Try a task first
Find your next lesson by making something.
Six short exercises with sample inputs, answer checks and course options. No signup, new subscription or paid API key required. These are our own exercises, not assignments from the courses we link to.
Python developers choosing an agent course · About 45 minutes
Build a support lookup before an AI agent
Make a five-case support lookup and find out whether your next gap is Python, tool use or answer evaluation.
Try this exercise →JavaScript developers comparing AI courses · About 45 minutes
Make an AI result usable in a web interface
Build a small result card with success, missing-answer and invalid-data states before paying for a model API.
Try this exercise →Office workers starting without code · About 25 minutes
Turn a messy announcement into a checked action list
Use an invented event brief to measure whether AI reduces your checking work, with a source-based answer key.
Try this exercise →Learners choosing a machine-learning foundation · About 35 minutes
Catch data leakage before training a model
Split a tiny repeated-customer dataset and compute a baseline by hand before selecting a machine-learning course.
Try this exercise →Managers choosing practical AI literacy · About 30 minutes
Choose an AI pilot your team can evaluate
Compare three proposed team tasks using measurable outcomes, data access and the cost of a wrong answer.
Try this exercise →Independent builders deciding between a course and a learning community · About 30 minutes
Find the next lesson from a project that stalled
Turn one unfinished project into a concrete next experiment and a question that another builder can answer.
Try this exercise →