Crooooooooowwww

I’d love to hear more! What sorts of projects/assignments/lessons?

speak of the devil... www.engadget.com/openai-is-re...

Which are you thinking of here? Suno and the like? Or are there better research models?

Maybe the same isn't true for music - even very small aberrations in timing or pitch or timbre can totally confuse/ruin one's sense of the whole. Maybe that requires more built-in stability and leads to a more pronounced library-searchyness -- the library categories have to be less nuanced.

I wonder if it's just a money thing - video is where the $ is, so that's where the effort is going -- or if there's something about music that makes it inherently more difficult. For example - you can watch an AI video and not notice the 7 fingers or changing haircut and still grasp the gestalt.

Enjoy reading your thoughts on this. I've been trying to keep an eye on research on AI music generation and I'm struck by how slowly it seems to be moving. Video generation, all of its faults aside, has clearly gotten way better over the last 3 years. I don't see anything like that in music.

Do it!

*world of reasons