Louis Castricato, a computer scientist, left his research on large language models at Brown University to start Overworld, a company focused on developing AI that can understand and interact with the physical world, not just process language. This shift reflects a broader trend among AI entrepreneurs who are exploring “world models,” which teach AI systems to navigate and react in real-world environments. Prominent figures like Fei-Fei Li and Yann LeCun are advocating for this approach, emphasizing that true AI intelligence requires understanding space and time, not just text. While AI chatbots have transformed office work, they lack the ability to interact physically, highlighting the limitations of current generative AI models. This new direction in AI research aims to overcome these limitations by enabling AI to predict and respond to physical interactions, potentially revolutionizing fields like robotics and interactive gaming.
QUESTION: How might the development of AI with world models change the way we interact with technology in our daily lives?
Top developers are pivoting from chatbots to physical AI
