We're Screwed - Suno v4

  • Thread starter Thread starter Monkey Man
  • Start date Start date
I've been watching progress in this area with a sense of dread.

Right now with v4 I'm at "peak dread".

What I've been through for the past 30 years just to get the music thing happening again is insane, and now I'm trying to resist wondering, "Why the Hell did I bother?".
 
Monkey Man said:
I've been watching progress in this area with a sense of dread.

Right now with v4 I'm at "peak dread".

What I've been through for the past 30 years just to get the music thing happening again is insane, and now I'm trying to resist wondering, "Why the Hell did I bother?".
Click to expand...
remember, it can only imitate, it doesn't innovate.
 
IMG_6602.jpeg
 
Monkey Man said:
Yeah, but the vast majority (99.x%) of music being consumed out there isn't innovative. :dunno:
Click to expand...
Truth. Artists have been following the same cookie cutter template for many years now, what is it, I- V - vi - IV? Not surprised AI could easily get a foot in the door there. The surprising thing for me is graphic design, I think that's basically completely gone at this point.
 
Erock said:
how so? It's still all based on what it can consume. :dunno:

edit: maybe I should have added "yet" ?
Click to expand...
generative AI takes information, rules, etc., and can synthesize results from LLMs, training data etc. For example the music genres have rules, characteristics, etc., which it can dynamically combine to generate a new result.

It would be like a human combining two or more genre music elements to create a "fused" new genre based on those existing genre patterns.

FWIW, I've been working in AI for a few decades.

edit: what is limiting is the amount of data, rules, training, prompts, etc. but the knowledge bases are growing in size, facts, data, complexity, rules, so will the AI solutions...it's minimally creative, like a preschooler, but still creative
 
rsm said:
generative AI takes information, rules, etc., and can synthesize results from LLMs, training data etc. For example the music genres have rules, characteristics, etc., which it can dynamically combine to generate a new result.

It would be like a human combining two or more genre music elements to create a "fused" new genre based on those existing genre patterns.

FWIW, I've been working in AI for a few decades.

edit: what is limiting is the amount of data, rules, training, prompts, etc. but the knowledge bases are growing in size, facts, data, complexity, rules, so will the AI solutions...it's minimally creative, like a preschooler, but still creative
Click to expand...
Damn. So what does this mean with quantum computing around the corner?
 
You must log in or register to reply here.
Back
Top