Tutorials · Chapter B (2/4) · ~10 min

AI for Teachers

Try it → see it → read → next

Design learning support without outsourcing judgment.

Try yourself

Teachers

Differentiator board

Pick a student level. AI adapts the explanation — you keep the judgment checklist.

What is temperature in a language model?

Student level

Recap

What you just did

DifferentiatorBoard adapted an explanation by level while keeping teacher judgment checked on.

Teach

How it works

Begin with the learning target, not “make a worksheet”:

Learning target: students can compare fractions with unlike denominators.
Learners: age 10, mixed confidence.
Create:
1) a 3-minute explanation using a pizza example
2) one worked example
3) four practice questions from easy to challenging
4) an answer key with common misconceptions

Use plain language. Do not collect or mention real student data.

Then switch modes without changing the goal:

Keep the same learning target.
Rewrite question 3 as a hint ladder:
- first hint points to a strategy
- second hint starts the setup
- final step explains the answer
Do not reveal the answer in hint 1.

Review every output for correctness, accessibility, reading level, and cultural assumptions. A generated rubric or lesson plan is a draft; teacher judgment decides what enters the classroom.

Use it

When you'd use this

Producing extra examples at varied difficulty
Turning a lesson into retrieval practice or exit tickets
Rephrasing instructions for clarity and accessibility

Watch out

Do not paste identifiable student records, disability details, grades, or private writing into an unapproved tool. Never let AI make final grading, discipline, or placement decisions. Follow your school’s policy and provide human review.

Try next

Try this next

Pick one learning target. Generate an explanation, a misconception check, and a three-question exit ticket. Solve every question before sharing it.