I hope it's ok that I make this point

[PorterhouseMusic]

I think many on this forum have an "audio vision", sounds they hear in their head that they want to put in recordings. And everyone is limited/liberated by whatever tools they're using to get to those sounds. I use Logic. It has fake drums, sampled drums, and drum-like noises, and drum machines based on different players and styles. When I make percussion arrangements, I usually use all of those options willy nilly, and I feel like I'm in a candy store gorging myself. I recorded my bathtub running, then used that water sound through a vocoder plugin to make a "water running instrument." I play my saxophone into my SSL2+ and layer it in with fake brass instruments. Whatever I do, I'm chasing a sound and grabbing it the best way available. Personally the idea of AI creating melody, and performing vocals, is super offputting. Especially AI that's asked to perform like a well-known artist, that really creeps me out. But I can imagine someone who's chasing a certain kind of sound, even a specific kind of voice, using AI to achieve that, yet still creating an original piece of material that has merit. I'm too old/set in my ways/disinterested to ask the robots to sing for me, but I won't stand in the way of someone else using whatever tools they can to express what they hear in their head.

Absolutely.

 

[K-dub]

Don't over complicate it.

Whatever you do: If it sounds good, it is good.

However you get there.

AI is a tool. It is unlikely to sound good on its own because it has no clue how it sounds. It cannot judge itself for discernment of taste. It doesn't care. It does what it is taught to do and cannot make critical decisions about the quality of its work. It can only compare what it produces to known models.

I use AI. I use EQ.

I'm the one who makes decisions regarding what I find acceptable. I don't care much how I get there.