What AI is
Mastery transfer: school email spam shield
Prove you can move the same lens to a new scene: a school spam and phishing detector for student email — then finish the poster case.
1Learn the idea
Read
The big idea (carry it intact)
See it
Fits
- Suggest reply
- Flag odd purchase
- Draft from bullets
Not by itself
- Spreadsheet formula
- Doorbell circuit
- “Smart” ad copy
Fits = smart tasks · Not AI by itself = fixed rules
Keep the durable model:
Designed system doing a bounded task from inputs to outputs — with stakes, checks, and honest labels.
For the email shield:
- Inputs: sender address, links, wording patterns, student reports, maybe past phishing examples
- Bounded task: sort messages into allow, warn, or quarantine
- Outputs: inbox placement, warning banners, admin alerts
- Stakes: missed real opportunities, clicked scams, panic, trust in school email
- Checks: false quarantine rate, false allow rate, easy release button, who reviews edge cases
Calling it AI may be fair if it scores patterns from examples. Calling it AI does not mean it understands the principal’s intent. The label still does not reveal perfect reliability.
Read
Transfer map (same lens, new scene)
| Skill from earlier pages | Email-shield version | | --- | --- | | Feed inputs/outputs | Message features → quarantine decision | | Thermostat boundaries | Sorting mail is bounded; judging friendship is not | | Chatbot fluency trap | Official-looking text can still be phishing | | Cafeteria stakes | Missed coach email vs clicked fake reset | | Yearbook failures | Lookalike domains and spoofed names | | Grades-sheet habits | Spot-check known safe/unsafe messages | | Robot cleaner decision | Honest poster wording with human override |
If you can fill that table without copying adult workplace stories, you have transferable mastery.
Read
Finish the course case
Your school wants a poster: which tools may we honestly call AI?
Draft four poster rows using tools from this topic:
- For You feed — yes/maybe/no + one-line why
- Classroom thermostat schedule — yes/maybe/no + one-line why
- Homework chatbot — yes/maybe/no + one-line why
- Email spam shield — yes/maybe/no + one-line why
Add a footer rule in kid language: “AI means a designed system doing a bounded task from inputs to outputs. The word does not mean always right, always fair, or alive.”
Go deeper
Before you start
Why this matters
Your school turns on a “smart inbox shield.” It parks suspicious messages in a quarantine folder: fake password resets, “free concert tickets,” and messages that look like they came from the principal.
One real message from a coach is quarantined because it has a shortened link. One fake message slips through because it uses the school colors and a familiar name.
You already practiced feeds, thermostats, chatbots, scanners, yearbook tags, grade predictors, and robot cleaners. Now transfer the lens cold.
Related lessons
Check your understanding
Page assessment
Answer from memory. Completion is saved from this evidence, not from opening the next page.
All responses are required.