Testing Like a Builder
Unit 4 teaches students to think like an AI builder. Across five 30–45 minute lessons, they find failure modes, test and iterate, judge results with simple metrics, redesign a workflow to reduce risk, and defend their tradeoffs.
The 5 lessons
- 1
When AI Gets It Wrong
30–45 minStudents identify common failure modes — the unusual or edge-case inputs where AI breaks down.
Vocabulary: failure · limit · error · edge case
- 2
Testing Like a Builder
30–45 minLearners discover why builders test repeatedly to find weaknesses before a system is widely used.
Vocabulary: test · iterate · refine
- 3
Metrics and Criteria
30–45 minKids use simple criteria and metrics to judge whether AI results are accurate, useful, and safe.
Vocabulary: metric · criteria · quality · reliable
- 4
Redesign the Workflow
30–45 minStudents redesign a weak AI workflow — adding prompts, checks, and revision steps — to reduce risk.
Vocabulary: workflow · revise · safeguard
- 5
Builder Review
30–45 minLearners present a redesigned workflow and explain the tradeoffs behind their design choices.
Vocabulary: tradeoff · evidence · justify
Get the full 6–8 Teacher Pack
Every lesson includes ready-to-teach plans — warm-ups, mini-lessons, activities, assessments, and printables. No prep required.