A path that fits your curiosity.
Start with a guided model, inspect an AI system, or practise a decision. These are suggested routes through live lessons. Everyone can explore any topic.
Start with a goal
New to AI? Try “Understand your first model”. You can explore any lesson before choosing a path.
More goals: applications, projects and deeper practice
Understand your first model
Explain how a network learns, spot overfitting, and evaluate a small classifier.
Starting point: School algebra, percentages and graph reading. No programming needed. Refreshers are available inside lessons.
How progress and assessment work
Evidence you can collect: Five lesson completion records, separate understanding checks, and the existing classifier project with a frozen evaluation and model card.
Limits and future work: The expanded student path will add data quality, classical ML, real datasets and independent debugging. This starter uses the lessons available today.
Path version 1. Choosing this path changes your suggested next activity and its completion total. Earlier achievements, experiments, project reports and other path preferences stay saved. This is not full-program completion.
Your activities
0 / 6 activities completed
0 / 6 understanding checks passed
0 skipped and still outstanding
Skip for now changes the suggested next step, not the requirements or your understanding record. Restore an activity whenever you like. Optional refreshers and coding are outside this total.
Previewing this path. Choose it above to save skip preferences; all lesson links are open now.
- LessonTo explore
Your first neuron
Completion outstanding · Understanding check to try
- To explore
Completion outstanding · Understanding check to try
- To explore
Completion outstanding · Understanding check to try
- LessonTo explore
Send the error backward
Completion outstanding · Understanding check to try
- To explore
Completion outstanding · Understanding check to try
- To explore
Completion outstanding · Understanding check to try
Project evidence
Build and evaluate a classifier: evidence still to collect; understanding check to try.