Voices
LeCun Takes the World-Models Case on the Road: RAISE Summit Paris Pitch Against LLM-First AI
At the RAISE Summit in Paris (July 9), Yann LeCun reprised his core thesis that LLMs don't understand the physical world and that progress requires 'world models' — internal representations that let an agent predict the consequences of its actions to enable reasoning and planning. It's continuity, not a new claim, but it's his freshest public appearance and reinforces the JEPA/AMI direction as the leading non-LLM bet now that AMI is funded (~$1B seed). Dated July 9, so slightly outside the 48-hour window — included as tracked-leader output not previously covered.
Source
↳ Follow the thread