Free AI foundations · Lesson 2 of 12

Compare options and understand uncertainty

Define a problem and avoid treating a prediction as certainty.

Understand

A useful problem description names a starting situation, possible actions, constraints and a desired result. A search method explores possible paths from the start toward that result. A shortcut for prioritizing promising paths is called a heuristic. It can speed up a search without guaranteeing the best answer. A generative model describing a plan is not automatically performing a reliable search or checking every constraint.

For decisions involving another player, a game tree represents possible moves and responses. A minimax strategy chooses the move with the best worst-case outcome, assuming the opponent responds optimally. Everyday planning often has uncertain outcomes rather than a perfectly rational opponent, so those assumptions may not fit.

Probability describes uncertainty, not a promise. The starting frequency of an event matters when interpreting a signal. Suppose 10 of 100 fictional messages are urgent. A filter flags 8 of those and also flags 18 non-urgent messages. Only 8 of the 26 flagged messages are actually urgent: about 31%. “Catches 80% of urgent messages” does not mean every flag is 80% likely to be right. Counting cases this way is a practical form of Bayesian updating: combine the initial frequency with evidence.

Try it in your life or work

Keep: A decision table with options, constraints, uncertain assumptions and a reason for your choice.

Use fictional or non-sensitive information. Written practice does not require an AI account. The live demos are optional; other tools have their own terms and costs.

This is a self-check, not a scored certification. Answers and exercise notes are not submitted to AstraCruz.