world model

On October 28th, Google DeepMind yesterday released a new generation of world models, Genie 4, supporting real-time production of images, proto-audio and physical simulations, with single interactive simulations lasting about two minutes, which is about one minute longer than the previous generation. The user enters a text that generates an explorable virtual world and controls the movement of the role and adjusts the lens through the keyboard. In the course of the operation, the user can also continue to enter a hint to change the weather, add a person or object. Genie 4 supports original audio generation for real-time environmental sound, music and humans..

On December 21, according to the Wall Street Journal, Meta is mobilizing to develop a new generation of artificial intelligence models in a super-smart laboratory led by the co-founder of Scale AI, Alexander Wang. Currently, the company is working on two new projects: an image and video model code-named Mango and an entirely new text model, internally called Avocado. 1AI notes that the media quoted Meta's internal question and answer session on Thursday..

According to December 12, the technology media Ars Technica today (12 December) published an article on Runway's release of the GWM-1 “World Model” series, claiming that it could maintain a few minutes-long visual consistency, and that the company's strategic focus was expanding from film production to robotics, physics and life sciences. Note: World Models is an AI system that understands and simulates the physical world, not only generates images, but also predicts that "if I throw an apple, it will fall on the ground"..

August 6 news, Google DeepMind official yesterday (August 5) published a blog post, announced the launch of a "world model" called Genie 3, the model allows AI systems to interact with realistic real-world simulations, an important step towards general artificial intelligence (AGI). According to Google, Genie 3 can be used to train robots and self-driving vehicles to interact with realistic recreations of environments such as warehouses. Experts at Google's DeepMind Artificial Intelligence division believe that world modeling is a key step towards AGI...

Recently, Feifei Li, a professor at Stanford University, talked with a16z partner Erik Torenberg, and Martin Casado, an early investor in World Labs, about 'world modeling' and other hot topics. When asked why she started a basic modeling company instead of continuing with academics or participating in existing big models, Feifei Li said that her goal has always been to solve the real 'Polaris Problem', not to start a business per se. According to Feifei Li, language is a way of capturing the world with information loss, and while powerful, humans live in...

January 8 news, technology media The Verge yesterday (January 7) released a blog post, reported that Google DeepMind is planning to set up a professional team, to create a "world model" that can simulate the real physical environment. The project is led by Tim Brooks, former co-head of the OpenAI Sora project, who joined DeepMind in October this year, and is responsible for Google's video generation and world simulator work, a move that is seen as Google competing with rivals in the field of general artificial intelligence (AGI)...