Tech News
← Home  ·  All topics

World Model

11 GoKawiil briefs on this topic · RSS for world model

42AI raises $62.5M seed to launch 'behavioral world model' for trading firms

Israeli startup 42AI has exited stealth with $62.5 million in seed funding for a predictive system it calls a behavioral world model. The company says the model infers individual psychological tendencies from data like voice sentiment, support calls and transaction history, and is already deployed at a handful of financial brokerages to forecast how traders might react to volatile markets.

Developer runs tiny AI world model and game-playing agent on Raspberry Pi Pico 2 W

A hobbyist built two small neural networks—a looped transformer 'world model' that generates game frames and rewards, and a lightweight MLP 'actor' that chooses actions—running together entirely on a Raspberry Pi Pico 2 W microcontroller. The world model, ranging from 26,000 to 131,000 parameters, predicts the next frame and reward from the current frame and action, while the actor (9,000-16,000 parameters) picks moves based only on generated frames, never touching the real game logic. The project also supports a streaming reinforcement learning agent that can learn to play while running inside the simulated 'dream' itself.

Google unveils Genie 4 world model with native audio and two-minute scene consistency

Google announced Genie 4, a real-time world model that builds on Gemini Omni and can maintain a consistent simulated environment for up to two minutes, double its Genie 3 predecessor. Users can alter scenes in real time with natural language prompts, including weather, camera movement, and new objects or characters, while the model generates ambient sound and adaptive human dialogue. Genie 4 works alongside Gemini 3.8 Flash, which breaks down prompts into steps, and is currently limited to select trusted testers in a research preview.

Google unveils AI world model that generates synchronized video and human speech

Google introduced a new AI world model capable of simulating environments by generating not only video but also accompanying human voices. The company described the system as part of its broader effort toward understanding and simulating the world, which it frames as a key step toward artificial general intelligence.

Google DeepMind Debuts Genie 4 World Model With Audio and Longer Scenes

Google DeepMind released Genie 4 in research preview, a world model that generates photorealistic, interactive environments with real-time audio including music, ambient sound, and dialogue. The new model allows mid-simulation prompt edits like changing weather or camera angle, maintains consistent scenes for up to two minutes, and offers improved physics and more lifelike characters compared to last year's Genie 3. It's currently available to select testers, with broader access planned for Google AI Ultra subscribers.

Google DeepMind unveils Genie 4 world model with audio and improved physics

Google DeepMind has released Genie 4, an upgraded world model that generates interactive virtual environments frame by frame from text prompts, now with audio support and more realistic human characters. The new version uses Gemini 3.8 Flash as a guiding agent to track goals and keep simulations consistent, though DeepMind says scenes still only hold together for a few minutes before losing context. Genie 4 also retains the ability to alter weather, add elements, or move the camera mid-simulation, as in Genie 3.

Google unveils Genie 4 world model with real-time editing and audio

Google announced Genie 4, a new AI world model that generates interactive experiences lasting up to two minutes, doubling the length possible with its predecessor Genie 3. The update adds real-time editing via natural language prompts, audio generation, and runs on Gemini 3.8 Flash, though Google warns object permanence can degrade over longer simulations.

Mirror Particle builds foundation model to simulate changing human behavior

Mirror Particle, a two-year-old San Francisco startup, is developing a custom foundation 'world model' that aims to predict consumer behavior and its underlying causes, rather than relying on large language models fine-tuned to role-play demographics. CEO Abhivyakti Ahuja says the model focuses on tracking how people change over time, including the triggers behind those shifts. The company has raised an angel round and says it is close to closing its first venture round.

Developer trains 12.5M-parameter world model on RTX 3080 Ti to play Pokémon Red

A developer known as stmonty built a small AI 'world model' that learns to play Pokémon Red by observing screen frames rather than being told game rules or win conditions. The model, with about 12.5 million parameters, was trained entirely on a single consumer RTX 3080 Ti GPU using over 42,000 grayscale frames from more than 1,000 short in-game runs. It draws on JEPA-style world model research co-authored by AI scientist Yann LeCun, who left Meta last year to pursue similar work at his new AMI Labs.

AMD to buy Fei-Fei Li's World Labs for $8.2 billion in all-stock deal

AMD announced an agreement to acquire World Labs, the San Francisco AI lab building 3D 'world models' founded by researcher Fei-Fei Li, in an all-stock deal valued at roughly $8.2 billion. Li will join AMD as chief scientist and executive vice president, while World Labs will operate separately from AMD's chip business until the deal closes later this year.

Runway launches Solaris, an AI model that renders interactive interfaces frame by frame

Runway introduced Solaris, which it calls an interface world model, letting users build interactive apps, websites and digital scenes by directly clicking and dragging on-screen elements instead of writing code. The system pairs a language model that determines how an interface should evolve with a rendering engine that generates each frame live, treating clicks and drags as signals to produce new visuals in real time. Runway demonstrated use cases such as virtually trying on clothes by dragging garments onto a photo of yourself, plus text-prompt edits like recoloring or repositioning objects.