This course collects the full AI Starter Course in one place so readers can move through it lesson by lesson.
Use the lesson list below to jump into the course or return here any time to continue where you left off.
Course lessons
- 01
Lesson 1
What Is Artificial Intelligence? A Practical Beginner Definition with Examples
A beginner-friendly introduction to what artificial intelligence is, with simple examples, interactive visuals, and an easy first step into the math used throughout the course.
- 02
Lesson 2
How Machines Reason: Goal Trees and the Logic of Problem Decomposition
Learn how machines break one hard decision into smaller checks, and why AND/OR trees make reasoning paths easier to follow and understand.
- 03
Lesson 3
What Are Expert Systems? Rule-Based AI, MYCIN, and the First Commercial AI Wave
Learn how expert systems turn explicit rules into recommendations, why that made early AI useful, and where rule-based reasoning still helps today.
- 04
Lesson 4
Search as Problem Solving: How AI Navigates a State Space
Every AI that makes decisions is searching through a space of possibilities. This post formalizes what a search problem is, introduces depth-first, breadth-first, and iterative-deepening search, and explains why the choice of algorithm determines whether you find a solution at all.
- 05
Lesson 5
Heuristic Search and A*: Finding the Optimal Path Without Exploring Everything
Uninformed search explores blindly and hits an exponential wall. Heuristic search adds a guide: an estimate of how far you are from the goal. This post builds up from greedy best-first search to A*, proves why it is optimal when the heuristic is admissible, and shows it applied to our hospital triage diagnosis graph.
- 06
Lesson 6
Local Search: When the Path Does Not Matter, Only the Destination
A* finds the optimal path from start to goal. But many AI problems have no meaningful start, no path to trace, and no graph to traverse, only a space of candidate solutions with measurable quality. This post examines hill climbing, random restart, and simulated annealing, and explains why the landscape metaphor reveals everything about where these algorithms succeed and where they fail.
- 07
Lesson 7
Constraint Satisfaction Problems: When Rules Define the Answer
Discover how CSPs turn scheduling and assignment puzzles into tractable search problems, with arc consistency and backtracking search at their core.
- 08
Lesson 8
Automated Planning: Teaching AI to Think in Action Sequences
Explore how STRIPS-style planning lets AI reason about goals, preconditions, and effects, turning abstract objectives into executable action plans.
- 09
Lesson 9
Introduction to Machine Learning: Learning from Data Instead of Rules
Understand the core shift from hand-coded AI to learned models: what a learning algorithm actually does, the three learning paradigms, and the bias-variance tradeoff that governs all of them.
- 10
Lesson 10
Decision Trees: Splitting Data into Clarity
Learn how decision trees recursively partition feature space, why information gain drives splits, and how to use them to classify triage patients one question at a time.
- 11
Lesson 11
Neural Networks: Learning Representations Layer by Layer
Understand how artificial neurons, activation functions, and layered architectures give neural networks their expressive power, then watch a feedforward network classify triage patients.
- 12
Lesson 12
Backpropagation: How Neural Networks Actually Learn
Demystify backpropagation by walking through the chain rule layer by layer, then see how gradient descent uses those gradients to drive weight updates and reduce prediction error.
- 13
Lesson 13
Deep Learning and CNNs: Structure That Reflects the Problem
Explore how convolutional neural networks, residual connections, and transfer learning enable deep models to extract meaningful patterns from images, signals, and structured sequences.
- 14
Lesson 14
Knowledge Representations: How AI Structures What It Knows
Survey the major ways AI systems encode knowledge, semantic networks, frames, ontologies, and description logics, and understand when symbolic knowledge representations outperform learned ones.
- 15
Lesson 15
Modern Architectures: The Transformer and the Era of Foundation Models
Understand how the self-attention mechanism works, why Transformers replaced RNNs across language and biology, and what foundation models and fine-tuning mean for clinical AI.
- 16
Lesson 16
Probabilistic AI: Reasoning with Degrees of Belief
See how Bayesian networks encode conditional independence, how inference propagates evidence, and why probabilistic reasoning gives AI systems calibrated uncertainty instead of brittle yes/no verdicts.
- 17
Lesson 17
Uncertainty and Graphical Models: Reasoning Over Incomplete Information
Go beyond Bayesian networks to explore Markov models, hidden states, and temporal reasoning, then see how uncertainty quantification separates trustworthy clinical AI from confident guessing.
- 18
Lesson 18
Wrapping Up: A Map of AI and Where It Goes Next
Bring the full course together: trace how symbolic AI, search, machine learning, deep learning, and probabilistic reasoning form a coherent toolkit, and see where the frontier lies today.