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/
Oct 16, 2026, 6:03 PM@nixCraft it's fine, he "understands the code" he submitted 😆
@wizbean this is how angry people get once they're vibesick ... I've seen this kind of crash-out several times
Oct 17, 2026, 9:50 AMLike a kid who gets sick from candy. Now this whole GenAI industry starts looking like the fast food industry to me.
You must admit a Whopper is useful!
Oct 16, 2026, 6:05 PM@nixCraft AI code assistance is best for mondaine tasks, like boiler plate code, solutions used by many programmers, like the stuff you can find on stack exchange and friends. It is less useful for very mathematical issues. At least this is my impression. Most likely, Claude could create my Python code faster than I do, but I usually don't know what I want to build.
I would love to see studies on how well these tools perform with code close to machine hardware.
@prefec2 @nixCraft Your impression is correct. I could be wrong about this, maybe there's a switch in there somewhere that says "Do deterministic things now, become calculator" but to my understanding the output will always have some level of randomness involved. Really not a great fit for deterministic needs. Like math.
@leontheloner @prefec2 @nixCraft there is a bunch of switches ("tools"). Basically, if the language model, reasons that the user is asking a math question, then the calculation is sent to a calculator app.
The most fucking wasteful amount of compute to open a calc app, jesus.
@leontheloner @prefec2 @nixCraft There is a 'temperature' setting that controls how random the output of AI models is, and its set very low for coding tasks. When choosing the next token, models ran with a higher temperature will pick from the top choices randomly, where low temp will typically go with the top choice.
I don't think I would count low temperature as being deterministic, but maybe somewhat less random.
@leontheloner @prefec2 @nixCraft We use claude for most coding tasks at my work. Engineers are still responsible for code they merge. Overall its maybe a 2x speedup. Reviewing PRs generated by Claude isn't really any different than reviewing a PR generated by my senior devs. Its correct a lot of the time, and the quality is generally good. Its not always correct and its still critical to have domain experts involved providing guard rails.
All of the ethical concerns are still absolutely valid.
Either AI is as incompetent as I have personally found it to be, in which case damn near everyone else is suffering from mass psychosis, and every software I use is going to be ruined.
Or AI is somehow actually good, in which case my days of solving problems with creativity and logic are numbered, I have no future, and my entire existence will soon be unnecessary.
Either way makes me want to jump off a f***ing building. Jesus F***ing Christ.
Oct 17, 2026, 6:08 PM@argv_minus_one @HeckinChonker @leontheloner @nixCraft I had similar thoughts, but maybe it's different oder something in the middle. And you still have to understand the problem so you can explain it in a prompt. My normal way of solving issues is to express my thoughts in code, fail, and by doing this better understand what the problem is. Had I typed in my problem directly in an AI, it would have provided a solution for my limited understanding of the issue.
On the rare occasions when I had a problem and could not find any solution, bots have proven profoundly unhelpful.
I posted about one such incident last month. Here's the thread: https://mastodon.sdf.org/@argv_minus_one/117215411079697884
I'm sure the AI bros would tell me I was just prompting it wrong, but…well, see for yourself. I described the problem with exemplary precision and thoroughness, and in return I got unmitigated garbage.
My email inbox is also daily bombarded with confidently-incorrect slop from my meat-proxy coworker, which has further soured my opinion of AI.
If it's this disastrously wrong this often about things I know, what *else* is it frequently disastrously wrong about?
So, seeing the likes of Linus Torvalds sing AI's praises has made me feel like I'm the only remaining sane man in a world gone completely mad.
@argv_minus_one @prefec2 @HeckinChonker @leontheloner @nixCraft You are not alone. To me it is obvious that the less competent you are at something, the more you feel the improvement by using AI, and everyone knows someone that is overexcited because of that boost and tries to evangelize everyone else as if he/she has discovered fire.
@argv_minus_one @prefec2 @HeckinChonker @leontheloner @nixCraft An additional problem with software is that even a catastrophic global failure like the one that happened with the Windows boot sequence some years ago has little to none legal consequences, as i.e. a bridge that collapses has, so people write crappy software or claim the authorship of software generated with AI without fear of the consequences.
@nixCraft Idk how Linus uses AI. I hate AI, but sometimes is kinda useful ONLY IF it's used locally hosted, like with Ollama.
I'm more concerned about environmental impact than stealing knowledge, although both are major concerns
@epistomai there are better tools than ollama for local AI. https://sleepingrobots.com/dreams/stop-using-ollama/ explains why.
@nixCraft But also… “Amanda Caswell is an AI journalist, certified prompt engineer, and technology commentator whose work and expertise have been featured on Fox News and CBS News […] TNS owner Insight Partners is an investor in: Anthropic”
(and the article is sprinkled with weird ambiguities. Wouldn’t surprise me if it was LLM-summarized or LLM-rewritten)
Oct 17, 2026, 7:38 AM@nixCraft And how did it go with humans? I've read this is a problem humans tried to fix multiple times and failed. Should we abandon using humans becaus they failed? Are they the wrong tool for the job? I don't really get what point you are trying to make. You wrap it up so it seems very one-sided.
@Chapz @nixCraft You'd rarely get the point of most things out of one post unless they were specifically trying to make one, but if you read multiple posts by NixCraft you might get a larger picture of their general stance and opinions regarding LLMs. For one comparing LLMs and humans in the manner that you are doing is frowned upon.
Oct 17, 2026, 12:20 PM@707Kat @nixCraft I know, it's sad to see this in tech. I always thought the tech sphere was objective and critical, but fair... but it seems that more and more of them have a very fixed view, no will to even consider or hear out the other side, so it ends up just being... Well, in lack of better terms, propaganda, because its wrapped so that it seems like objective news or discussion instead of simply a very private, at times extreme, opinion.
@Chapz @nixCraft This "fixed view" and "extremism" as you put it, comes from having things they care about destroyed by the people developing the technology you think they should have a neutral opinion about.
I've seen it and even if I can see some use cases for the tech, I'm taking a stark position against it for a mural of reasons. Many have to do with ethics, harms and societal factors. Technology isn't as neutral as we make it out to be. The way it's implemented, used and developed matters.
RE: https://mastodon.art/@707Kat/117303119585307183
Perhaps this talk by @Felienne might help put it into a context where it makes more sense.
A lot of people now opposing this technology used to be hopeful about machine learning and automation, many still are, but consider what we're seeing with LLMs and hybrid implementations harmful. We all have our reasons, but as a mass of overlapping greys it might look black/white from afar. Zoom in and you'll see the nuances.
Also see this thread by @glyph: https://mastodon.social/@glyph/117446111212505873
Ah, love the tiny bit at the end where it says their owner are an investor in Anthropic.
It's good that they are self-reporting affiliation though.
@nixCraft 🤦 WHAT THE HECK, MAN... What's wrong with you, Linus?
@nixCraft Wasn't Claude itself propietary? Not even Deepseek with a libre corpus...