Yang Likun's recent output density is too high, from criticizing LLMs to the robot route and then to the insider story of Meta's resignation, the disagreements among the three Turing Award giants are deeper than I thought.

In an unsupervised learning podcast, Yang LeKun puts forward a contrarian view on the limitations of large language models, explores a robotics technology path, explains why he left Meta, and differs from Hinton and Bengio in their views on the Turing Award. He also predicts 2027, introduces the new company AMI and its bet on world models, compares OpenAI and Anthropic to Sun Microsystems, and advises PhD students to stop researching LLMs, sharing insights into AI safety, the mechanisms behind breakthrough research, and Meta FAIR’s gains and losses.