Simple AlgorithmArtificial Intelligence
Artificial intelligence · from intuition to research

Don’t just use AI.
Look inside it.

A course in three levels of depth (Input Layer, Hidden Layer, Deep Layer). Every idea is a model you can train, poke and break, and your understanding is measured, not just your clicks.

166lessons
3depth levels
26simulations
71 hof material
The knowledge network

All of artificial intelligence, in one picture.

The course is a single tree with three layers. The inner ring is the Input Layer: ten big ideas. Each feeds Hidden Layer subtopics, and each of those Deep Layer research dives, like signals flowing deeper into a network. As you learn, your network lights up. Click a segment to zoom in.

Layer I
Input Layer · 10 topics · 2.8 h
Every big idea in AI, built on intuition and live demos you can poke.
Layer II
Hidden Layer · 39 subtopics · 15 h
The real machinery: algorithms, math and models, derived and implemented.
Layer III
Deep Layer · 117 deep dives · 53 h
Theorems, proofs, bounds and the open problems at the research frontier.

Train every idea

26 live simulations: pathfinding, game trees, Bayes nets, gridworld RL, neural nets, attention, diffusion and more. You don’t read about gradient descent; you watch it.

Predict, then observe

Commit to a prediction before every reveal. The surprise is where intuition gets rewired.

Understanding, measured

Bayesian knowledge tracing plus your stated confidence. Confident mistakes are flagged as misconceptions.

Memories need replay

Like an agent’s replay buffer, what you learn fades unless revisited. Spaced reviews keep your network sharp.

Syllabus

Every lesson in the artificial intelligence course

166 lessons in 10 topics. Read any of them here; a free account adds the live simulations and keeps track of what you understand.

2Search & Problem Solving

Finding solutions by exploring a space of possibilities.

  1. Uninformed Search
  2. Heuristic Search & A*
  3. Games: Minimax, Alpha–Beta & MCTS
  4. Constraint Satisfaction Problems

3Knowledge, Logic & Reasoning

Representing what is known and deriving what follows.

  1. Propositional & First-Order Logic
  2. Knowledge Representation & Graphs
  3. Automated Planning
  4. Neurosymbolic AI & Reasoning Models

5Machine Learning Essentials

Learning patterns from data instead of programming rules.

  1. Supervised Learning
  2. Generalization & Overfitting
  3. Unsupervised Learning
  4. Evaluating Models

6Neural Networks & Deep Learning

Layers of simple units that learn rich representations.

  1. From Neurons to Networks
  2. Backpropagation & Optimization
  3. CNNs, RNNs & Residual Networks
  4. Representation Learning

7Language & Large Language Models

From word counts to transformers that write, code and reason.

  1. From N-grams to Embeddings
  2. Attention & Transformers (coming soon)
    • The Mathematics of Attention (coming soon)
    • Positional Encodings (coming soon)
    • Efficient Attention & Long Context (coming soon)
  3. Large Language Models: Pretraining & Scaling (coming soon)
    • Scaling Laws (coming soon)
    • In-Context Learning & Emergence (coming soon)
    • Mechanistic Interpretability (coming soon)
  4. Fine-Tuning, RLHF & Prompting (coming soon)
    • RLHF & Preference Optimization (coming soon)
    • Prompting, Retrieval & Tool Use (coming soon)
    • Evaluating Language Models (coming soon)

8Perception & Generative AI

Seeing, hearing and creating images, sound and video.

  1. Computer Vision (coming soon)
    • Object Detection & Segmentation (coming soon)
    • Vision Transformers (coming soon)
    • 3D Vision & Scene Understanding (coming soon)
  2. Generative Models (coming soon)
    • Variational Autoencoders & the ELBO (coming soon)
    • Generative Adversarial Networks (coming soon)
    • Diffusion Models (coming soon)
  3. Speech & Audio (coming soon)
    • Speech Recognition (coming soon)
    • Speech Synthesis (coming soon)
    • Audio & Music Generation (coming soon)
  4. Multimodal Models (coming soon)
    • Contrastive Vision–Language Models (coming soon)
    • Vision–Language Models & Grounding (coming soon)
    • World Models & Video Generation (coming soon)

9Reinforcement Learning & Agents

Learning to act from trial, error and reward.

  1. RL Foundations (coming soon)
    • Bellman Equations & Contraction (coming soon)
    • Dynamic Programming in RL (coming soon)
    • Exploration Strategies (coming soon)
  2. Temporal-Difference Learning & Q-Learning (coming soon)
    • Convergence of Q-Learning (coming soon)
    • Deep Q-Networks (coming soon)
    • Off-Policy Learning & Importance Sampling (coming soon)
  3. Policy Gradients & Actor–Critic (coming soon)
    • The Policy Gradient Theorem (coming soon)
    • Trust Regions: TRPO & PPO (coming soon)
    • Actor–Critic & Soft Actor–Critic (coming soon)
  4. Agents in the World (coming soon)
    • Multi-Agent Learning & Game Theory (coming soon)
    • Robot Learning & Sim-to-Real (coming soon)
    • LLM Agents & Tool Use (coming soon)

10AI, Ethics & Society

Fairness, safety, privacy and governance of powerful systems.

  1. Bias & Fairness (coming soon)
    • Fairness Impossibility Theorems (coming soon)
    • Bias Mitigation Techniques (coming soon)
    • Explainability Methods (coming soon)
  2. AI Safety & Alignment (coming soon)
    • Adversarial Examples & Robustness (coming soon)
    • Specification Gaming & Reward Hacking (coming soon)
    • Scalable Oversight & Alignment Research (coming soon)
  3. Privacy & Security (coming soon)
    • Differential Privacy (coming soon)
    • Federated Learning (coming soon)
    • Attacks on Models (coming soon)
  4. Governance, Impact & the Future (coming soon)
    • The AI Regulation Landscape (coming soon)
    • The Economics of AI (coming soon)
    • Forecasting AI Progress (coming soon)

Your weights are random. Let’s train them.

Free account, one login for every course. Progress syncs to your profile, so pick up on any device.

Create your account