At the core of world model research is the belief that artificial intelligence (AI) must go beyond just processing text to truly understand and interact with the world. Louis Castricato, a computer scientist, realized this after years of studying large language models, which power chatbots like ChatGPT. Feeling that research in this area had reached a plateau, he left his doctoral studies to start Overworld, a company focused on developing AI that can comprehend and navigate real-world environments. While AI chatbots remain profitable, with significant investments from companies like Anthropic and OpenAI, a new wave of AI entrepreneurs is exploring “world models.” These models aim to teach AI systems and robots how to respond to physical surroundings. Esteemed scientists, including Fei-Fei Li, emphasize the importance of world models in advancing AI technology.
QUESTION: How might the development of AI that understands and interacts with the physical world change the way we live and work in the future?
Top developers are shifting from chatbots to physical AI. Here’s why
