable, and they work 5 years at a company and "turn it around" before moving onto the next. But if you look at the company after they leave, it's in shambles. And yet, they are the most sought after CEOs. You don't want the Arizona Green Tea guy, no, you want the guy who ran Chipotle into the ground. But hey, at least he increased their stock price for, like, a little bit.
Arizona comes to mind because it's the outlier that everyone knows about, the company that bucks the capitalist pressure. Your choice to use it basically proves my point.
Capitalism outside regulatory capture looks like the East India company. You are basically giving me 'no REAL communism has ever been tried'.
Why it must be "major"? Major is counted by stock price, you created a circular test.
And that still does not make stock the product.
Because they so much want to be a service business than a software business. Microsoft execs are losing sleep over becoming the next IBM, not realizing they are already there and have been for a long time.
Their main problem is that they never really learned how to compete on merit, just on first-to-market and all kinds of legal (and illegal) tricks.
I’m actually somewhat stoked about generative AI from a “good enough” perspective, because at this inflection point where a lot of countries and organizations are looking for Microsoft alternatives (digital sovereignty, etc), this is the best time to be able to build and deploy alternatives with the productivity advantages (if any) AI might provide.
Big Tech thinks they have a moat, when it’s really diffuse power being made available via genAI to build software good enough to replace them.
Big tech's moat is mostly built up around perceived security (not actual security) and abuse of monopoly positions, AI is going to make it easier for them to maintain this because it serves as an addictive component for the users, or so they hope. In practice it only appeals to the lowest common denominator, which is exactly how they built their empires in the first place. AI allows non-experts to pretend they are experts with confidence, and to produce output they claim as their own to which they have no real title. As a democratizing principle that's great, but as a quality-of-service-provided item it could easily become a disaster and I think MS et-al are betting that they will be able to 'fake it until they make it' on the quality front to avert that disaster.
So far, I'm not seeing it. All I see is a massive leap forward in the first two years that still had some fundamental problems and a lot of fancy packaging of the same broken stuff since then. We're looking at band-aids here, not actual progress.
To be pedantic, IBM is a service company
This follows from the parents statement.
And IBM could have been AWS a decade earlier had they so chose.
That's the point.
Or do things that actually work. Why, for example, can I not translate a PowerPoint using Copilot in Powerpoint? Why do I need to save it, then upload it into ChatGPT, translate it, then download it again, and open it in PowerPoint for further editing. But at the same time get all kind off nonsense I don't want pushed at me in Windows, like that MSN news clickbait crap.
Exactly, even those of us that like Windows have a hard time talking about it when Microsoft treats it so badly, I really miss Balmer era in regards to Windows.
The only good thing that came out of Satya era has been the Windows Terminal and WSL.
As long as companies, and consumers, still pick Windows and Office, then why spend the resources. Making Windows better won't move the sales number significantly, but removing the ads and the potential AI upsell is a direct hit to revenue.
The sad reality seems to be that Microsoft do not care about the majority of their products anymore. Only Azure, Microsoft 365 CoPilot, CoPilot and maybe CoPilot.
I'm not familiar with many "consumers" who still pick a Windows and Office, and in this generation, there are very few consumers picking xbox. Outside of enterprises, they seem to be losing market share everywhere, and at this rate they'll be akin to IBM or Oracle in a few years.
Office is part of the "Productivity and Business Processes" at Microsoft. That business unit had $120B of revenue in 2025.
Microsoft 365, which I believe includes Office, makes up $95B of that amount, which is split between Commercial (92%) and Consumer (8%)
From there you can see why they're focused on Enterprise.
Source: https://www.bamsec.com/filing/95017025100235?cik=789019 (page 39)
But "we're fixing Windows and Office" doesn't get you a fat bonus, and it doesn't get investors a double-digit growth quarter. Anyone who tries to do the right thing will be disposed of. Why don't people understand this?
Nobody gets a bonus and a new boat doing that.
I will say that with enough group policy and sysinternals turning absolutely everything off, turning all of the settings to maximum performance lowest flashiness, no web results, killing Cortana with reckless abandon my Windows installation is actually what I would consider to be snappy. I was surprised.
It doesn't make it any better that Microsoft does this, but as a piece of practical advice, it seems like it can be done. There does still exist a core of Windows under all that garbage that is fast.
All that tinkering is getting you dangerously close to daily driving linux. And the advantage there is that the maker isnt actively trying to get in your way
I've got two laptop in my new job. They sent me a windows one, when I asked for a linux one. Had to set up the laptop to begin working.
Honestly, I had to do a lot of workarounds to get comfy. There's annoying stuff I cannot uninstall.
Still wonder why the OneDrive mobile can’t find a file and the photo backup has been broken for months. But I have a copilot button in notepad.
Satya’s days are numbered. The OpenAI investment will be a very costly and painful lesson.
And photo back up has never included metadata like locations when backing up to OneDrive making it usekdss
Onedrive, that useless app that creates a mess in the desktop if you have a laptop and a desktop like most users
The new broken Briefcase model from Windows 95, only far far far worse!
What progress.
This is what's so amazing to me, a primary feature of cloud storage is multi-device usage. But by default this ridiculous product causes a mess if you use it on 2 or more computers, and many apps save app data to my documents so you have useless (at best) or conflicting at worst bloat being copied between computers. I want to use it but Microsoft makes it a huge pain if I do.
Most users do not own two computers.
A huge percentage do, whether it's most or not would be a stupid metric by which to design a cloud storage product in such a way that causes issues when one does sign in on multiple devices.
No, but the biggest users who pay the most do.
well, onedrive is the ms solution to dropbox which is main purpose is syncing between computers
I was hoping for a real look at weaknesses in Microsoft’s AI products. They ship lots of “AI features,” but only Copilot and Azure’s ChatGPT hosting see broad use. Instead, the article mostly reads as anti-MS/OpenAI without much detail.
From my experience, Microsoft’s GPT-5 integrations in Word, PowerPoint, and their ChatGPT clone struggle with basic tasks. Copy/pasting from ChatGPT still works better.
To be fair, building solid AI features is hard when model capabilities change so quickly. Reasoning and tool use only became reliable in the latest models, and when these Office features were planned, GPT-5 didn’t exist.
> Copy/pasting from ChatGPT still works better.
Side tangent: Copy/pasting from the Windows Copilot app is absolute dogshit. It makes no sense that the simple action of "copy this text" is this broken.
I have found it copies the markdown that is used for the formatting?
Perhaps this hopeless implementation is a wonderful microcosm and accurate barometer of the usefulness of AI-assisted development and tools.
They bought Dragon a few years ago, and 2 years ago they debuted the Dragon Ambient Experience, then renamed to Dragon Copilot. We had dozens of doctors try it, after a handful of months most had quit, it was a bad product. We switched to a competitor at literally 1/6th the price, and we don't even have to offer it, the doctors tell each other about it and they ask for it.
Nadella has done a lot of listening through is CEO reign but it looks like MS is back in a "don't listen to customers, tell them what they'll get" phase.
So, you're saying it's a Bad Dragon?
Indeed!
Lol the Copilot app isn't even that useful on iOS for a blind person. On Android, you type something in, hit sent, and the app pipes the pure output of the AI, Markdown formatting and citation markup included, to the screen reader. That's at least something. I mean it's crumbs, yes, but we blind people are very, very used to crumbs.
On iOS, you type a message and send, and... nothing.
Big corporate AI products are all currently stupid bolt-ons that some committee decided solved a problem.
When the internet came out, did many legacy companies lead the way with online experiences, figuring out what the real killer apps now that everyone was connected were? I don’t know for sure, but I doubt it, I think it gave rise to some of the present crop of big tech, and others reinvented themselves after the use cases were discovered.
All that to say, I expect the same here. In 10 years there will be AI uses we take for granted, built by companies we haven’t heard of yet (plus the coding apps) and nobody will talk about stupid “rephrase with AI” and other mindless crap that legacy companies tried to push.
> Big corporate AI products are all currently stupid bolt-ons that some committee decided solved a problem.
Or not even .. maybe someone said all products need to be AI enabled, so now they are. Just append "AI" to the product name, add bolt-on to call an LLM to do something, and declare mission accomplished.
Big corpos have reached the stage where they can hire ex-politburo apparatchiks from the soviets or china, straight into C-suite roles, and nothing will materially change.
I moved largely away from Microsoft products in 2010 personally, and 2015 or so professionally. Just came back to them last year for a client.
What I am seeing are two really disturbing patterns: 1) really, really bad stuff that has been there for decades is still there, weighing everything down, for "backwards compatibility", and 2) a lot of horrible fluff and nonsense everywhere.
#1 is pretty self-explanatory, as I am editing docs in Word I am finding my muscle memory circa 1999 or so still works with it due to all of the formatting bugs that still exist.
#2 I saw with Windows 11 and the crazy adverts everywhere, seemingly random UI choices, half-broken or implemented tools and applications, and now random UI buttons thrown everywhere.
The 1-2 punch is devastating and makes using Microsoft products 10x more depressing than it was 10-15 years ago, and it wasn't a happy fest then either for me.
So to me, AI is just "more of the same" with Microsoft. It is more random shit thrown on the wall to see what will stick.
Backwards compatibility has always been a Microsoft staple. What used to be a huge selling point - depending on the audience - is now clearly a crutch. Right now, it seems that the tech debt has finally started making the whole stack lean like the Tower of Pisa.
I feel like the case for Microsoft inability to execute in a lot of verticals should really be studied, not saying this as a sound bite, I'd genuinely like to know how that is possible.
Their investment in OpenAI, giving them what was, at least ~1-2 year ago if not now, the best possible LLM to integrate in the office suite yet they are unable to deliver value with it.
Their ownership of Xbox and Windows should have allowed them to get a much better foothold in gaming yet their marketplace is still, to this very day, a broken experience with multiple account types. It's been 10 years.
The counter point is Azure obviously which still has great growth numbers, but that's a different org.
From the outside, it just seems like they should be doing better than they are. They have much better business integration than Google and Amazon. The fit is obvious and people are borderline hooked on excel. Why aren't they dominating completely?
What fascinates me is how someone at MSFT managed to get their own button (the Copilot) button approved. Having your own button or tab is a big deal (I remember we debated having our own shopping tab on the Instagram app while I was at Meta - and we had to abandon it when it didn't get the traction we were hoping for).
These people have a UI button AND a hardware button on actual consumer devices, and it doesn't do anything. How?
AI assistance is a gold rush — promotions are to be made and huge complex system to be over-engineered in Big Tech. The race to stake out the future empires is underway, and there is no time to think about the quality control, UX etc. But who am I kidding though, there is no time to think about those things during the chillest of times either, as any user of Power Automate can concur.
It's being treated like a gold rush but I don't think it really is. This is like dotcom 1.0 all over again. They didn't know what the best use cases for the internet were but they still poured billions into it. The gold rush didn't come until the 2000s when social media took off.
Normally you get a frontier exploration phase where fringe people experiment with the new technology and try to figure out what it's good for. It feels like we just skipped that step entirely.
It's sad that the "gold rush" on social media just means "using the Internet" is a depressing experience of being tracked incessantly and sold to, constantly. A lot of the content isn't to educate or delight, but to build a following exclusively so that you can sell to said following.
It's more like the "poo rush", where you are the one getting pooped on.
It seems to me that America is a great big nomenclature butchering machine. If they can butcher what words like crypto or AI mean, they’re willing to put their whole economy behind that. I need to rethink my investment strategies.
This is true. Just look at what selling of products with "memory" means - "the tablet has 256GB memory"; no it doesn't. It has 8GB RAM and 256GB non-volatile storage.
I suppose "AI" is easier to say than "word-guessing non-deterministic instruction transformer", but it does carry "intelligence" into the hearer, when the truth is that "AI" is not intelligent - it is a great mimicker of everything it has read.
I'm always glad to see others notice this too. cheers
Would you be so kind to elaborate this?
Between the RAM shortage and the forced migration to Win11 along with a forced HW upgrade, I'd start shorting MS asap. This bit about lack of adoption of Copilot is just icing on the cake.
The AI beating will continue untill the buying improves. And the use will be forсed by changing the OS.
> FirstPageSage AI Chatbot Usage Chart (December 3, 2025)
What a bizarre way to organize the chart. Clearly Anthropic is leading -- their early bet on "programming" as the main use case is paying off.
The recent report from Openrouter [1] confirms as much: coding is the number one use case, with role playing / fantasy writing in second place.
I have a feeling the remaining use cases will never dominate, instead they will slowly mature into acceptable practices across the various industries. That will probably take longer than the investment bubble can hold though.
1: https://openrouter.ai/state-of-ai
Claude Code usage probably isn't counted as "chatbot" use. Also, I think you're overestimating how many people program vs. how many people are using AI chatbots as the new websearch. Orders of magnitude more of the latter.
Sure - US has 1M developers vs 300M pop. At least the Claude Code developers are paying for it though, vs only 5% of users paying for ChatGPT.
Mustafa is a side-character, he's the friend of a brother of someone who knows what they're doing. Competence isn't something that people pick up through proximity, and Microsoft is finding that out in real-time.
I personally never found Microsoft software anything but shoddy: Microsoft word, internet explorer, power shell, outlook, powerpoint... these all make me shudder... I'll give them visual studio, that one is pretty good. I think Excel is fine. Skype was okay I guess. Anyway, I think the shoddiness can't really be the actual threat to Microsoft given their track record.
When I joined my current employer, a few years ago, one of the things I thought might be good was that for the first time in a fairly long career I'd be using Visual Studio.
Previously I'm a vi or vim user for everything, for many years.
But I can say after a few years experience I'm not really impressed. It's too big and too slow. It has a few things I kinda like, a lot of half-working things I'd love if they worked consistently (e.g. some things I work on can be debugged, some can't, experts might know why but I don't) but as they are they're too unreliable to really change how I do things - but overall it's not enough to e.g. miss VS when I'm writing my own stuff, still in vim on Linux.
Actually all of the Microsoft technologies I've run into were disappointing with only two exceptions which I'll get to. Powershell felt like they hadn't really learned the right lessons from the Unix shells for example, and Entra ID (called "Azure Active Directory" when I started caring) is a confusing mess.
Two exceptions: 1. C# is a pretty good language. Mostly it's a better Java. Is that amazing? Not really, but it's still pretty good, it delivers reasonable performance, there's a large ecosystem, I don't hate writing it.
2. Azure itself has to have a way to "cut off" payments because Microsoft sold a product where students can get a limited amount of credit. The student doesn't have any money, so if they had $50 of Azure credit and Microsoft lets them spend $85 before turning off all the Azure systems that credit was funding, well, too fucking bad - Microsoft eats the $35 loss. Accordingly Microsoft are better (not perfect, but better) than AWS or Google's thing at actually turning stuff off when it exceeds what you asked to spend.
Skype was an acquisition.
And became pretty bad afterwards
Microsoft's problem isn't bad AI, but a fundamentally wrong UX pattern. They're trying to shoehorn Copilot into linear workflows - Word, Excel, Outlook - but chat is a terrible interface for editing documents or analyzing tables. Google with Gemini is winning not because their model is better (though it is good), but because they integrate AI better as a tool, not as a conversational partner. This difference in mental model defines utility
I've used Copilot ih VSCode quite extensively, using both GPT-5(.1) and Sonnet 3.7 and Sonnet 4.5 as the LLM backend. (Before our company got the deals in place for OpenAI and Anthropic to use their tooling directly)
The results are measurably worse every time vs. doing the exact same operation with Codex Cli or Claude Code.
There's something in the Microsoft AI framework that makes the worse and more "stupid" than the competition.
My first experience of copilot was learning that my Microsoft Office subscription was increasing in price.
Then I dug a little deeper and saw that I could exclude it and go back to my old price.
For what I do with office, I have no use for copilot so you can guess what I did.
An unfortunate reality that software makers need to grasp is that a lot of people don’t use their software properly to begin with, meaning the addition of AI doesn’t push the needle - it’s just yet another “time saving feature” that the user won’t touch.
Maybe they could add a helpful paper clip to improve sales.
Edit: Or better still, convince all of their customers to throw away perfectly good hardware and upgrade to one with a single extra chip, creating a hazardous waste epidemic for landfills as a nice side effect. It's especially important to do this in the middle of a RAM and HDD shortage.
Really, I'll just never be half the great business strategist that these guys are. <sigh>
It doesn’t matter. Microsoft has a monopoly on the desktop, and they have no qualms using it to displace competing products. They did it with Teams, and they’ll keep doing it because they know there’s no appetite for anti-trust prosecution anymore (or maybe they feel comfortable arguing they’re no longer a monopoly because they have no presence in mobile).
Every procurement team is going to point to copilot, saying it’s included with the other Microsoft services a company is already paying for, so duplicate AI products won’t be approved for purchase.
Microsoft is laying claim to the desktop real estate, so in a few more generations of the technology, they’ll have the customers and competitors will already be starved out.
Maybe Microsoft should go all in and replace the CEO by an AI?
Not just about the products imho. I do some consulting for law firms who typically use the MSFT stack, and I was excited about the private ChatGPT services in Azure, because from my (admittedly limited) sample of law firms, nobody likes using Copilot and LLMs need to be private/secure. The amount of outdated and poor quality documentation for Azure services is amazing given how nascent these services are.
Is there an excellent "AI Free" linux distro that one can escape to when AI is inescapable from both Windows/MacOS
All of them? I know of no Linux distros that do anything in particular to integrate AI.
Although knowing Canonical they might add something to Ubuntu sooner or later.
Counter: I use various co-pilot features at work, they are very helpful, save me hours, and many folks in my team do to.
Their copilot stuff is such a mess. Over promising, dressed with marketing slop on top of an under-performing product.
However; their azure offerings are somewhat decent. AI Search is getting quite decent, access to Open AI models served through foundry is quite the differentiator (although the rate limits are an issue), their reference architectures are very helpful and their ancillary services, like document intelligence integrate well.
I'm a bit worried that their marketing dept has gotten their crappifying and obfuscating eyes set on those services though, given the recent rebranding of Foundry and what not, but the underlying services are decent.
> Their <snip> stuff is such a mess. Over promising, dressed with marketing slop on top of an under-performing product.
Updated as it was almost close to being a generic comment about AI overall.
See WinUI after Project Reunion announcement 5 years ago, unfortunately fits exactly the same description, and we are way past COVID to use that as an excuse.
Nobody sane wants to use their shoddy products.
Their office products can't even agree having the same menu options for the same functionality.
People use them out of corporate inertia: Windows laptops are cheaper than Macs and that brings the whole office suite in.
> People use them out of corporate inertia
And now with how things are going with the American gov, foreign companies see security flaws on two fronts: Microsoft's AI's unreliability+invasiveness and the US gov's unreliability+invasiveness relating to its stateside companies.
Microsoft is using the deep penetration of SharePoint in companies to sell Copilot license. At least in France it's well and alive and I see much more Copilot licenses than actual OpenAI uses.
I just had a sales call with someone from Microsoft who was looking for an AI tool to automate some Excel work they were doing. I doubt they'll buy our product, but it gave me a good laugh.
BTW. Feeding that AI boom, that is driving up the cost of hardware... while at the same time asking people to buy new computers only to run the next version of their OS.
Not a great plan.
Just remember what SPJ said what the problem with Microsoft is.
Which is what?
... they have no taste.
Aw, that won't matter when management just forces you to use Copilot or else you're fired.
Unfortunately that is probably how it will end.
That's how it already is in some companies.
I have not used Windows in ages. But I use Visual Studio under Linux that has some AI included. Does this count?
MS Bob -> Clippy -> Copilot -> ?
You have forgotten Cortana
So did Microsoft
Cortana... ah yes, that thing that I immediately disabled. I had forgotten its name.
Microsoft doesn't just have a shoddy AI problem. Microsoft has a direction problem. I'm no fan of Ballmer, but his Microsoft seemed like they knew what they were doing, and were actually trying to be good at it. Nadella seems extremely clueless and seems happy to just ignore and later axe consumer products that don't generate immediate revenue.
And noone should actually be shocked about his ineffectiveness. Covid was a great example of how clueless his leadership has been. Skype used to be a verb people used in common parlance, and yet they dropped the ball and let Zoom take over both consumer and enterprise segments while focusing on "restructuring" Skype into Teams for no reason whatsoever.
Prior to Covid, he was ready to let Windows run its course and axe that too. The sudden demand for sub-$500 laptops during the pandemic showed him that people still liked Windows and wanted a good OS from Microsoft. But instead of capitalizing on it to give customers what they wanted, he just gave us an ad-filled spyware with AI slop.
I have zero hope in any product with a Copilot in its name (including GitHub). At this point, unless there's a change in leadership, it's only a matter of time before XBox faces the axe.
Or is it that nobody wants to use integrated (i.e. force fed) AI products?
Satya is doing with AI what Ballmer did with cloud.
Right direction, wrong execution.
I am honestly struggling to figure out Microsoft's licensing system. We have it here and there are some on my team that want to try "CoPilot" (not even sure what that means now) but there is no licenses option for me to enable them to do that. I am sure they are targeting the larger places than my 50 person shop but it is so confusing I don't have the time in my day to figure out how to enable copilot for my team.
The more MS pushing a product - the less people want to use it.
I see mentions of Gemini as a fast growing alternative to ChatGPT. Isn't anyone troubled by the fact that for consumers there is no way to keep your data from being used for model training if you want to maintain history of your Gemini chats.
ChatGPT respects privacy and allows for maintaining history while also opting out of using ones data for model training
I trust Google ad monopoly to keep my data actually secure. They have a great track record of not sharing their datasets with anyone because this gives them an edge pushing ads down people throats. Google is honest about what they doing. Google also not going away anytime soon so they also not going to sell off their datasets to highest bidder.
And I don't trust Sam Altman and AI.com at all since their whole thing was built on lies. They could start regaining the trust by changing their company name.
It's not just about protecting data in the old sense - typically from other corporate entities. It would suck if your information somehow made it into a generally available model that then leaked some of that anyone asking a question
> I trust Google...
Yup, yeah, sure. The company that attempts to open your password-protected zip files. Let us not give it a free pass either.
There is no good incumbent.
You just made that up, huh?
https://grahamcluley.com/shouldnt-gmail-zip-files-password-i...
Its been a while so i had trouble finding it (but Grok obliged)
Moreover its always the edge cases that people are 'OK' with, but again if they can do it (setup the infrastructure) for one thing they can do it for anything, and it makes 'trusting' them seem naive. Since trusting was the original statement.
Open VSCode, close co-pilot again.
That monkey face simply won’t go away.
Who are the likely successors if Satya steps down?
> I suspect the issues are deeper for Microsoft, who have worked tirelessly under Satya Nadella to create doubt around its products
This reads more like a hit piece than good faith article
(But yeah the MS AI products especially on consumer level are pretty terrible)
Microsoft’s early/rushed attempt at AI (the various things called “Copilot”) does start to feel like how they lost the mobile wars. They had a tech fairly early (Windows Mobile) but utterly failed to execute on it and ultimately, allowed a competitor to dominate with a better product.
It’ll be interesting to see how this plays out. With the sums of money involved it could end up being make or break?
[dupe] Discussion: https://news.ycombinator.com/item?id=46148748
Gemini is really great now. Fast is insanely fast and handles 90% of queries. Deep Research works better than OpenAI's deep research given their search expertise.
It is going to be very hard for OpenAI or Microsoft to compete with this now that Google has gotten their act together.
That’s the joy of not having to compete, your stuff is just there.
I mean have you tried them ? I did and they're beyond terrible, of course they're not the one I pay for
When you're on the internet for a while, you develop an intuition for the average level of Microsoft hate that is supposed to exist at any given time. It's a rather stable measure.
I'm seeing an unusual increase in that, and after some thought, I think the only plausible reason is that another company is afraid of them and creating anti-MS editorialized buzz.
My gut feeling tells me this is all related to long-term sustainability of AI products, and both GitHub and VSCode are major points of anxiety for other AI coding platforms.
We will eventually know if I'm right or not. If we see moves that point to other companies trying to decouple from GitHub and VSCode, then it's probably what is actually happening here.
Isn’t the problem more so that nobody wants to buy it and Microsoft is leveraging its power and influence to successfully coerce and pressure people into buying them?
You using OpenAI’s products? Aren’t Microsoft getting cuts as owners?
Anyone remember Cortana? It seems like MS doesnt learn
Wait, Cortana doesn't exist anymore?
It doesn't (but it seems pretty much every Windows install has it installed, and if you try to open it, it just tells you that there's no Cortana anymore.)
BTW. They did the same with Maps and the PDF reader: the two apps I used the most on my Windows tablet. An "upgrade" replaced the app with a nonce. That just made me so incredibly angry.
Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
But this also means they end up with average products. They don't have the talent to do something exceptional.
This has worked well for them when they can just come in and copy something (say AWS in Azure) and not pay the innovation cost, but AI seems different for some reason, perhaps in the same way search was. You need the top 20% in order to really be successful.
That’s exactly what happened. For the past decade, the crème de la crème went to Google and Meta, which offered nearly double what Microsoft paid new graduates. Microsoft hired the next tier, after the top talent had already been skimmed off by Google and Meta.
Intel/AMD hired 2 tiers below MS. Intel's "great place to work" corporate propaganda was known as "Great place to leetcode" by engineers there.
this is just not true. Building great products with average talent is a sign of great management, and it's been done before in both business and sports. moneyball is about this idea at some level.
Plenty of SV is building below average products with exceptional talent.
> Building great products
This is where they are failing.
> Plenty of SV is building below average products with exceptional talent.
Yes, you can hire exceptional talent and give them poor directions, resulting in poor products.
But to hire mediocre talent and still produce competitive products you must have an unfair advantage of some sort. The Windows and Office monopolies gave Microsoft that unfair advantage. But it is becoming clear that this unfair advantage does not extend to AI.
> Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
This is such a BS attitude. A person’s ability doesn’t map directly with their salary.
The very best developers I’ve ever worked with? Small companies outside of major cities (for life style reasons). Not everyone is trying to max their income; some want a good job doing something they love.
Small companies can offer more meaningful impact, if you’re not the type who likes being a tiny cog or commanding armies.
Sure, but Microsoft isn't attracting those devs either.
You are coping. Overwhelmingly the quality of developers is reflected in their pay packages. Overwhelmingly most people are motivated by money/benefits/QOL/GPUs.
This attitude is how you maximize your TC.
> You are coping.
This is an amusing response. I personally have chased the TC (to an extent), climbed the ladder, and sit at the top (below the c-suite). This isn’t cope, it’s a genuine reflection based on many years of experience having worked for multinationals in various countries.
Perhaps compensation should be a reflection of ability, but the real world is far more nuanced.
The only time they copied something successfully and did not rely on major tie-ins with their existing monopoly was Xbox and that division lost money hand over fist for a long time.
> Microsoft has a problem that they hire the middle block of talent in the market. They do not chase the top 20% most expensive nor the bottom 20% least expensive.
That's only if you're only looking at the top 10% or w/e of the market in the first place. New grad software engineer salaries at MS are higher than the median software engineer salary in the US.
Microsoft has this problem with most of its products.
It's not just AI, it's a market fit and quality problem.
They don't need to solve it, however.
Their strategy has been quite clear: make it barely usable so that is passes muster to auditors, integrate it with systems that corporations need, and sell them on the integrations.
Teams and Azure suck?
So what?
Big companies will pay for that, because it's integrated with their ldap, has an audit trail, gives them the ISO-whatever stamp, and lets them worry about something else.
That the users are miserable is almost never the question for the ones signing the checks.
In a world where box-checking is paramount, this approach is a winning strategy.
Nothing an ad carousel full of AI-slop, embedded in the OS won’t fix
because it is absolute shit and forcefeeding customers it will only make them resent it or use it less
https://www.visualcapitalist.com/microsofts-revenue-by-produ...
Helpful chart to draw conclusions
This is from 2025: https://www.visualcapitalist.com/how-microsoft-makes-its-bil...
Copilot and AI are not on the list anymore.
I doubt it's useful to draw conclusions in today's world. The chart is almost two years old.
Is Bing now called "Copilot" too?
I don't understand why they name everything Copilot now, it's the most confusing thing in the world. Copilot AI, Copilot 365 Copilot. GitHub Copilot makes the most sense I think because it's Copilot AI but for coding.
Or "Watson". I lose track.
Copilot .NET Core, not to be confused with Copilot Core, and Microsoft Core Copilot Plus.
Not so helpful, it's too out of date
That's 2 years old?