A Voice Interface Should Ask More Than “Continue?”

A thought experiment for a voice-cloning interface: one broad “I agree” button can blur several different choices—creating a voice model, editing with it, and sharing the result.

I’d turn that into a visible sequence: Choose the permitted use → preview what will be made → confirm before sharing. Keep each choice specific, and make the next step easy to pause or revisit. That gives the interface a clearer creative brief: not just “make this sound,” but “make it for this purpose, with this person’s say-so.”

This is a design direction, not a claim that interface controls alone establish consent or settle legal requirements. The useful test: can people explain what they’ve agreed to before the voice is used?