Interesting, however, how much decision-making is still done by the human user. You are the one who chooses the video avatar; you are the one who chooses the kind of voice that will be on the video. Even when you generate a custom avatar, the site gives you four choices to pick from. You still get to exercise considerable creative control. I see it as analogous to how a composer of orchestral music sends a composition out into the world, without having full control over what the individual instruments in the orchestra will sound like (as opposed to rock music, where every aspect of the final recording is decided by the musicians in the studio).
I played around with it a little bit and got this version of Allen Ginsberg reading an excerpt from part 2 of "Howl", it's absolutely atrocious: https://studio.d-id.com/share?id=967387b1d5229f8f1b045faed9a43998&utm_source=copy
Interesting, however, how much decision-making is still done by the human user. You are the one who chooses the video avatar; you are the one who chooses the kind of voice that will be on the video. Even when you generate a custom avatar, the site gives you four choices to pick from. You still get to exercise considerable creative control. I see it as analogous to how a composer of orchestral music sends a composition out into the world, without having full control over what the individual instruments in the orchestra will sound like (as opposed to rock music, where every aspect of the final recording is decided by the musicians in the studio).
I'm mostly working with visual AI. here's a set of illustrations for Radiohead's song "Fitter Happier" that I made a while back: https://www.ruins.blog/p/fitter-happier-illustrated