Google DeepMind released the Genie 4 world model in research preview, generating photorealistic and interactive worlds with real-time audio, steering, and mid-simulation prompt adjustments for up to two minutes.
The new system, which was released as a trial version for researchers, produces realistic graphics and sounds that match the setting, and allows users to type instructions during the simulation to change the environment in real time.
How the new system works
The technology can hold a scene steady for up to two minutes. Gemini, the company's other AI system, also participates in the process, acting as a director, organizing the next steps and controlling the frames, while Genie takes care of creating the graphics.
When will it be available?
Genie 4 is currently available to select testers. In the coming weeks, it is expected to be released to subscribers of the company's AI services, partners who are early adopters of its programming tools, and to the corresponding application development platform.
G
HACKS.GR / GOOGLE SEARCH
Hacks.gr on Google
Add us to your preferred sources for more updates from Hacks.gr.
Preferred source
↗
end of article
Top comments (0)