Search AI Products and News

At the 2025 GTC conference, Nvidia CEO Jensen Huang announced that Nvidia is collaborating with Disney Research and Google DeepMind to develop a physics engine called "Newton." This engine aims to simulate robot movement in real-world environments, providing technological support for future entertainment robots. Image note: Image generated by AI, image licensing service provider Midjourney. Disney will be among the first companies to utilize Newton technology, and this technology is expected to be applied to its...

Recently, a study led by Meta's Chief AI Scientist Yann LeCun revealed how artificial intelligence can develop a basic understanding of physics by watching videos. This research, conducted by scientists from Meta FAIR, Paris University, and EHESS, shows that AI systems can acquire intuitive physical knowledge through self-supervised learning without predefined rules. The research team employed a new method called Video Joint Embedding Predictive Architecture (V-JEPA).

The CAS Institute of Physics has released the MatChat AI large language model for material synthesis, which uses the synthesis processes of inorganic materials extracted from over 4 million papers, processing the data into 13,878 high-confidence synthesis path descriptions. The model fundamentally possesses generative and reasoning capabilities in the field of material synthesis. Experimental validation shows that the model outperforms ChatGPT in predicting the synthesis of complex inorganic materials. This achievement demonstrates the innovation of large models in the materials domain.