Artificial Intelligence / Unit 11 AI 0623 worked example, 60 to 90 seconds
Three Systems, Placed
The idea
Placing three everyday systems in the design space — by hand, one dimension at a time — is the course's map used as a map.
In practice
A mapping app first: the road graph is given, observable, and deterministic enough — Unit 2's home cell, which is why its routes come with guarantees. Next, a benefits screener: a learned classifier deciding about people — partial observability of lives compressed into features, a pile with fingerprints, lopsided stakes, and a feedback loop waiting wherever denials shape next year's data. The clean-cell guarantees are gone; every Unit 10 question is live. Last, a chat assistant: Unit 9's machinery, fluent and confabulation-prone, sitting in an easy cell while it only talks — and sliding toward the corner cell the moment it is chained into long tasks in changing worlds shared with other agents, where the multiplication rule takes over. Same exercise, any system: read each dimension, find the cell, and the method's powers and failure modes follow.
Where people slip
The placement is the audit's first step — a system's cell predicts its failure modes before a single test is run.
Your place on the trail
Every lesson rests on earlier ones and opens later ones. Tap any to jump.
Builds on
Unlocks
- Nothing yet depends on this.
Something wrong in this lesson? Flag it · Sources for Artificial Intelligence