Module 5: Evaluation, uncertainty, and error analysis

AINS6001 — Foundations of Artificial Intelligence

Essential Question

How do we know whether an AI system is useful, robust, and honest?

Core Moves

• Define the problem boundary

• Identify evidence and assumptions

• Build or evaluate the artifact

• Communicate limits and next actions

Lab & Assignment

Build an evaluation table with accuracy, error slices, and uncertainty notes.

Use the assignment notebook to turn lab evidence into a defensible recommendation.