Google DeepMind has released Genie, an image-to-video AI model that can generate interactive games from images or sketches. It has the potential to empower game developers and offer cost-effective prototyping, but faces challenges from competition and technical limitations.