Google Project Genie lets you create interactive worlds from a photo or prompt
Summary
Ars Technica covers Google's Project Genie, a research prototype that lets users generate interactive AI-generated worlds from a photo or prompt, delivered through a dedicated web app to Google's AI Ultra subscribers. It builds on Genie 3's world-model capabilities, but remains limited by 60-second runs, input lag, and content restrictions, with access priced at $250/month and not yet integrated into Gemini.