Sander Saar, founder of AgentCDN, making videos work inside AI

Transcript

1. 0:00

   Tap on any outfit to try it out on yourself. That's just one of the 5 AI products Google dropped today at their I/O developer conference, and you can actually use right now on your phone. First up, Virtual Try-On. Go to the Google app, hit the Labs icon, then Try Things On. Upload a full-body pic of yourself, and then pick any outfit, men's or women's, works on both. The AI puts it on you. If you like what you see, you can actually buy it right from there. There are apps out there with similar capabilities like DressXMe and Doji, but this can be really powerful when built right inside Google Search.

2. 0:29

   Next, Imagen 4. It's built right into Gemini app, where you can generate images with insane control over text and characters and with much higher fidelity and realism than before. You can also edit existing images. It's on par with the best out there, like ChatGPT, but maybe even better. Project Mariner is like having a virtual assistant that navigates the websites for you. I asked it to find concert tickets for under $50 this weekend, and it searched multiple sites, compared prices, and found many options.

3. 0:54

   It is a browser within a browser, or it can also take over your browser if you like so. Flow is a new video creation app with a V03 video generation model to create videos with synchronized audio. I added a text prompt for a Google product manager rapping about their latest releases on an I/O stage, and here's what I got.

4. 1:10

   At Google I/O 2025, we shared how we're using our cutting-edge technology to build products that are intelligent and personalized.

5. 1:18

   You can also check out Flow TV to see the 3 short films that are made entirely with this technology. Stitch turns descriptions and screenshots into working app designs. I said a minimal, clean iOS app for AI-driven health tracking and AI health coach and got this, complete with working code. You can continue the development or export to Figma in If you watched until this far, here's a bonus one.

6. 1:38

   Google just announced at Google I/O that Gemini Live is now available for free to everyone on Android and iOS in the US.

7. 1:44

   So you can see what you see from the camera and hear what you hear to have a real-time live conversation. For most of these, you need to get an AI Pro subscription for $20 a month or have the $250 maxed-out limits with Ultra. I'm happy to see that this year's I/O is less about mockups and more about shipping products. What are you most excited about?