Constructionist AI literacy means students learn AI by building and testing things
Current synthesis
AI literacy becomes more coherent when students use AI to build, test, revise, annotate, and explain something inspectable, rather than only receiving answers from a tool or discussing AI risks abstractly. AI literacy should teach students what to do with AI, not just what to think about it We could be building a better world with our students. Why are we disempowering them instead? What Happened When We Taught AI Literacy Like Writing
Caulfield provides the clearest constructionist version: students build AI-assisted classifiers and data projects, audit outputs, revise categories, and explain the disciplinary judgments behind the work. We could be building a better world with our students. Why are we disempowering them instead? Agentic AI can preserve thinking when students have to design the work
Kentz’s classroom examples point in the same direction from the assessment side: AI interactions become useful when students question, annotate, compare, and reflect on them as evidence of thinking. When Students Interview Jay Gatsby What Happened When We Taught AI Literacy Like Writing AI chat transcripts can make student thinking visible
Potkalitsky’s system-altitude frame should inform this synthesis lightly: it supports the need for a portfolio of AI interaction contexts, but the wiki does not yet have enough independent material to turn altitude itself into a separate district-planning synthesis. Beyond the Hype: Why Your School’s AI Literacy Strategy Needs System Altitude AI literacy requires different kinds of AI interaction
The practical implication is that strong AI literacy tasks should leave behind inspectable artifacts: classifiers, transcripts, annotations, decision logs, revised prompts, models, source checks, or other evidence that students made judgments rather than merely received outputs. Agentic AI can preserve thinking when students have to design the work AI literacy should teach students what to do with AI, not just what to think about it In an AI world, schools need visible thinking, not just policing final products
Practical implications
- Design AI literacy tasks around making, testing, and revising artifacts, not only discussing appropriate use.
- Ask students to document the judgments they made: definitions, categories, constraints, rejected outputs, source checks, and revisions.
- Treat AI chats, classifiers, models, and generated drafts as evidence to inspect rather than as finished products to accept.
- Keep system-altitude and equity questions nearby, but do not force them into a synthesis until more sources independently support those frames.
Related pages
- AI literacy has to be taught inside real subjects
- Students need to check AI answers against real evidence
- AI is changing what knowledge work asks people to do
- In an AI world, schools need visible thinking, not just policing final products
- AI literacy needs a mix of interaction contexts
Source trail
- We could be building a better world with our students. Why are we disempowering them instead?
- What Happened When We Taught AI Literacy Like Writing
- When Students Interview Jay Gatsby
- Beyond the Hype: Why Your School’s AI Literacy Strategy Needs System Altitude
Synthesis history
- Created after Clay approved synthesis-review feedback on 2026-06-20: proceed with constructionist AI literacy, defer system altitude as district planning, and skip the equity thread for now.