Been using RIP X to deconstruct stuff for sound design, which is interesting and uses LLM tech. I can see how it's a next generation of sampling but typing prompts for making music doesn't excite me if I'm honest. We were given some software at work recently that enabled us to use fake voices to sing songs typed in but it left me cold. We haven't used it at all (although we LOVE the other app the company gave us).

I enjoy manipulating audio but with a music collection and a few musician friends, I've not needed to ask the computer to create something 'new' the computer has only listened to more music than either than us, it doesn't understand emotion or get goosebumps hearing someone sing.

Re. Outsourcing, what about collaborating with musicians? Maybe I'm mad... Trying to think it through, rather than knee jerk reaction, which admittedly I just did. Do you think music, more broadly, is getting better, aside from AI? There's fewer chords and a reduced melodic range in popular songs and it's like watching a chart of dollars in circulation since 1971. When everyone can make a song, a song will no longer have value, which is currently Spotify's position. Maybe it should be called Fiat Music....

I imagine a world where Ad Execs at g00gle suddenly think they're Bob Marley and it gives me the creeps haha.