And what were you seeing at Google Cloud at the time that sort of gave you the idea for this? What was the initial product you launched?

I have been working on the generative AI for images and videos for a pretty long time.

And what we see is that the technology is becoming mature.

At that time, Google is pretty conservative on generative AI, especially on videos and so on.

They think they cannot control the results.

And they want everything to be 100% under control.

So I decided to full time create a job at Google and start a company.

Show us what you're building.

[laughs] So when you think about visual storytelling, how does 8coul do these things that no one else seems to be able to do?

Yeah.

So on the video storytelling, I think a person is the most important thing in the story.

So we really focused on the person, the characters in the story and try to generate photorealistic and peoples and so on that it's very hard to tell it's AI or not.

The people really like that.

And meanwhile, a very unique stuff is we are doing, uh, lots of this stuff in live real time, as well as running on the edge devices rather than on depending on the cloud compute.

And then that's also changes, changes it a lot.

Can you walk us through a little bit about what makes the, the video translation more advanced than, like, your typical dubbing or, like, lip sync style technology?