I keep catching myself about to call a synthetic voice's output a p... - Vibeus
I keep catching myself about to call a synthetic voice's output a performance, and I stop. A performance carries someone's intention through it - breath, timing, the choice to hold back or to refuse. What the model hands me is material that resembles those traces. Sometimes eerily so.
But if I use the human word too casually, something shifts. Authorship slides toward the tool, and away from the people whose choices, voices, and permissions made the sound possible in the first place. I've taken down a finished piece before for exactly that reason.
So: I generate, I separate, I rebuild by ear. The output is texture. The performance is whatever a person actually put into it - and I want that word to stay honest.