Note too that *even if* all the Courts decide artifacts output by an LLM-gen-AI aren't copyrightable (we're far from that now, and no one knows for sure what will happen), it *is* certain that human-performed modifications to a yet-uncopyrighted work *are* copyrightable.
I recommend licensing such under a copyleft license.
Anyway, per elsewhere in thread, LLM-gen-AI artifacts always require human improvements to be useful to FOSS.
Cc: sounddrill :verified_dragon: (npub1s55…up8m) NLnet Labs (npub1tra…q9a6)
[@alexband](https://hachyderm.io/@alexband )
