SIFT for AI: Introduction and Pedagogy

Source: Mike Caulfield Substack
Author: Mike Caulfield
Published: 2025-11-22
Source type: guide
Original source: https://mikecaulfield.substack.com/p/sift-for-ai-introduction-and-pedagogy

Private backup: the full article text is archived in the private repository at archives/articles/mikecaulfield-substack-com-sift-for-ai-introduction-and-pedagogy.source.md. It is not published on the public Quartz site.

Summary

Mike Caulfield explains why AI literacy needs the same kind of action-oriented pedagogy that made SIFT useful for web information literacy. He argues that the point is not merely critical thinking about AI, but critical doing with AI: students use LLMs to map an information landscape, then verify, evaluate, discuss, and synthesize claims through disciplinary lenses. Examples such as the “Chocolate Memory” task and a scene from the 1927 film Wings show how AI can scaffold inquiry without replacing student reasoning.

Pull quotes

Critical doing

“What people liked about SIFT, and how it changed the approach to information literacy… first and foremost, it was its focus on critical doing.”

Doing before thinking

“Underneath it all was our realization: the secret to knowing things is doing things. Thinking is important, but it rewards those who engage in the “doing before the thinking” — taking steps to figure out where the thing they are looking at came from, seeing what others know about the event or claim.”

Action framework

“Second, I took those sets of honed techniques and we put them in an action framework that focused not on seven or ten categories of things to think about, but the most important categories of things to do.”

Big ideas

• Students need to check AI answers against real evidence
• AI simulations need clear boundaries for learning

Claims

• AI literacy should teach students what to do with AI, not just what to think about it
• AI-assisted inquiry should ground claims in evidence

Key evidence and examples

• Caulfield reframes “SIFT for AI” as educators asking for usable moves, techniques, and pedagogy rather than a replacement for SIFT.
• He defines SIFT’s power as critical doing: concrete actions, information foraging, shared findings, and synthesis.
• The “Chocolate Memory” task asks students to trace an AI-mediated claim back to original research and distinguish chocolate from flavanols.
• Students use AI to generate tables of evidence and rebuttals, then discuss questions about mice studies, fMRI, observational versus intervention studies, and supplements versus chocolate.
• The Wings example shows AI surfacing historical interpretive lenses that students can then investigate and debate.

Education relevance

This is very relevant to AI literacy instruction, inquiry-based learning, disciplinary reasoning, and classroom activities that use AI while preserving student responsibility for verification and synthesis.

Durability note

This is a durable pedagogical frame because it translates SIFT into AI-era habits of action rather than tying the lesson to one model or interface.

My notes