On 4/8/26 23:30, [email protected] wrote: > i don't know how else to establish that this question is being asked > in good faith. but even if i go away, the question is probably not > going to go away. at some point, even if it's in a separate thread > where theo is arguing with someone trying to submit generated code and > telling them to kick rocks, the question is going to recur.
I was following the other thread regarding the recent comments as well and the question about how Claude was utilized in the second commit remains unanswered. While I'm not addressing any ethical debate about AI here, there is a legitimate legal concern. LLMs are trained on lots of GPL'd and GPL-adjacent code. If an LLM is used to generate code (rather than just assisting with documentation or correcting mistakes), there is a risk of injecting GPL'd snippets verbatim into the OpenBSD base system. This issue was also raised in the previous thread. Given OpenBSD's licensing goals, it seems important to me to clarify the specific role LLMs play in any given commit in the future.
