Turing Award winner and former Meta Chief AI Scientist Yann LeCun has confirmed the launch of Advanced Machine Intelligence Labs (AMI), a startup focused on developing world model AI systems. LeCun will serve as executive chairman while Alex LeBrun, co-founder and CEO of medical AI startup Nabla, transitions to lead AMI as CEO.

According to the Financial Times, AMI is seeking €500 million (approximately $586 million) at a €3 billion valuation before even launching—a significant but not unprecedented ask in today's AI landscape. For context, former OpenAI CTO Mira Murati's startup raised at a $12 billion seed valuation.

World models represent an alternative to large language models, attempting to understand environments and simulate cause-and-effect scenarios rather than simply generating text. LeCun argues that LLMs lack real-world understanding, leading to hallucination and context problems. By training on video, sound, and sensor data, world models aim to build systems that can predict outcomes and maintain context across complex, multi-step tasks.

AMI's first application will be in healthcare through a partnership with Nabla, focusing on AI agents that can handle scheduling, documentation, and billing while maintaining context throughout workflows. The company joins other major players including Google DeepMind and Fei-Fei Li's World Labs in the race to develop world model technology.