Research Scalable AI for Science Why AI Scientists Need High Bias and High Variance The Station From AI Agent to AI World: A New Paradigm for Genuine AI Scientists The Thinker Task Decomposing reasoning into Fast and Slow Thinking to train LLM's System 1 & System 2 The Planning Trinity Why General Intelligence Needs Intuition, Reason, and a World Model Thinker A model-based reinforcement learning algorithm that lets an agent learn to plan Coagent Networks Train neural networks without backpropagation, following biological learning rules
