Cool idea — make a music coding language for composition and use frontier LLMs to generate it, then marry it up with a web DAW for music generation. They’re super anti-generative audio model for obvious reasons.
This approach is smart because in theory the music quality will scale with the LLM’s intelligence. Zero compute overhead for them
Cool idea — make a music coding language for composition and use frontier LLMs to generate it, then marry it up with a web DAW for music generation. They’re super anti-generative audio model for obvious reasons.
This approach is smart because in theory the music quality will scale with the LLM’s intelligence. Zero compute overhead for them