42 new free courses added this week from Harvard, Google and freeCodeCamp
freeCodeCampGoogleedXSee what's newA practical route into applied AI. Start with classical machine learning so you understand validation and overfitting, add neural networks, take one deep theory course so the field stops being magic, then learn how generative models are actually used and where they fail.
Every step below is free to enrol in. Work them in order — each one assumes the previous.
Train, validate and improve your first models in a single afternoon.
Why this step: Validation and overfitting first — everything else builds on it.
Build and train neural networks with Keras, including regularisation that works.
Why this step: Neural networks with Keras.
Patrick Winston's celebrated MIT lectures on how intelligence can be represented.
Why this step: The theory course that makes the rest click.
How large language models work, what they get wrong, and how to prompt them well.
Why this step: Prompting, evaluation and the failure modes.
HTML and CSS to a deployed React portfolio.
Python, databases and production web applications.
Statistics, SQL and Python analysis that stands up to scrutiny.