Logan Kilpatrick

And that's the way we now, we, I think around like IO this year, um, or even it might've been even before that we, we released this, um, this build tab on the left-hand side of AI studio, where you can go in and actually have the model, like build you an entire app.

Um, and for a lot of the things, like it'll build you an app that's in some way powered by Gemini, which is actually really cool.

Um,

And so I spent a lot of my time, again, we're in this weird transitory period where I know how the product is going to evolve.

So I'm using the current product and it makes it, it's hard because I'm like, ah, I want the future version so badly.

So I have to like pull myself out of using it too much.

But it tends to be like this,

prototyping of AI experiences is how I spend most of my time.

It's like, ah, wouldn't it be cool if you could do this thing with AI in a product?

And AI Studio is awesome because like I can just put in a really simple prompt and it just like is connected to my Gemini API keys and it's connected to Gemini and all of our other models by default out of the box.

So it makes that iteration loop and process much more simple.

I don't need to worry about like

what's the right code snippet to call Gemini or whatever.

It's just like baked in.

And I think we're really going to lean into that and double down on that.

And there's so much more that we can do.

So it'll be awesome to see how folks feel about it.

and use it this is one of the cool things for us about like we i know we don't talk about this a ton but um ai studio as a product is built on all the google cloud infrastructure um so deploy as an example is using cloud run behind the scenes the apis are using cloud projects we're using cloud billing we're using the cloud api key system so it's like all of these subtle ways in which we can slowly start to expose

to users, the complexity and AI builders and developers, the complexity of actually building like a full stack app without having to sort of, there's obviously lots of complexity of what you could build.

And how do you do that in like a really incremental way so that you don't,