jesus christ… it’s been said over a million times, the development of the linux kernel is not some fucking 5 star github repo. it goes through a shit ton of review, it’s seen by a lot of eyes. a small piece of text that says it was assisted or not by ai is irrelevant if the code is good or not, which is the work of maintainers to decide.
but why remove it ?
Because rules have to be enforced, and investing time and effort into detecting and punishing rule-breakers would be a pointless endeavor when, disclosure/AI or not, all code has to be reviewed anyway.
According to Linus it’s free advertising for the AI companies (wowe look at how much code Claude (partially) committed buy le Anthropic now!) and also people don’t disclose all their other tools either so this is just an exception which can be removed. And I bet there’s going to be a ton of LLM based reviewing on all code that’s submitted to double check for potential bugs anyway.
Anyone who is allowed to work in the Linux kernel isn’t going to (be allowed to) push random slop.
maybe we should ask the maintainers, since it’s they’re the ones where this disclaimer could be useful or not. :)
It’s been said a billion times: you don’t get to decide what I find relevant. The only reason these clowns want to hide their behavior is because they’re ashamed of the criticism. There is no other reason to hide attribution, your patronization notwithstanding.
the developers of the linux kernel, people who have probably been coding for at least 20 years, are clowns.
They why try to hide it? If AI is not problematic at all, why do 90% of AI bros hide it and refuse to disclose and when they do are often extremely cagey about it?
For these types of projects, it is standardized what compiler they are building against, kernel /proc allows you to see what compiler was used.
Why does AI, which Linus himself has compared to compilers numerous times, get a complete pass and absolutely nothing has to be specified about them whatsoever (in the anti-AI-Disclosure advocation space of course)?
The coder doesn’t disclose what compiler he tested against. Doesn’t say what OS they where using or programs.
90% of AI bros hide it and refuse to disclose and when they do are often extremely cagey about it
source?
i have seen the complete opposite, just because some post doesnt start with THIS WAS DONE BY AI doesn’t mean someone’s trying to hide it. i recently saw a post where they shared the repo, you could clearly see claude as a contributor in the repo and many of the commits were done by it. but apparently they were trying to hide it. these kinds of complaints sound that came from people who have never used git, let alone touch a single line of code.
the comparison is not correct. a compiler is deterministic, asking about it is asking about a standardized tool. ai is stochastic, asking about it does not guarantee reproducibility, which is the whole point, not suspicion about if the compiler did or did not translated it properly into machine code. furthermore, precisely since ai is stochastic, even if you knew which model and version you used, you’d get different results, which needs review anyway.
Eh, I see every 4th project post in open source communities being a cool looking project with no Claude contributed, no AI disclaimer, no agents.md, however when you look at the user’s commit history on the git service, it is
- 10-20 contributions (mostly issues) before 2025
- maybe a couple readme or documentation commits in total
- mid to late 2025, user starts committing 500-5k lines per day, every 3-5 days, evolving from large commits to broken up 100 commits today (following the 2025-2026 timeline where AI started doing the same)
- exact same commit structure that AI uses, and filled with jargon
- no comments at all on the post, or few comments but never responding to anything technical about how the program works.
Probably some confirmation bias going on on both our sides.
I don’t understand. Being moral is irrelevant?
Completely this.
As long as the code is good,. does it matter where it came from?
Yes, fuckstick. That’s the whole point to a lot of people.
I swear, the level of smugness of AI true believers is totally inappropriate.
Also, are we forgetting that AI models were trained by stealing from a fuckton of people? How are folks ok with this? Why is the discussion solely about how well the models work? Where is everyone’s moral compass? And why would they defend a tool explicitly built to replace them? Not to mention how inefficient it is (energy and money-wise).
Genie is not going back into the bottle.
Genie requires a fuckton of food. If you don’t feed it, it will die.
we’ll be waiting in the grown ups room while you figure out how to answer the question “why?”
This decision baffles me.
As some one here summarised it: it is to stop AI companies from getting free advertisement and endorsement, the kernel itself goes through an unbelievable amount of scrutiny so the disclosure requirement was irrelevant
I guess, as he suggested, it’s time to fork it.
the hardest part is to come pu with a good project name and mascot.
It would be interesting if this results in Linux being forked or perhaps even more interesting developers and users opposed to AI moving to GNU Hurd or any of the BSD variants.
good luck with that. software development has changed and it will never be the same. you will not be able to achieve the level of security required without ai assistance, you can’t beat a machine. and this is something you’ll see everywhere and is the tip of exponential curve.
Atleast NT is not vibecoded yet, not sure about BSD.
NetBSD has a clear stance on no ai.
As much as NetBSD takes a stand against AI, it’s not banned. They just said AI code has to be signed off by the core team.
LLM code can be committed with “written approval by core.” https://www.netbsd.org/developers/commit-guidelines.html#%3A~%3Atext=written+approval
Also Haiku OS and Ironclad OS
NT isn’t open source too. So nobody cares about NT.
Good
