Analyze pre-generated examples
Use curated screenshots, frame strips, prompt packs, or facilitator-provided outputs. Your job is to name defaults, assumptions, failure modes, and evidence without producing anything new.
Critical / No-AI pathway
You can investigate AI systems without generating new AI content.
This pathway is not a fallback or a consolation prize. It is a full participation route for people who object to direct tool use, lack access, need a lower-risk classroom option, or want to focus on evidence, consent, policy, and critique.
Use curated screenshots, frame strips, prompt packs, or facilitator-provided outputs. Your job is to name defaults, assumptions, failure modes, and evidence without producing anything new.
Act out prediction, pattern matching, diffusion, temporal drift, or model-card review with people, paper, cards, whiteboards, or classroom materials.
Write an evidence-based critique of a system, output, policy, dataset, interface, classroom use case, access tier, or consent risk.
Turn one camp idea into a lesson, worksheet, discussion protocol, gallery walk, unplugged game, or assessment that works without student logins or generation.
Define what can be recorded, prompted, screenshotted, named, archived, or shared. Make consent visible before anyone is asked to participate publicly.
Use frozen examples to compare vague prompts, revised prompts, model defaults, confident claims, video failure modes, and access differences.
Bring one thing that shows what you found: a critique, model card, consent checklist, classroom activity, annotated example, comparison, or unplugged protocol. It does not need to be generated by an AI system.
Point to the evidence behind the artifact and name what can be shared publicly. If the work involves another person's words, image, or classroom context, consent comes first.