Gains:
- Ability to produce questions, examples and materials appropriate to the level and verify each solution independently
- Ability to design distractors to target real misconceptions and point out where analogies break down
- Ability to prevent oversimplification from turning into physical errors and confirm the pedagogical value of the material
Teaching physics is much more than imparting formulas. Students come to physics not with an empty mind, but with intuitions from their daily experiences — and these intuitions are often physically wrong. Misconceptions such as "A moving object requires constant force to stop" or "heavy objects fall faster" are the biggest obstacles to teaching. In this unit, you will learn how to use artificial intelligence (AI) in producing physics course materials, preparing problems and examples, and addressing misconceptions; and you will learn how to audit each material for physical accuracy and pedagogical appropriateness. Basic principle: AI produces blueprints; the teacher confirms accuracy, level and pedagogical value.
Where is AI strong in teaching, where should it be supervised?
AI; It is very fast in generating practice questions, examples from daily life, analogies, explanations adapted to different levels, and multiple choice questions with distractors. But three points require careful monitoring: (1) physical accuracy — AI may produce an incorrect number, unit, or concept; (2) the limit of analogies — every analogy collapses somewhere and can create a new illusion; (3) quality of distractors—incorrect choices of a multiple-choice question should be designed to capture real misconceptions, not randomly.
Quest
Contribution of AI
Teacher supervision
practice question
Draft + generates solution
Physical accuracy, level fitness
Everyday example/analogy
Provides an interesting analogy
Pointing out where the analogy breaks down
Concept explanation
Establishes a simple explanation
Ensuring it does not cause errors
Multiple choice distractor
Suggests wrong choice
Aiming at true delusion
Level adaptation
Simplify text
Don't let oversimplification turn into a mistake
Step by step: reliable course material
1. Clarify the learning objective and level. Which concept, which grade level, which prior knowledge? Like "Newton's second law, for 10th grade high school students who don't know derivatives."
2. Ask the AI for the material, but verify the solution. Independently (by hand or in code) provide solutions to each problem it generates. When generating questions, AI may make mistakes in the solution — especially in sign and unit.
3. Consciously target misconceptions. Ask the AI for questions and distractors that reveal typical students' misconceptions. A good distractor represents a real mistake that the "careless student" will fall into.
4. State the limits of analogies. With each analogy, add the note “this analogy works here but breaks down there.” An analogy without a specified limit creates new errors.
5. Review the pedagogical value. Does the material actually teach the concept or just make you memorize the formula? The purpose of teaching physics is understanding.
Hint: When you have the AI produce text that explains a concept, you also ask “what new misconception might this explanation create in students?” ask. This reverse aspect allows you to foresee the misunderstanding that can arise from oversimplification or an incomplete analogy. For example, it is useful to explain electric current with the analogy of "water in a pipe", but it may lead the student to think that "the current is consumed and ends"; This limit should be stated from the beginning.
three mini cases
Case 1 — Wrong solution key. A teacher asked AI for a 12-question exam and answer key. In two questions, the AI got a sign wrong in the accelerated motion formula and the answer key was wrong. The teacher solved all the questions manually and with a Python code, catching two errors. Without verification, students would be scored incorrectly for their correct answers.
Case 2 — The collapsing analogy. A teacher asked the AI to explain heat conduction with an analogy; AI described heat as a "flowing liquid". This analogy was helpful at first, but students thought heat was a "conserved substance" (caloric fallacy). The teacher prevented the mistake by clearly explaining where the analogy broke down.
Case 3 — Provided a good distractor. A teacher asked the AI for a distractor for a multiple choice question on free falling and had it add a distractor representing the fallacy "heavy objects fall faster." A significant portion of the students chose this option in the exam; The teacher thus had the opportunity to measure and directly address common misconceptions in the classroom. The well-designed distractor became a teaching tool.
Four copyable templates
1) Generation of verifiable questions:
Generate 5 practice questions for [topic, grade level, methods allowed]. Give the complete solution for each question, STEP BY STEP, with unit tracking. I will verify the solutions by hand/code; Mark the step you are not sure about. Keep the difficulty level of the questions appropriate to the specified level.
2) Misconception-targeted distractors:
Write a multiple choice question on [concept]. Let EACH of the incorrect choices represent a ACTUAL misconception students have about this topic. Also explain which misconception each distractor targets. State the correct answer and why it is correct.
3) Analogy with specified limits:
Describe [physical concept] with an everyday life analogy. Then, clearly state WHERE this analogy works, WHERE it breaks down, and what new misconception it might lead to. Level: [grade].
4) Level adaptation + accuracy check:
Simplify the following physics explanation for [target level]. BUT do not let the simplification turn into a physically incorrect statement. Mark if something is lost or distorted while simplifying. Text: [here]
Weak prompt / Strong prompt
Weak: "Write 10 questions and answers about Newton's laws."
Result: Questions with uncertain levels, unverified solutions, and random distractors; There is a high risk of incorrect answer keys.
Güçlü: "For the 10th grade of high school, produce 5 questions on Newton's second law; give the solution of each step by step and with unit tracking. Make one question multiple choice and set up the distractors to target real misconceptions such as 'there is no motion without force'; write down which misconception each distractor represents."
Result: Material that is level-appropriate, verifiable, error-targeted, and has high pedagogical value.
Common mistakes
- Not verifying the answer key. When generating questions, AI may make sign/unit errors in the solution; Each solution must be provided independently.
- Not stating the limit of the analogy. Every analogy collapses at some point; If the limit is not stated, a new mistake will arise.
- Using random distractors. False options that do not target real error neither teach nor measure.
- Oversimplification. Oversimplifying a concept can make it physically incorrect.
- Not stating the level. If the target class and preliminary information are not given, the material will either be too easy or unusable.
Caution: Before giving a course material to students, independently verify the solution to each problem and check each explanation for physical accuracy. An incorrect answer key or a misconception-producing analogy leaves students with misunderstandings that are difficult to correct. AI-produced material is not a substitute for approval from a competent teacher; The responsibility is on the teacher.
In summary
In physics teaching, AI is a powerful drafting tool for generating questions, examples, analogies, and level-adapted explanations. But three points of control are essential: the physical accuracy of the solutions, the limit of the analogies, and the targeting of the distractors to real misconceptions. Good teaching aims at understanding, not memorization; AI material should enter the classroom only after it has been checked by the teacher for accuracy and pedagogy. In the next and final unit, we will bring together the verification discipline that we have studied throughout the module under the headings of hallucination, ethics, confidentiality and scientific integrity.
Application task
Choose a physics subject that you have taught or learned. Have the AI generate a few solved questions with template 1 and a misleading multiple choice question with template 2. Verify each solution manually or with code; Check at least one solution for errors. Then ask for an analogy with the 3rd template and evaluate its limit. Write down in 5-6 sentences: were there any errors in the answer key, where did the analogy break down?
checklist
- [ ] I clarified the learning objective and grade level.
- [ ] I have independently verified the solution to each problem.
- [ ] I checked that the distractors were targeting real misconceptions.
- [ ] I pointed out where the analogy I used breaks down.
- [ ] I have checked that the simplification does not turn into a physical error.
- [ ] I did not give the material to students without verifying its accuracy.