Channel Guide

Reinforcement Learning 3 series · start here for RL

›

Builds up from a basic Q-learning agent to a world model, step by step.

Model-based RL with Lunar Lander — building the full PETS algorithm from the ground up.

Self-play reinforcement learning plus Pong from scratch. Complete series.

Robotics 1 series · ongoing

›

Building a physical robot arm from scratch — hardware, joints, firmware, and everything that broke.

AI Security 1 series · 3 videos

›

Testing three AI assistants against a deliberately vulnerable web app — ChatGPT, Claude, then Gemini.

AI Tools standalone videos

›

Head-to-head: which AI coding assistant is most useful for reinforcement learning troubleshooting?

First impressions from a Claude user — what Gemini CLI does differently.

Coding offline with local models — how well does Qwen Code hold up?