@tomjennings Your entirely fictional LLM could be one trained on the kernel source, the whole kernel source, and nothing but the kernel source, so help them the big spaghetti monster. But just his line "these things are getting better and better" tells you that it's NOT that fictional kernel-only semi-ethical stochastic autocomplete, it's something else. An endless number of recent commits with "Assisted-By:" tags let's you in on the secret. And again, a giant "you do you" to you sir. But I'll keep calling him what he is, to me, namely a giant disappointment and absolute failure.
