@david_chisnall@infosec.exchange I'm curious what you all have in mind when we talk about "a strong no-LLMs policy". Personally, I picture a stance against any use of LLMs on ethical grounds, whereas NetBSD and other projects only talk about restricting code that's generated without human involvement. There's nothing about using LLMs to dig into bugs and vulnerabilities, or about people rewriting LLM generated code and passing it off as their own. Now more than ever, it feels like sending patches is going to become a question of trust between committers 🤔