When the machine sounds sure: an AI literacy seminar
A TOK-style seminar where students grade AI answers that are fluent, confident — and wrong.
How to use it
Groups receive three AI answers on the same prompt (one solid, one subtly wrong, one fabricated) and must rank them with evidence before the reveal.
Adaptations and limits
For younger groups, use image-based cases instead of essays.
Why it works
Confident-error cases train calibration better than abstract warnings; the reveal moment anchors the lesson.
Take this resource to your library
With your free account you keep it, adapt it to your group and get an unbranded classroom link to share the activity with your class.
Editorial resources are free for everyone; with an account they are also yours.
Staff room
★ 4,9 · recommended by 35 teachers
-
Ana Ferreiro Deputy head · British school, Madrid
We used it as-is in the department meeting and then in class: the adaptation levels save the hard part, which is differentiating without rebuilding everything.
-
Marcus Webb IB Coordinator · International school, Lisbon Community
With Year 11 it works better splitting the session in two: first half guided, second with the material in students' hands. Suggestion: add a 5-minute plenary.
-
Lucía Prat Secondary tutor Community
I shared it with my class through the classroom link, no sign-ups and no branding: it looks clean on the board and every student opened it on their tablet first try.
It will be published anonymously if it passes editorial review.
Received — thank you.
It will go through editorial review and be published anonymously.
Rather sign it and follow the conversation? Create your account with just your email.
Join in to take part
Commenting, rating and uploading your resources is free with your account.
Got a better version or a resource of your own? Share it with the community.