One pre-trained robot model now matches or beats the specialists that were fine-tuned with reinforcement learning for the very tasks they were built for.

scaling laws Whether more compute keeps buying more capability, and where the curve now bends -- pre-training, post-training, test-time.

robotics Machines that act in the world: humanoids, training data for robots, and how far behind language models they run.

reinforcement learning Training by reward: environments, verifiable tasks, value functions, and whether it works or merely beats what came before.