Free courses from Ionate
Learn with Ionate
How AI learns, taught by the people who build it.
Ionate Technologies builds reinforcement learning systems and AI agents. Here we teach how they work, in courses of short lessons you can take at your own pace. Free, with no account to make.
Every lesson comes with
- A lesson you can read on its own, in plain words
- The video, once it's out
- Simulations that train live in your browser
- A short check, with every answer explained
- The slides it's taught from
- Your progress, kept in this browser
The series
Courses, one series at a time
Series 01 · Free course
RL from Zero to Agents
13 episodes · 1 in production · 12 coming soon
Reinforcement learning from the first idea to training language models and agents.
Open the seriesSeries 02
Coming soon
A new series is in the works.
Series 03
Coming soon
A new series is in the works.
New here? Start with · RL from Zero to Agents · Ep 00
What Is RL?
First the car crashes. Then it wobbles. Then it drives a clean lap, and not one line of its code says how to steer. How did it learn this when nobody told it how to drive?
About 25 minutes · Nothing to know firstStart the lesson
- Explain how reinforcement learning differs from supervised learning.
- Name the parts of the RL loop: agent, environment, state, action and reward.
- Describe how trial and error, and the trade-off between exploring and exploiting, let an agent improve.
How it works
Every lesson, the same three steps
Read, or watch
Each lesson explains one idea in plain words, with the video beside it once it's out. Every term links to its definition.
Try it
Run the simulation on the page. Guided experiments set it up for you and say what to look for, and why it happens.
Check it
A few questions at the end, with an explanation for every answer. Mark the lesson complete and this browser remembers where you are.
The slides each lesson is taught from are there too: open them beside the page, or print them.
RL from Zero to Agents · Episode 0
Simulations you can train in your browser
Nothing here is a recording: the agents learn on your device, and every number on screen is computed as it runs.
- A maze that learns its way outA Q-learning agent learns a gridworld from rewards alone. Bribe it with a coin and watch it stop going for the goal.Try it in Episode 0
- A car that learns to driveFive distance sensors, three steering choices and a reward for progress. It crashes hundreds of times, then laps cleanly.Try it in Episode 0
Who makes this
Ionate Technologies
Learn with Ionate is made by Ionate Technologies, an AI lab and engineering studio. We build reinforcement learning environments, train models in them and ship AI agents, and we teach what we build. Everything runs on free hardware: your browser, and Kaggle's free GPUs later on. About Ionate