Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Microsoft starts canceling Claude Code licenses

theverge.com

Post reply on HN
Print view View original

441–450 of 489 posts

Re: Microsoft starts canceling Claude Code licenses

#441

Post by CamperBob2 »

NiloCK wrote: post #202
serf wrote: post #184

Earlier quoted context omitted.

the obvious answer is because it's easier , faster, and more efficient to flip a true to false right in front of you than it is to prompt an llm. if your response is "my prompts don't produce code that needs values flipped, ever." then I would wager you're only touching very simple things with an LLM. for me I don't care about the token cost and prompt writing so much as the fact that it's just faster to change 0 to…

The thing that drove me away from manual edits was that I found myself confusing the LLM all the time. It would read or write, some code, I'd twiddle with things, and then the LLM's future references to the same code would be a mess. On balance, and via dictation, it feels likely to be faster overall to just enact the changes I want 'inline' of the conversation thread. Is this stuff any better now? I think current ha…

I try to remember to tell it that the file changed, and should be reloaded. That usually avoids confusion.

Re: Microsoft starts canceling Claude Code licenses

#442

Post by Terretta »

epistasis wrote: post #366
Terretta wrote: post #344

Earlier quoted context omitted.

Anyone (including ANTHROP\C) "recommending selective use of cheaper models" is spending costly human time (which costs more over time) on correcting the machine (which costs less over time). This is a bad trade. In cost per line of code, we have verified this is always an error unless your time is worth less than the machine (unlikely unless you consider your time to have no cost rather than considering it as your ho…

I disagree heartily with everything here, both in personal experience from the models, and in values about coding. I don't care bout cost, I care about getting good results fast. Cost per line of code is not a suitable metric for anything. It's as silly as measuring engineers' performance by lines of code. More lines of code is worse than fewer lines of code. When you say "we have verified" whoever that "we" is makes…

Yes, cost per line of code itself is an error.

Only cost for effective* outcome matters. And if your lines of code have a cost, you would want fewer lines of code to achieve the outcome, not more.

Are you sure you disagree with that?

* If your place of work starts talking "efficiency"**, run. Find somewhere the conversation is *effectiveness* — at the goal/outcome level.

** Not to mention that "efficiencies" is MBA speak for "right sizing" away effectiveness.

Re: Microsoft starts canceling Claude Code licenses

#443

Post by Terretta »

majormajor wrote: post #408
Terretta wrote: post #344

Earlier quoted context omitted.

Anyone (including ANTHROP\C) "recommending selective use of cheaper models" is spending costly human time (which costs more over time) on correcting the machine (which costs less over time). This is a bad trade. In cost per line of code, we have verified this is always an error unless your time is worth less than the machine (unlikely unless you consider your time to have no cost rather than considering it as your ho…

> The single best fix for results-per-total-cost is to ensure it reads and thinks about whole content, not snippets, and thinks with the smartest model, not agents. I haven't seen "just absorb a giant ball of context and do the right thing the first time" be cracked yet, even for Opus 4.7. At the end of the day, code is code, and we have decades of lessons about how to make code more reliable and maintainable. Compos…

I haven't seen "just absorb a giant ball of context and do the right thing the first time"

Exactly.

No more than sitting down and writing code before a product concept or spec or architecture comes out right the first time, or fifth.

Absorb the concept, make a shape of outcome, then a spec, then hold its hand to architect a series of iterations, either component by component or thin vertical slice or whatever combination lets you iterate in working increments...

Your brain, machine leverage. After all, it types faster than you. But it should type what you want.

You know what it should type, right? If you don't, you're gonna have a bad time anyway.

Re: Microsoft starts canceling Claude Code licenses

#444

Post by ryanhecht »

Anon1096 wrote: post #306
cameronh90 wrote: post #294

Earlier quoted context omitted.

Microsoft have historically tended to dogfood their own products. Obviously you want to be aware of what else is on the market, and use the right tool for the job -- but equally if you have a directly competing product, you'd prefer your org's telemetry and suggestions are directed towards improving your own software rather than your competitors'.

This was always a little weird to be because Microsoft internally is actively hostile to cross-org collaboration. If you worked in most of Azure you basically have 0 lanes of communication with someone from the Windows team and vice versa. Triply so for stuff like Kusto or Teams which you'd be dogfooding daily. I guess if there's a horrible stop the world bug it'd get surfaced through telemetry but normal user feedba…

> Microsoft internally is actively hostile to cross-org collaboration

The Copilot CLI has ushered in the beginning of a change in this dogma -- I've helped dozens of Microsoft engineers get access to GitHub source code so they can contribute to Copilot CLI! It's fun to subvert expectations when a Microsoft IC pitches an improvement and I can respond with "submit a PR!"

Re: Microsoft starts canceling Claude Code licenses

#445

Post by ryanhecht »

maxignol wrote: post #242

This might actually be clever since Microsoft dev will be longing claude code features and might result in copilot getting way better

That's what we've spent the last five months doing! Have you tried the Copilot CLI recently? We've onboarded loads of feedback from Microsoft devs who were switching from Claude Code -- I'm proud of how far the team has come! This announcement comes at a time where Copilot CLI usage has been greater than Claude Code usage at Microsoft for several weeks; we've been winning hearts and minds!

Re: Microsoft starts canceling Claude Code licenses

#446

Post by lumost »

apsurd wrote: post #399
lumost wrote: post #354

Earlier quoted context omitted.

This, I happily used the opus 4.6 fast mode to the tune of 5k for a project. The delivery of the project justified the 5k, if I only spent 500 but delivered the project 1 month later - I would have been in the dog house.

Your project cost $5k in tokens? How does that work? over what time? My understanding is that most developers are given pro max plans at $200/m and are expected to max that out. I've been getting by on the $200/year plan by smoothing usage continuously over time. The pay per use is for the API so does it mean you're using the API in a custom setup?

Maxed out the 200 dollar per month plan then incurred the overage. This was over the span of 1 month or so.

Re: Microsoft starts canceling Claude Code licenses

#447

Post by tharkun__ »

theptip wrote: post #436
tharkun__ wrote: post #432

Earlier quoted context omitted.

No sawdust is bad. But it's also bad if you cut all your boards into sawdust. Completely. Obliterated. No useful output, only sawdust. % of AI suggestions accepted vs. edited is also a BS metric that Anthropic et. al. like to push, similar to LoC, because they're large numbers and large numbers must be good, right? Well guess what, I have auto-accept on and then adjust after it's "done". And I do it by telling it wha…

> % of AI suggestions accepted vs. edited is also a BS metric I disagree. It’s a valuable metric if you are building an agent / skill infra layer. Think of it like error rate on your API. Green metric does not mean your system is healthy, but if it’s red you have an issue you definitely need to fix. Your example scenario is detectable in the non-naive implementation anyway; the o11y layer (usually OTel these days) tr…

Not the one down-voting you btw. Disagreeing is fine by me.

I would ask you tho: What incentive do AI vendors have to even try and detect this? It's in their interest to use the most naive interpretation, i.e. what my original comment mentioned, as it shows how "good" their models are, coz nobody ever changes much if anything ;)

Never mind that they really can't unless they're going "creepy mode". If I use Claude/Codex et. al. to agentically write something, then let the session just sit while I go about in my IDE changing things and then I commit and push, are you telling me that the vendors do or should track all of the changes made to the files they touched and report back to base what got overridden by me, the human?

Re: Microsoft starts canceling Claude Code licenses

#448

Post by 8note »

theptip wrote: post #390
bob1029 wrote: post #363

Earlier quoted context omitted.

I think quantifying tokens used is analogous to quantifying the amount of sawdust generated on a construction site. Churning out useful code quickly is not solved by using more tokens per unit time. Most non-technical leaders can grasp this one and are likely more interested in the strategic game theoretical dynamics that are being forced by way of implied token consumption expectations (competition between developer…

To understand the token count thing - spending tokens is necessary and not sufficient to demonstrate that you are adopting AI. Where we were 6mo ago is that a lot of big orgs realized they were behind, and needed some way of measuring if the tools were usable at all. No sawdust at all on your job site, and you can tell nobody is cutting wood. Now that tooling is more mature, you can measure things like % of diffs AI-…

> % of AI suggestions accepted vs edited

this has to be the worst metric.

anytime the llm wants me to read a diff of one file, im just gonna send it forward so i can read the whole diff

Re: Microsoft starts canceling Claude Code licenses

#449

Post by solenoid0937 »

arkadiytehgraet wrote: post #437
solenoid0937 wrote: post #435

Earlier quoted context omitted.

You have no substantive arguments other than calling people you disagree with shills. > so the anecdata does gradually become data. No, it does not. Countless social phenomena demonstrate how factually incorrect misconceptions spread rapidly. Frequency illusion is real and contagious. > How do you know that [they are not serving quants] Lots of ways to tell, if you weren't busy calling people shills. First, Anthropic…

Nothing to do with disagreement, I only call "Anthropic shills" people who are explicitly and shamelessly shilling for Anthropic. You still ignore the point that shilling adds bias to all your comments, so other readers have to actively keep it in mind to adjust for it. Stating that you are an Anthropic shill helps everyone around. And somehow you managed to be peddling LLM-hype shit so hard, that you are the only on…

> Yes, it does, it is literally the definition of data

No, feelings are not reliable data when frequency bias and misinformation exist. There is a reason most experiments isolate out bias as much as possible.

> Moreover, benchmarks (that matter) are not run continuously all the time.

So there's no data?

> What do you think different token distribution means?

You clearly did not understand anything I said. Stated simply: If you were being served a quant, you'd be able to tell by looking at the token distribution, latency, and TPS. You don't need to trust the labs' word for it.

> they pay you your salary, duh.

In fact, I get paid by a FAANG, though I do use Anthropic products heavily. Further, I don't really need money, I have more than enough. So much for reading my history.

> You are actively poisoning this forum

Your degenerate discussion - calling people shills instead of engaging with the argument, insulting them when your arguments are disproven, your inability to hold a rational debate that's not angry and emotionally charged - that is what is poisoning this forum.

Frankly, if you react this angrily and emotionally to a simple rational premise (that frequency bias leads to the perception of models being worse than them actually being worse), you're ngmi unless you're already independently wealthy.

I would recommend a therapist, it helped me when I had similar behavioral issues. (Claude is a great therapist, by the way ;)

Re: Microsoft starts canceling Claude Code licenses

#450

Post by arkadiytehgraet »

solenoid0937 wrote: post #449
arkadiytehgraet wrote: post #437

Earlier quoted context omitted.

Nothing to do with disagreement, I only call "Anthropic shills" people who are explicitly and shamelessly shilling for Anthropic. You still ignore the point that shilling adds bias to all your comments, so other readers have to actively keep it in mind to adjust for it. Stating that you are an Anthropic shill helps everyone around. And somehow you managed to be peddling LLM-hype shit so hard, that you are the only on…

> Yes, it does, it is literally the definition of data No, feelings are not reliable data when frequency bias and misinformation exist. There is a reason most experiments isolate out bias as much as possible. > Moreover, benchmarks (that matter) are not run continuously all the time. So there's no data? > What do you think different token distribution means? You clearly did not understand anything I said. Stated simp…

> Feelings

Nice gaslighting, Anthopic shill. No one said a word about feelings, only you (to derail the conversation). People reported their own experience and frustration with the model being unable to complete tasks they previously could. I said, get a better playbook before coming back. Or is it the best LLMs can do for now? Sad, then.

> No data

There is data, which you try to gaslight into being "feelings", Anthopic shill.

> Stated simply: If you were being served a quant, you'd be able to tell by looking at the token distribution, latency, and TPS.

Did you just repeat what you said before while ignoring the actual meaning of the words and my explanation of what YOU wrote? Is it what LLM told you to do, Anthropic shill? And you claim I have no substance. Maybe spend a week or so getting educated, before blindly copying and pasting LLM output, Anthropic shill?

> I get paid by a FAANG

Yeah, in your dreams maybe, Anthropic shill. I did read your comment history, and this is likely part of the story you try to build around your Anthropic shilling persona. Not a single fact that would prove that and believe me, I tried looking for it. Only endless claims of "I work at a FAANG" (no one who actually works here writes it like this).

> I use Anthropic products heavily

This is obvious, as 90% of your comments are LLM generated, Anthropic shill.

> calling people shills

Clanker, I called only you a shill, not people, tell your LLM to update its context. And I called you shill not because of any arguments, but because of your comment history unapologetically shilling for Anthropic and peddling LLM hype.

> arguments are disproven

You ignored half of my arguments, and for the rest you just repeated what you wrote before, not even understanding what the words you typed meant. Nice gaslighting, Anthropic shill.

> insulting

And you said you were not offended. Once again, Anthropic shill, being called a shill is not an insult. This is your fate, to be called an Anthropic shill, while you are on their payroll, astroturfing online communities with your LLM-bullshit peddling. Or do you expect being a propagandist to be a pleasant experience? People with no morals like you coming into this forum spreading their employer's bullshit deserve all the hate they get and more.

> you're ngmi. Hope you're already independently wealthy.

Your LLM outputs the same thing as in other comment for no good reason. Can't Anthropic afford good models for its shills, or is it the best SOTA can do now?

I would recommend you abandon this account, because it's now burned for all shilling intents and purposes.

Post reply on HN