Google DeepMind dropped the next generation of world model builder with Genie 3. The simulation in Genie 3 takes a leap from mere game-specific few seconds of 320p video generation of GameNGen, and Genie 2 to a whopping output of 720p with navigation in real time that is extended to multiple minutes.
Genie 3 can generate vibrant ecosystems with intricate real-world landscape, animal behavior, and more. The new AI model also generates complex natural phenomena including lighting, water etc. Genie 3 also maintains environmental consistency for several minutes, “with visual memory extending as far back as one minute ago.”
Besides building an interactive world with a single prompt, Genie 3 users can insert further text prompts to manipulate the world in real time “like changing the weather or introducing new characters.”
Stating the purpose of building a world model generator, Google DeepMind in its blog remarked
World models are also a key stepping stone on the path to AGI, since they make it possible to train AI agents in an unlimited curriculum of rich simulation environments.”
Company’s Noble Laureate CEO Demis Hassabis in his tweet stated, “- one day we will be able to build the Holodeck for real!”
Reacting to Genie 3, Elon Musk dropped a “Congrats” on Hassabis’s tweet.
Genie 3 is here – it can generate an entire world simulation that you can interact with in real-time, just from a text prompt! It’s pretty mind-blowing really when you stop to think about it, and it’s rapidly improving – one day we will be able to build the Holodeck for real! pic.twitter.com/wJbXfEFJm2— Demis Hassabis (@demishassabis) August 5, 2025
This line of work (and world models in general) is very close to my heart. Back in the 90s when I was designing sim games we could only dream of one day having tech like this. Incredible work from the Genie team! In limited preview for now, see blog: https://t.co/k5vjtlFKfh— Demis Hassabis (@demishassabis) August 5, 2025
What if you could not only watch a generated video, but explore it too? 🌐
Genie 3 is our groundbreaking world model that creates interactive, playable environments from a single text prompt.
From photorealistic landscapes to fantasy realms, the possibilities are endless. 🧵 pic.twitter.com/P0cwFvf5d2— Google DeepMind (@GoogleDeepMind) August 5, 2025
Beyond navigation, users can insert text prompts to alter the world in real-time – like changing the weather ⛅ or introducing new characters 👤
This unlocks a new level of dynamic interaction. pic.twitter.com/vvy9U2KOav— Google DeepMind (@GoogleDeepMind) August 5, 2025
See Also: Is ‘The Matrix’ Real? Physicist Makes Explosive Claims Of Our World Being A Simulation
See Also: This Disturbing AI Short Generated On Google Veo 3 Is Straight Out Of A Black Mirror Episode: Watch
