Most people think of AI video as "type a prompt, wait 30 seconds, get a clip." That's one way to do it. But what's super exciting and almost feels like another side of this that doesn't get enough attention: generating video in real-time. As in, live, like 30+ frames per second while you're watching.
I've seen people in the first edition of this program doing some crazy, creative stuff with this (and world models, too!), so I'm curious what they end up building this time.
Most people think of AI video as "type a prompt, wait 30 seconds, get a clip." That's one way to do it. But what's super exciting and almost feels like another side of this that doesn't get enough attention: generating video in real-time. As in, live, like 30+ frames per second while you're watching.
I've seen people in the first edition of this program doing some crazy, creative stuff with this (and world models, too!), so I'm curious what they end up building this time.
these autoregressive models are wild