Skip to content
Idaho AI Classroom
All lessons
AI FluencyAppropriate Use9-12.AIF.AU.1

The Right Tool for the Job

9–1050 minutes90 minutes (extend hands-on and discussion)

Standard quoted exactly

Understand when AI versus non-AI solutions are appropriate.

Clarification. Appropriateness is context-dependent. The goal is for students to develop the critical thinking skills to evaluate this on a case-by-case basis.

Student-friendly learning targets

  • I can decide whether a task is a good fit for AI, a traditional tool, a person, or a mix.
  • I can explain why appropriateness depends on the goal, the data, and who is accountable.
  • I can recommend a tool for an Idaho workplace or school task and name the risk if the tool is used alone.

Essential questions

  1. When is AI the right tool, and when is it the wrong one?
  2. What does a task's purpose tell us that the tool itself cannot?
  3. Who is still responsible if an AI suggestion is wrong?

Objectives

  1. Apply a three-question test (purpose, trusted data, accountability) to school and workplace tasks.
  2. Sort tasks into AI-appropriate, non-AI, or mixed, and justify each sort in writing.
  3. Contrast at least one Idaho workplace example with one school learning task.
  4. Name a human check that should stay in the loop when AI is used.
  5. Use only district-approved tools, or a paper packet, so no student account is required.

Key vocabulary

Appropriate use
Choosing a tool because it fits the goal, the data, and the people involved—not because the tool is new or popular.
Context
The situation around a task: who it is for, what is at stake, how much time you have, and what counts as a good answer.
Non-AI solution
A way to finish a task without artificial intelligence, such as a calculator, a spreadsheet formula, a sensor, a checklist, or asking an expert.
Human-in-the-loop
A person who checks, approves, or corrects a machine result before it is used.
Automation
Using a machine or program to do a repeated task with little help from a person, usually to save time or reduce errors.
Task fit
How well a tool matches what the task is actually for. A tool can be powerful and still be a poor fit.
Verification
Checking an answer against a trusted source, a measurement, or another person before you act on it.

Teacher background

This lesson trains judgment, not tool use. Ninth and tenth graders already meet AI in homework, social media, and after-school jobs; the work of this period is to slow that meeting down. Appropriateness is context-dependent, which is the exact clarification in the Idaho standard. Treat AI as one option among many: a calculator, a soil probe, a phone call to an agronomist, a spreadsheet, or a human expert. Two pitfalls show up fast. First, students turn “AI versus not-AI” into a morality contest. Keep the frame practical. A frost call for a potato cellar in Bingham County is a different decision than drafting a personal narrative for English. Second, teachers over-demo a chatbot and the class never practices the decision. Script a short projected demo, then move to paper task cards so rural rooms stay in the lesson when Starlink drops. Name Idaho workplaces out loud: Idaho Power load forecasts on the Snake River system, Micron process control in Boise, county 911 dispatch, and infrared wildfire maps from the National Interagency Fire Center. Students should leave able to say, for a given task, what a human, a traditional tool, and an AI system each do well—and what each cannot do.

Materials and prep

Materials

  • Whiteboard or chart paper and markers
  • Printed Right Tool task cards (8 school and Idaho workplace tasks) — one set per pair
  • Right Tool Memo half-sheet (one per student)
  • Sticky notes
  • Projector and teacher workstation for an optional district-approved chatbot demo
  • OFFLINE fallback: printed “fake model output” packet with two scripted prompts and canned replies (comfort a classmate; list public Idaho snowpack sources)
  • Timer visible to the class

Before class

  • Print and cut task cards; stack them in envelopes labeled Station A.
  • Print the fake-output packet even if you expect to demo live—Starlink and guest Wi-Fi fail.
  • Write the three-question test on the board before the bell: Purpose? Trusted data? Who is accountable?
  • If using a district-approved tool, test the two scripted prompts on the teacher account only. Do not log students in.
  • Identify one local example (ag, wildfire, hospital, county, trades) students will recognize.

Instructional sequence

Person or machine?

7 min
  1. 1. Draw two columns on the board: “Needs a person” and “A machine could try this.”
  2. 2. Read four tasks aloud: calculating a tip, comforting a friend after a loss, spotting frost risk from a weather table, and writing a college essay about your family.
  3. 3. Students hold up one finger (person) or two (machine) for each task. Tally the split; do not debate yet.
  4. 4. Circle the frost item and the essay item. Tell students both will return, and that “it depends” is an allowed answer once they have a test for it.

The three-question test

13 min
  1. 1. Write the standard on the board exactly: “Understand when AI versus non-AI solutions are appropriate.”
  2. 2. Teach the three-question test: What is this task for? What data can we trust? Who is accountable if the answer is wrong?
  3. 3. Work one Idaho example live: Idaho Power forecasting evening load on a January cold snap. Historical spreadsheets may be enough; an AI forecast may help; a human dispatcher still calls the play.
  4. 4. Contrast with a school task: a lab conclusion meant to show the student understood the experiment. An AI draft is a poor fit because the goal is the student’s thinking.
  5. 5. Read the clarification aloud: appropriateness is case-by-case, not a forever yes or no. Post it where students can see it during independent work.

Four-card sort

10 min
  1. 1. Pairs receive four printed cards: a potato grading camera on a Magic Valley packing line, a 911 call, a personal narrative, and a county budget spreadsheet.
  2. 2. For each card, pairs mark AI, not-AI, or mixed, and write one reason that uses the three-question test.
  3. 3. Two pairs compare. Any mixed vote must name who stays in the loop (grower, dispatcher, teacher, clerk).
  4. 4. Cold-call two pairs to defend one card. Push for purpose, not “because AI is smart.”

Right Tool Memo

10 min
  1. 1. Each student draws or is assigned one new task: St. Luke’s shift scheduling, a Nez Perce fisheries count, a welding bead inspection, or a history DBQ.
  2. 2. Complete the memo: task goal, best non-AI option, possible AI role, risk if AI is used alone, and a recommendation.
  3. 3. Remind students: no names, photos, grades, or private writing go into any digital tool. Paper memos stay in the room.
  4. 4. Collect memos as the exit ticket, or photograph the board tallies if you need a faster close.

Real-world examples

  • A Bingham County potato cellar uses a soil probe and a frost forecast. AI might help scan a weather table, but a grower still walks the cellar because a wrong call ruins a crop.
  • The National Interagency Fire Center in Boise uses infrared maps and models for wildfire. Crews still look at real wind, fuel, and road access before they move people.
  • Micron in the Treasure Valley uses automated inspection on wafers. A technician still decides whether a flagged lot is scrapped, reworked, or released.

Hands-on activity

Use / don’t use / use with a check

12 min
  1. 1. Station A (always on): students sort eight mixed school and Idaho work tasks onto three desk mats labeled Use, Don’t use, and Use with a human check.
  2. 2. Station B (teacher-projected): on a district-approved tool, the teacher pastes two scripted prompts—one poor fit (comfort a grieving classmate) and one better fit (list three public sources for Idaho snowpack). Students watch; they do not type.
  3. 3. If the network is down, Station B is the printed fake-output packet with the same two prompts and canned replies. The lesson does not pause.
  4. 4. Each student adds one sticky note to a class chart with a one-line reason.
  5. 5. Three volunteers defend one sticky using purpose, data, and accountability. Teacher records a model sentence on the board.

Discussion questions

  1. Which of today’s tasks would you never hand to AI, even if the output looked perfect? Why?
  2. When is “mixed” a stronger answer than a hard yes or no?
  3. How should a rural clinic or a county office decide differently from a large Boise employer?
  4. If a tool is free and fast, why might it still be the wrong tool for schoolwork?

Differentiation

Support

  • Give a filled-in sample memo and a word bank (purpose, data, accountable, mixed).
  • Reduce the independent task to two boxes: best non-AI option and one risk.
  • Allow oral dictation of the memo to a peer scribe or the teacher.

Challenge

  • Require a second Idaho workplace and a different recommendation for each.
  • Ask students to write a one-paragraph policy a shop teacher or coach could post.
  • Compare two tools for the same task and argue which failure is more costly.

Multilingual learners

  • Provide the three-question test and card titles in English and Spanish.
  • Allow the memo reason in the student’s strongest language, with key terms in English.
  • Pair with a bilingual peer for the card sort; the written memo remains individual.

IEP / 504

  • Offer large-print cards and a digital copy of the memo with extra spacing.
  • Allow extra time or a shortened sort (four cards instead of eight) without reducing the three-question test.
  • Permit a graphic organizer instead of paragraph writing; success is the recommendation plus one reason.

Assessment

Formative

  • Warm-up finger votes and the four-card sort reasons during guided practice.
  • Sticky-note placement and the one-line defense in the hands-on close.
  • Teacher circulation notes: every mixed vote names a human check.

Summative

  • Right Tool Memo scored on four traits: purpose named, non-AI option, AI role or refusal, and a named risk.

Success criteria

  • I used the three-question test in my reason, not “AI is cool” or “AI is bad.”
  • I named at least one task that should stay with a person or a traditional tool.
  • I identified who is accountable if the chosen tool is wrong.

Responsible use, ethics, and privacy

Responsible use

Students do not log into public chatbots for this lesson. If a live demo happens, the teacher uses a district-approved tool on a teacher account, with scripted prompts that contain no student names, class lists, or private writing. The default path is paper. Fast is not the same as allowed: a tool that is free on a phone may still be unapproved by the district. When students recommend AI for a task, they must also name a human check. The class rule is simple and posted: we choose tools on purpose, we check the result, and we keep schoolwork that is meant to grow our thinking in our own hands.

Ethics

Appropriateness is not only about accuracy. Using AI to comfort a classmate, to write a college essay in someone else’s voice, or to make a public safety call without a person in the loop treats people as problems to be optimized. Students practice asking who is helped, who is skipped, and who is blamed if the output is wrong. Idaho public schools serve rural towns, Boise-metro suburbs, and tribal communities; a one-size rule would hide those differences. The ethical habit this lesson builds is case-by-case judgment with reasons, not a cheer or a ban.

Privacy

FERPA still applies when the tool is “just a demo.” Never put student names, photos, grades, ID numbers, or private writing into an unapproved tool. Do not project a student’s memo, email, or face. Printed task cards use fictional or public workplace details (Idaho Power, NIFC, a generic packing line), not real patient, student, or employee records. If a student volunteers a family farm story, keep it in the room; do not type it into a prompt. District-approved tools only, and paper whenever that is the safer path.

Reflection

  1. Which task today changed your mind, and which question in the three-question test caused the change?
  2. Where in your own schoolwork is AI a mixed tool rather than a yes or a no?
  3. What will you check by hand this week even if a machine offers an answer?

Homework

Take-home half-sheet (20–25 minutes, no account). Pick one real task you will face this week (homework, a shift, a chore, a sport). Complete the three-question test on paper. List a non-AI option, a possible AI option, and your recommendation in five to eight sentences. Do not use a chatbot. Bring the sheet back; we will swap and peer-check the reasons.

Closing

AI is a tool with a job description, not a default setting. If you can name the purpose, the data, and the person who is accountable, you can choose AI, skip it, or mix it—and you can explain that choice to a boss, a teacher, or a grower. That judgment is the standard, and it travels from this classroom to Idaho work.

Extensions and cross-curricular links

Go further

  • Block period: add a structured debate (AI vs. non-AI vs. mixed) using one wildfire card and one English-class card; require evidence from the three-question test.
  • Invite a local grower, millwright, or county clerk for a 10-minute “what I still check by hand” interview; record a backup audio file for rural bandwidth days.
  • Students rewrite one school honor-code line so it names appropriate vs. inappropriate AI use without banning all tools.
ELA
Purpose and audience decide the tool, the same way they decide a genre. A personal narrative is evidence of the writer, not a product to outsource.
Mathematics
A calculator is a non-AI tool with a clear job. Contrast it with an AI that might invent a number. Students practice asking whether a result is computed or guessed.
Science
Lab conclusions must show the student’s claim, evidence, and reasoning. AI can format a table; it cannot stand in for the observation.
CTE / Social studies
Ag, energy, health, trades, and county government all mix sensors, spreadsheets, and human calls. Civics: who is liable when a public agency uses a model.