I remade the Silicon Valley title sequence for Bengaluru, where everything is code, including the music. No AI video generation model involved.
If you've seen the HBO show, you know the opening: a miniature tech city the camera glides across, buildings rising as it goes. No city deserves that treatment more than Bengaluru. So I made 'Namma Silicon Valley'. Flipkart, PhonePe, Swiggy, Zerodha, CRED, and 30+ others rising out of a toy city, with Vidhana Soudha, Cubbon Park, Lalbagh, Chinnaswamy Stadium, metro, ISRO, and, of course a pickleball court.
The video was only half the point. I made this to test how far AI could go. Rather how far I could push it, and where model capabilities stand today. The realization I've come to is that if you can describe what you want clearly enough, the execution will take care of itself. You're pretty much in the director's seat now, and the job that comes with the seat is honing your vision. You have to be able to see the output you want before you can ask for it.