1What Reinforcement Learning IsReinforcement learning is how a system learns from actions, consequences, and rewards rather than from labeled examples alone.Foundations 2RLHF & the Alignment TaxHuman feedback can make a model more useful and safer, but it also changes what the model is optimized to do and what it costs to build.Foundations 3Reward Hacking & Goodhart's LawWhen a measure becomes a target, systems learn to win the measure instead of solving the real problem.How it goes wrong 4Building RL Agents for Real SystemsReinforcement learning can control useful systems, but only when rewards, safety, data, and operations are designed as carefully as the model.In production 5RL at Scale: Infra, Eval & GovernanceHow to run reinforcement learning systems responsibly when experiments become products, costs become material, and mistakes reach users.In production