Veo 3

“I've noticed that the model will often do something weird if you don't put in enough text to fill eight seconds of audio. For example if you make it sound like it's supposed to be a street interview, and it's an 8:02 video, and there's only two seconds of audi…”

“Some of the more developer-oriented, like, API platforms, like Fall or Replicate, are offering generations where you pay per video. It's priced around, like, 75 seconds. 75 cents per second today.”

“And what's very different about it is it generates audio natively at the same time it generates video.”

“We obviously made a choice that we want VO three and Gini three to be separate projects this year, right?”

“VO three is clearly a higher quality threshold than Gene three, right?”

“But my hunch is it's largely VO three, because traffic to that To the Google Labs spiked 15% in the month VO three was released, which is really exciting.”

“It's kind of like the VO three moment when I don't think it was intuitive to people that video would be cracked necessarily by bringing audio together with video in the same place. And that ended up being the thing that made AI video go viral.”