Foundations

A 12-week, 24-lesson curriculum covering symbolic AI, neural networks, vision and NLP.

Stanford's natural language processing course, updated each year to cover current model architectures.

The course that trained a generation of computer vision researchers. Notes and assignments are public.

An interactive textbook where every concept comes with runnable code in PyTorch, TensorFlow and JAX.

A national-scale AI literacy course with no maths prerequisite. Finland taught this to its entire population.

Short, practical micro-courses that run in the browser. The intro to ML track takes an afternoon.

MIT's week-long bootcamp, published free every year with lectures, slides and labs.

Classic machine learning in 26 lessons, using Scikit-learn and real-world datasets.

Google's own internal ML primer, with interactive exercises and no setup required.

A free certification track covering TensorFlow and the core ML workflow end to end.

Builds a neural network, then a GPT, from scratch in plain Python. The clearest explanation of backpropagation anywhere.

Top-down deep learning: you train a working model in lesson one and learn the theory as you need it.

OpenAI's own educational resource for reinforcement learning, with reference implementations.

Harvard's AI course: search, knowledge representation, optimisation, learning, and neural networks.

Andrew Ng's five-course sequence. Still the most recommended structured path into deep learning.

The rebuilt version of the course that started the MOOC era. Assumes only basic maths.