Mine is the gap between having an idea and having something I can actually show people. I can sit with a melody in my head for weeks and by the time I've built it into a proper demo the energy is completely gone. Curious if this is a universal thing or if some people have actually cracked it.
’ve been experimenting with Suno and Udio lately, and while the "magic" is cool for 5 minutes, I find myself getting incredibly frustrated as a producer. The lack of control is soul-crushing—I have a specific melody in my head, but the AI just does whatever it wants. It feels more like gambling than composing.
To me, a "real" tool should follow my MIDI and my arrangement, not just spit out a finished .wav file I can’t stem out. I’m curious where you guys draw the line? Are you finding any AI tools that actually fit into a DAW workflow and respect your creative intent, or is it all just "prompt-and-pray" for now?
I use Suno and Udio for brainstorming, and the tech is crazy. But let's be honest: typing "synthwave pop" and hitting generate 50 times until the algorithm spits out a banger isn't really producing music. It’s pulling a lever on a slot machine.
Everyone is arguing about whether "AI music has soul," but the real issue is the absolute lack of creative control. You can't easily tweak a specific vocal run, fix one bad chord, or build an arrangement layer-by-layer. You just get what the black box decides to give you.
I feel like the future of AI music shouldn't be one-click vending machines. It needs to be actual tools integrated into DAWs (like directing an AI vocalist or converting MIDI to realistic instruments) where the creator actually drives the car.
Am I the only one who feels limited by the "prompt-and-pray" method?