Technology & AI

Google's Genie 4 AI world model reaches trusted testers in preview

Google DeepMind's latest system builds explorable virtual scenes with sound, speech, and responsive characters.

By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .

This story is no longer being updated. It reflects reporting up to .

What happened

Google DeepMind has unveiled Genie 4, which generates interactive simulated environments in real time, complete with sound and dialogue, and lets users steer scenes with text prompts. Access so far is limited to a small group of trusted testers, while Google AI Ultra subscribers and early developer partners are expected to get it later.

Why it matters

DeepMind says simulating the world is essential to building AGI, and Google says all of Genie 4's output will carry SynthID markers.

Timeline

  1. Oct 28, 2:22 AM ET: Google DeepMind unveils Genie 4, a model that simulates interactive worlds in real time with generated sound.
  2. Oct 28, 12:14 PM ET: Access starts with trusted testers, while Google AI Ultra subscribers and early API partners are set to follow.
  3. Oct 28, 12:59 PM ET: DeepMind describes the technology as still at an early research stage, despite striking demonstrations.
  4. Oct 29, 12:20 AM ET: Google's Sundar Pichai publicly praises Genie 4 for its lifelike, interactive simulations.

Coverage

More on this story

More from Technology & AI

All Technology & AI news · Technology & AI story archive