Reinforcement learning

3 videos across 3 channels

Reinforcement learning describes how agents learn by updating behavior based on prediction errors from rewards and punishments. The clips illuminate neural underpinnings—dopamine and serotonin shaping motivation and learning—and extend the idea to foraging-like decisions and real-world behaviors from dating to social media, while also exploring brain monitoring and AI tools that test and apply these concepts. They also touch on practical implications for autonomous systems and the economics of open weights and licensing that shape how learning policies are shared and deployed.