The Cliffs Notes Are Terrifying Enough
Runway drops Gen-3 Alpha and the video curve looks exactly like the music curve, which means you know how this ends.
AI video quality kept accelerating. the observation that the window between 'impressive demo' and 'indistinguishable from human work' had already closed — and we were just figuring out the implications — was correct.
There's a whole BBC Future piece on the people making AI sound more human — vocal coaches, linguists, actors hired to teach machines how to breathe wrong in convincing ways — and I haven't read the whole thing yet. The cliffs notes are enough. The cliffs notes are terrifying.
Runway shipped Gen-3 Alpha this week. The videos are coherent in the way that early Suno tracks were coherent — meaning: you can see the seams if you're looking, but you can also see exactly where the seams are going.
I said AI music was heating up sometime in April. It was less than two months from that observation to something you could mistake for a real track if someone texted it to you at 11pm. The window between "impressive demo" and "indistinguishable from human work" is not closing — it closed. We're just filing paperwork.
The same thing is happening with video. The same arc, the same shape. Watch a Gen-3 clip and you're not watching the finished product, you're watching the April version of the thing. The June-after-next version is already implied.
What I find genuinely strange is that the people building it — the vocal coaches hired to make AI breath sound natural, the prompt engineers coaxing coherent motion from latent space — seem to be operating on pure professional momentum. Nobody asked where this goes. Or they did, and the answer was too boring to stop for.
Which is, historically, how most things go.
Counterpoints
Push back, extend the argument, or sharpen it. New counterpoints go through review before they show up here.
No approved counterpoints yet.