2026-10-16 17:38 UTC
Replies (33)
-
@nixCraft@mastodon.social 2026-10-16 17:41
Just so you know. Linus who is adopting AI tools. Here is how it is going right now. Linus Torvalds let Claude Code fix the Linux page cache. Then reviewers found a race https://thenewstack.io/torvalds-claude-code-kernel/
-
@richardibbo@fosstodon.org 2026-10-16 17:43
Whilst it's true I've had some good conversations with Linus I don't agree with his ideas about AI. Come back to bite him ? Yes. I've started to cal him Baldy Baldy... @nixCraft@mastodon.social
-
@chessert@mastodon.online 2026-10-16 17:44
@nixCraft@mastodon.social Sigh. Guess it's time to start playing with the BSDs again, "just in case"?
-
@ajayiyer@mastodon.social 2026-10-16 17:54
@nixCraft@mastodon.social I think it is scary when someone at that level makes false comparisons between a compiler and an LLM. He's loves his "I'm just an engineer; I just build tools" philosophy. Surely he should know that not all tools are the same. Both "asbestos" and "mineral-wool" stop fire, but one is EXCEPTIONALLY more dangerous to humans than the other.
-
@derderwish@social.chaotikum.org 2026-10-16 17:58
@nixCraft@mastodon.social At least, the editor is not destroying the Internet and the environment while stealing intellectual property from small creators…
-
@IainWilliamMcclain@mastodon.social 2026-10-16 18:15
@nixCraft@mastodon.social I use AI to scan my code and to compile apps. Does that mean I don't know the code? Nope. I just know it takes AI 30 seconds what takes me three days.🤣
-
@Stefan_S_from_H@mastodon.social 2026-10-16 18:24
@nixCraft@mastodon.social there's hope that Claude suggests rewriting the kernel in C++.
-
@biglinter@mastodon.social 2026-10-16 18:29
@nixCraft@mastodon.social "nobody in their right mind"... what happened to the ring of maintainers? well, then itll take the time it takes to put in the code... seriously, just because youre flooded doesnt always mean youll have to open the flood gates... i remember linus saying that someday linux will become obsolete, so we are steering right into that i guess
-
@freevolt24@mastodon.social 2026-10-16 18:38
@nixCraft@mastodon.social Two things. First, most of the time, discovering problems are easier than building complex systems, and the same goes for ML/AI models, of course. Second, ML/AI is "bound" to make mistakes, and ML/AI scientists have been saying for years that you need mistakes for ML/AI to be useful (because otherwise, it means overfitting). I understand that ML/AI tools can be helpful for productivity, but it's guaranteed that it'll make mistakes (if not intentional).
-
@syntaxseed@phpc.social 2026-10-16 18:56
@nixCraft@mastodon.social Comparing deterministic tools to #AI is wild. Yes, I would create rules about not using compilers & editors if they gave different output each time I used them, occasionally gave wrong output, sometimes gave SLIGHTLY wrong output, and were just running predictions based on my past inputs rather than following sets of rules. #Linus
-
@jesse@social.jelliefrontier.net 2026-10-16 19:13
@nixCraft@mastodon.social Uf what a bad take. Not all tools are created equal. “No sharp knives in here because they can be misused for violence.” “Oh, so I guess we can’t use straws or spoons either?! For that matter, this food shouldn’t be on a plate!”
-
@jeantranscene@mastodon.social 2026-10-16 19:18
@nixCraft@mastodon.social I may be wrong about that but it seems to me that the debate shouldn't be about whether an AI wrote the code or not but about code quality and whether trusted humans validate it. If your contribution goes in a queue that is reviewed at human pace and you have a storage quota for your contributions there's no way slop of any origin get into mainstream. AI or not AI becomes irrelevant. But maybe I'm overlooking something.
-
@mxchara@cyberpunk.lol 2026-10-16 19:21
@nixCraft@mastodon.social I love it when techbros begin their argument with "I don't understand" as a rhetorical stance because it's always true in a way: they don't understand.
-
@jeantranscene@mastodon.social 2026-10-16 19:23
@nixCraft@mastodon.social It should be easy to maintain stats on the slop producers regardless of origins and slow the worst offenders even further somehow. There's always a human motivation behind any AI production. That shouldn't be too hard to control I think.
-
@loadhigh@bitbang.social 2026-10-16 19:24
@nixCraft@mastodon.social Man, that last paragraph, what a nasty guy. My respect goes to the maintainers who manage to handle working with an asshole like this. I know I wouldn't be able to do it.
-
@Gjoel@mstdn.dk 2026-10-16 19:41
@nixCraft@mastodon.social This post is actually a pretty good reply to that, even if it is on LinkedIn... https://www.linkedin.com/pulse/ai-tool-rich-hickey-kjl8e Use the tools you want - but don't use automation where your brain should do the job.
-
@MuhammadFreeSoftware@fosstodon.org 2026-10-16 19:48
@nixCraft@mastodon.social It's time we replace Linux with some other kernel by the way, it would be great if it was the Hurd
-
@tmw@ioc.exchange 2026-10-16 19:51
@nixCraft@mastodon.social i'm sick and tired of borderline-eugenicist freakshows like linus and other techbros who are so obsessed with "intelligence" that they dismiss anyone and everyone who doesn't appear to display it in a way that affirms their own egos. people like this shouldn't be leading or managing anything that involves other people.
-
@BeKindRewind@ioc.exchange 2026-10-16 20:30
@nixCraft@mastodon.social I feel like AI is spewing out code it was trained with off projects that people aren't aware of, there's no attribution of that code. So what happens if you take copywritten closed-source project code and add it to your open source project without knowing where that code snippet came from? Legally this sort of thing might kill open source projects as a result. Is Anthropic going to state all code generated is open source and copyright free? I don't think they can, and that is going to pollute the kernel? AI code is extremely problematic in this sense.
-
@wcbdata@vis.social 2026-10-16 20:41
@nixCraft@mastodon.social The thing is, he misunderstands the technology entirely. it's not a tool. It doesn't make coding easier; it generates stuff that looks like code and sometimes runs. It doesn't write documentation; it generates text that resembles documentation and sometimes reflects the way the code runs. This isn't like a table saw making accurate cuts easier, it's like paying a million people to repeatedly dump glue and sawdust together and then hand over the result that most resembles a board.
-
@RickTimmis@mastodon.social 2026-10-16 21:18
@nixCraft@mastodon.social I disagree, Linus is right. When I was a kid we used to write Hex for the 6809, then me and a couple mates wrote an assembler, all the way to today, one tool on top of another. I trusted my assembler, just like I trust my go compiler, and my AI coding agent. But I still write designs, tests, and docs. Well actually I read them more than I write them.. No matter what tools are used, the developers are always responsible for the output. You can always git revert 😁
-
@paelnever@masto.es 2026-10-16 22:47
@nixCraft@mastodon.social Sorry i don't see a word in his post about "look at 100s PRs per day" What are you talking about?
-
@paelnever@masto.es 2026-10-16 22:50
@nixCraft@mastodon.social Sure, that's the wet dream apple and microslop ever had, to get rid of Linux project for good and convince people to develop code with a license they can vampirize.
-
@remixtures@tldr.nettime.org 2026-10-17 00:12
@nixCraft@mastodon.social Where have all anti-AI zealots have been in the last year? https://asdlc.io/patterns/adversarial-code-review/
-
@remywhisker@mastodon.social 2026-10-17 00:33
@nixCraft@mastodon.social Last I checked compilers don't spit out random assembly. I seriously can't believe he is that ignorant about how AI works. What an absolute disappointment
-
@senslayer@mastodon.social 2026-10-17 00:45
@nixCraft@mastodon.social im normally on Linus's side when it comes to kernel ai policy (I for one am glad to see AI was able to find kernel bugs) and agree theres some place in the kernel for (ethical) AI. But I think he's mssing the fact that tools like compilers, git clients, rust, etc. were made to reduce footguns, making it easier to develop software, especially on a public codebase.
-
@mahadevank@mastodon.social 2026-10-17 01:24
@nixCraft@mastodon.social I can say from practical experience that you are correct. The problem with AI output is that it is by design plausible, so it looks like it's right, but it hides subtle issues which are not apparent at first glance. You would not expect this since human code in general would not contain such errors. I had solid problems with a developer I work with, who finally realized the problem and started cleaning up the codebase and verifying it.
-
@hszakher@mastodon.world 2026-10-17 07:58
@nixCraft@mastodon.social my immediate thought is "...Pride comes before destruction, and an arrogant spirit before a fall." (Prov. 16:18). setting himself as a reference, boasting of "intelligent people" he knows, calling people's concerns as "utterly stupid" without even recognising their valid points of slop, while using false equivalence between a tool (IDE) that 'show' what a human wrong, and a tool (AI) that traps 10 monkeys randomly striking keyboards to generate a plausible-looking code...etc. #shame
-
@rythur@mastodon.social 2026-10-17 11:57
@nixCraft@mastodon.social I fear sincerely that over-use of these tools may lead to a Tower Of Babel situation across many problem domains. I can't fully disagree with all of his statements, though.
-
@rodolforg@mastodon.online 2026-10-17 12:29
@nixCraft@mastodon.social he stupidly compares deterministic tools with statistics ones. Wow
-
@Photo55@mastodon.social 2026-10-17 13:54
@nixCraft@mastodon.social he is however very clever and practiced at this. And his explicit expectations seem appropriate. I would not, however, generalise from this. Or to it.
-
@sundaymidnight@mastodon.social 2026-10-17 14:04
@nixCraft@mastodon.social Let me explain the problem: before, you'd double-check before sending a PR. Now there's no incentive to review it. Your girlfriend's calling, you already have the code, so you just say: "Ah, it's fine... if there's a problem, they'll catch it anyway". On the other side, someone receives hundreds of commits and just hits approve, because it's humanly impossible to review them all.
-
@cocaine_owlbear@retro.pizza 2026-10-17 14:33
@nixCraft@mastodon.social I'm just gonna say that this is what we get for trusting one person to be in charge for this long. Basically, Linux has been under the thumb of a helicopter parent, and we all went along with it because by and large, Dad was a good guy. But now dad's hanging out at a bar that allowed in a Nazi, and he doesn't care as long as the beer keeps coming. Hell, he's thrilled because there's even MORE beer. Sure, it's Schlitz's and Natural Ice, but hey - beer is beer, right? And *that* my human friends is how you mix a metaphor. Owlbear OUT.