Raytheon awarded $71 million in Navy contracts for missile systems
Investing.com -- Google (NASDAQ:GOOGL) DeepMind has announced Genie 3, a new general purpose world model capable of generating interactive environments that users can navigate in real time.
The new AI system can create dynamic virtual worlds based on text prompts, maintaining consistency for several minutes at 720p resolution and running at 24 frames per second.
Genie 3 represents a significant advancement in world simulation technology, building upon Google DeepMind’s decade-long research in simulated environments. The company has previously developed AI systems for mastering real-time strategy games and creating environments for open-ended learning and robotics.
According to the announcement, world models enable AI systems to simulate aspects of the world and predict how environments will evolve and respond to actions. Google DeepMind views these models as important steps toward artificial general intelligence (AGI), as they allow AI agents to train in unlimited simulated environments.
The company introduced its first foundation world models with Genie 1 and Genie 2 last year. It has also developed advanced video generation models called Veo 2 and Veo 3, which demonstrate understanding of intuitive physics.
Genie 3 marks the first world model from Google DeepMind that allows real-time interaction while improving consistency and realism compared to its predecessor.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.