Short answer: Jordan Metzner says Google's Gemini image upgrade can merge two images, such as a person and a dog on a basketball court, or add a couch for interior design, and expects developers to build it into ecommerce apps.
Watch this answer from Episode 10 (starts at 8:41) · Episode 10 notes and transcript
From Episode 10: Future in Focus: AI Tools & Big Tech, starting at 8:41.
Sam Nadler [8:41]: So I know it was a bit of a flop for us, and I know it launched yesterday. Just now in this demo is a little bit of a flop, but what is the hype specifically? I mean, these tool there's been versions of these tools. Like, why is Nano Banana theoretically better? Like, what is the hype about? Is it all about the continuity?
Jordan Metzner [8:59]: Yeah. I mean, I think they call it, like, an upgrade, and we can jump into it. Like, Google Gemini's upgrade gets the bananas upgrade. And so what they say they're able to do is kind of, like, merge two images. So here's an example of, like, a person plus a dog, and they want them on a basketball court, and now they're together on a basketball court. I do have Gemini, and I could definitely try this in Gemini. We could like kind of see how well that works. And so maybe that model is being different than the one I'm I'm running through on my application layer. But here it shows you kind of like, using it for, photo design. Right? As I mean, for interior design as you're able to like add a couch or, add a wall to a room or if it's something like that. So I think, we're gonna just begin to see the power of this because, this is at the level in which developers will be able to apply this to applications, and I think you'll be able to see this, natively, probably on your iPhones and, Android phones, but also probably within a bunch of different types of applications as well that will help you ideate things for purchase and ecommerce and a bunch of other stuff, and then, hopefully help you buy them. I think, as we go through it, let's just jump through the news, and then we can kind of jump through the tool of the week after that. So, obviously, Google's dropping new models. What we haven't seen is a new model that's been better than the Veo 3 model from Google.
Jordan Metzner [10:15]: And I think, it's probably any day now that we're gonna see a new release. The Sora model from OpenAI is certainly not as nearly as good as the Veo 3 Fast. Obviously, there's been some open source models available, and there's definitely like Runway and some other players that have made some models, but I think, we'll see a Veo 3 Fast update to, like, Veo 4 or something like that from Google or some type of competition. And I think that'll really up the game in video, which is even more exciting than images, in my opinion, simply because of how important it is for social ad creation. Yeah.
