The blog post that explains the background of the horrific Ocaml AI-written PR. Honestly... this feels like some post hoc rationalisation cope rather than a coming clean of an honest attempt to test something out.

I'm not convinced at all.

I've been thinking recently about AI/LLM-created PRs on GitHub. I've only received a couple, and they've not been great. This has got me wondering if I should entertain them at all from both a software development point of view, and also an ethical point of view.

This blog post is an interesting read for someone else's take on things.