AnyoneLearnAI
Start learning
Page 2 of 8~96 min topic

What AI is

Build the model: classroom thermostat & timer

A useful AI definition starts with a designed system, a bounded job, clear inputs, and clear outputs — not with sci-fi robots.

~12 min this pageBuild the mental model — mechanism and task boundariesReviewed 2026-08-08

1Learn the idea

Read

The big idea

See it

AI = judgment-like software

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

Use this mental model every time:

Designed system → bounded task → inputs → process → outputs → consequences

Apply it to the thermostat schedule:

  • Designed system: building controls set by facilities or a teacher panel
  • Bounded task: keep the room near a target temperature at set times
  • Inputs: clock, temperature sensor, maybe an occupied/empty signal
  • Process: compare reading to target; turn heat/cool on or off
  • Outputs: warmer or cooler air; a display number
  • Consequences: comfort, energy use, a class that cannot focus if it fails

The classroom timer is even simpler: input is a start time and duration; output is beeps and a zero. Calling that “AI” usually stretches the word past honesty.

Read

Designed does not mean mysterious

“Designed” means people chose the goal. Someone picked 21°C, not magic. Someone decided the quiz lasts 12 minutes.

When a tool later uses machine learning, people still chose the goal. They chose training data. They chose what “good” means. The thermostat helps you practice seeing that human design layer before you meet flashier tools.

Ask:

  1. Who set the target?
  2. What counts as “good enough”?
  3. What happens when the sensor is wrong?
  4. Who can change or pause the system?

Those four questions work on a feed, a chatbot, and a spam filter too.

Read

Contrast: schedule vs photo tagger

Compare the thermostat with a yearbook tool that auto-tags faces in photos.

The thermostat’s job is narrow and physical. A wrong reading means a sweaty room. The photo tagger’s job uses patterns from many faces. A wrong tag can embarrass a classmate.

Same mental model. Different inputs, different failure costs. Your poster should not stamp every automatic thing with the same AI sticker.

Read

Decision steps for the poster

When a classmate wants to label something AI, run this sequence:

  1. Name the job in a verb phrase.
  2. List inputs and outputs you can observe.
  3. Say what success looks like without using the word smart.
  4. Note the biggest uncertainty (sensor, data, ranking, wording).
  5. Only then decide whether “AI” is an honest label, a maybe, or a stretch.

Try it on the thermostat, then on the timer. One may land in “stretch.” That is progress, not a gotcha.

Go deeper

Before you start

Why this matters

Picture your classroom on a Monday. The wall thermostat follows a schedule: warmer before first period, cooler during lunch. Nearby, a digital timer counts down for a quiz.

Someone jokes, “The thermostat is AI now.” Someone else says the timer is AI because it “knows” when time is up.

Pause. What job was each device designed to do? What would count as success? What would count as failure?

Your school poster still asks: which tools may we honestly call AI?

Check your understanding

Page assessment

Answer from memory. Completion is saved from this evidence, not from opening the next page.

1. What is the thermostat’s bounded task?
2. Name one input and one output for the classroom timer.
3. Why does “designed system” matter more than “it feels smart”?
4. Which of the four questions (who set the target, good enough, wrong sensor, who can pause) would you put on the poster first?
5. Transfer this model to a cafeteria barcode scanner. What changes in the consequences box?

All responses are required.

AI literacy · Human approval · Next