Google DeepMind has unveiled Genie 4, an advanced world model that significantly improves upon its predecessor with more realistic physics, human-like characters, and real-time audio generation. Integrated with Google's Gemini models, Genie 4 can maintain consistent simulations for up to two minutes and allows for mid-simulation edits, such as altering weather or adding elements. The technology aims to create infinite training environments for robotics and accelerate rapid prototyping in creative fields, with a research preview now available to trusted testers and wider rollout planned for Google AI Ultra subscribers and Gemini API partners. AI
IMPACT Enhances simulation capabilities for robotics and creative prototyping, potentially accelerating AGI development.
RANK_REASON This is a new model release from a frontier AI lab (Google DeepMind) with a system card.
Read on Mastodon — mastodon.social →
- Electronic Frontier Foundation
- Google AI Ultra
- Google DeepMind
- Medicare
- AMD
- Diego Rivas
- Fei-Fei Li
- Gemini
- Gemini 3.8 Flash
- Gemini API
- Genie 3
- GWM Worlds 2
- Rivian Automotive
- Runway ML
- Tejas Kulkarni
- Third Dimension
- Veo
- World Labs
AI-generated summary · Google Gemini · from 7 sources. How we write summaries →