Technology field notes
Teach AI verification with a receipt that is almost right
A 20-minute lesson for teams that want to catch a plausible error before it reaches a customer.

In this guide
What you can try
A 20-minute lesson for teams that want to catch a plausible error before it reaches a customer.
Give a learner a fictional order: two notebooks at $12 each, one $6 pen, and $4 shipping. The approved total is $34. Now hand them a polished AI-written reply confirming “your two notebooks, pen and free shipping for $30.” Most of the words are correct. The mistake is exactly the kind a busy reviewer can read past.
For the first five minutes, ask each learner to underline every claim in the reply and connect it to the order. Quantities, price, shipping and total each get their own mark. The person who checks only the arithmetic will miss that the assistant changed the offer.
Give them a different mistake on the second pass
Show a second order with free shipping explicitly approved, but put a quantity of three in the drafted reply where the order says two. This time, cover the first example. Ask learners to explain their checking procedure to a partner while using it. The Institute of Education Sciences practice guide recommends moving between worked examples and independent problems and using retrieval during instruction. Institute of Education Sciences • Organizing instruction and study
Close with a short response the class can use in a real workflow: “I verified quantity, unit price, shipping terms and total against the approved order. The unresolved detail is ___.” Keep the unresolved field available even when the learner expects a complete answer.
The next day, give the team an unfamiliar but equally small task, such as checking a service date and location. A person who can transfer the method has learned how to verify AI output. A person who memorized $34 has learned the answer to yesterday’s exercise.
Sources and further reading
Sources checked November 18, 2026. Review the linked publisher information for current details.
Put one idea to work
Use one lesson in a small project, explain what you learned, and practice again without relying on the answer.
Try the free workshop →Explore more from AI TeachersGet the free AI at Work Starter Kit →Keep practicing
Teach Your Team to Evaluate AI Work
A practical lesson for turning a promising AI response into a repeatable workflow that people can explain, check, and improve.
November 16, 2026 · 5 min readAgent-Friendly Website Lesson: llms.txt and Task Testing
A hands-on lesson plan for llms.txt, accessible pages and public APIs, including a fictional site, a six-step exercise and an assessment rubric.
November 11, 2026 · 8 min read