Week 1: What Can AI Systems Do?

Chapter question

What kinds of jobs can an AI system perform, and what must we inspect before we trust its result?

What you will understand

Artificial intelligence is a field, not one app and not one mechanism. Some AI systems classify an image. Some predict a number. Some search for a route. Some follow represented facts and rules. They still share a useful system pattern: input, representation, mechanism, output, test, and boundary.

You will learn to trace that pattern without pretending every system learns, thinks, or understands in the same way.

What you will do

You will complete Microsoft's source-faithful introduction to AI and knowledge representation chapters, then use the local AI Field Map Lab to compare rules, learned parameters, and search as distinct system mechanisms.

What you will produce

Your final artifact is an AI Field Map Evidence Record. It compares four systems by task, input, representation, mechanism, output, test, changed condition, and known boundary.

How the week connects

1. Lesson 1 introduces the field of AI and its major approaches.
2. Lesson 2 studies knowledge representation, expert systems, facts, rules, and inference.
3. The guided investigation distinguishes rules, learned parameters, and search.
4. The Lab runs the four-system comparison and produces the evidence record.
5. The assessment and Wrap-Up check and repair one weak conclusion.

Key vocabulary

AI system, task, input, representation, mechanism, output, classification, regression, search, rule, learned model, test, and boundary.

Approximate time

Two 42-minute lessons, one guided investigation, one Lab day, and one assessment/Wrap-Up day.

Final evidence required

Submit the exported evidence JSON or its complete accessible text equivalent. It must include all four system mechanisms, one exact baseline/changed comparison, a shared-system pattern, and a limitation that says what the local simulation does not prove.