generative ai (subtype: programming with LLMs)
@glyph I can't take anyone who is programming with an LLM very seriously when it comes to rights management, if they're not being meticulous about firewalling that stuff from the actual changesets they write.
It's just... can I trust you to give a shit about IP for *our* stuff if you're hooking our stuff up to the entropy machine? I don't think I can trust you to do that. And a lack of trust is a collaborative toxin…