2 docs tagged with "Audio"

What can become what? Every modality is both input and output. The matrix of possibilities — not a list of tools.

When a text prompt can produce a full song with vocals — what does a musician do that a model cannot?