Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Google’s AI research organization Deepmind, Tuesday evening, Google’s latest video-generating model, potentially videos can be used for video games, said Demis Hassabis.
Go Google in response to a post in X “Allow,” My Video 3 Videos of My Videos “Wen who can play” and “? “Shababan be responsible“Now it would be anything.”
Wednesday morning, Logan Kilpatrick, Google AI Studio and the lead product for the Gemini API, hit an answer with an answer: “🤐🤐🤐🤐”
Both articles from Google leaders are a little more than joints and Google spokesman said there is nothing about Techcrunch to share the moment. However, the construction of world-able-playing world models is not outside the world of opportunities for technology giant.
World models are different from video generation models. The past simulates the dynamics of the real world environment that predicts how the world will develop in response to agents. Video gene models synthesize the real video sequence.
Google, Multimodal foundation model, Gemini 2.5 Pro plans to turn into a world that simulates aspects of the human brain. In December, Deepmind opened Soul 2“Infinite” a model that arises from different playable worlds. The next month we stated that a new team formed for Google to work In AI models that can imitate the real world.
Others are working on building world models – most importantly, EU Pioneer Fei-fei li. Li, last year went secret with world laboratories, a start to build it Own AI system Like a video game, it creates 3D scenes from an image.
Veo 3 in still public imageYou can also make a sound to go along with clips – from speeches can create anything from recording. VEO 3 creates real actions by simulating real world physics, this is not yet a world model. Instead, games can be used for kinematic story in games such as Cutscenes, Trailer and Narrative Prototype
The model is still a “passive output” blessing model and you need to go to a simulator who is more active, interactive and predictive (or future VEO generations).
But the real problem with the production of video games is not only impressive visual; Real-time, consistent and controlled simulation. Therefore, Google can make sense to see a hybrid approach to Google and Genie because Google needs to follow the future or playing world development.
Google could compete with himself Microsoft, ScenarioThe runway, pika and in the end, Openai video generating model sora.
Given the planned movements of Google’s planned movements and deep pockets and distribution muscles in world models, the competitors would be smart to protect the next hour in this space.
