Learn AI › Interactive Learning

Practice

Interactive Learning

Exercises, scenarios and decision-based activities for practicing judgment about AI rather than reading about it.

Reading that a citation needs checking is not the same as checking one. Much of the difficulty in working with these tools is a judgment call made under time pressure: whether this answer is good enough to rely on, whether this document should go anywhere near a model, which of four surfaces the task in front of you actually belongs on. Those decisions are easier to learn by making them.

This collection is for activities you work through rather than read — guided exercises, scenarios that branch on what you choose, short challenges and games. They are meant to be quick, and they are meant to be the kind of thing you can get wrong harmlessly.

1. 8 scenarios · one decision each · 5–8 min · Beginner

   Agentic or Not

   Decide whether each AI setup is agentic, partly agentic or not agentic, then compare your reasoning with ours and see which part of the loop decided it.

   Start activity →

2. 5 questions + explorer · 10–15 min · Beginner

   Inside the Package

   Open a real AI Skill package, work out what its files are for, and see how a Skill’s description, its instructions and its supporting material fit together.

   Start activity →

Both work in the browser with nothing to install, keep no account and save no progress, so reloading starts them over. Several of the self-paced courses end in a capstone built the same way: a set of situations, one decision each, with the reasoning revealed afterward.