Grant

It looks good.

The the problem is I was doing it on my work account and I didn't have my API fired up so I but I have my personal account here and I did two different versions so I could show you at the end.

Yeah, I think it is.

Yeah, if we look at the code really quick just to see.

Because we could even tell it to do maybe the via flash.

It's already using fast.

But this is basically what I was going for, is every time you write a sentence, you can then see what that sentence looks like.

And that's not possible without everything that you've built here.

Yeah.

So to very easily put it together, to have it easily, without me having to set up all the API keys to VO and Imogen and all this stuff, or Nanobanana now, it just does it for you, which is the coolest thing about AI Studio, I think.

Spelling will soon be a lost art, unfortunately.

Do you find that you're iterating mostly with flash now?

You can say like, please don't make it whispering.

Make it a calm, peaceful voice.

You just want to have the option to do whichever one is most convenient in that moment.

I think that's a great way to do it because sometimes the friction is the typing and trying to put it in some rigid structure.

So if you can just talk through what you're trying to get across, then you can iterate.

And especially with Flash, if it just can spin up something super quickly, 64 seconds, I mean, that's great.

That's a good call.

Is there anything that you could foresee or predict on the horizon that could lead to that time latency