AI-as-learning-companion: leveraging LLM fallibility through a critique-refinement cycle to develop higher-order thinking

Design · All designs
Evidence · 2 claims (2 for) · 1 study (1 design), q2 · 1 of 1 report an effect size · 2 claims rest on one study

Description

The paper's central framework reframes generative AI errors and hallucinations as a pedagogical resource rather than a limitation. By treating AI as a "learning companion" whose imperfect outputs prompt analysis, evaluation, and reflection, instructors structure an "iterative critique-refinement cycle" in which "AI's imperfect outputs serve as catalysts for evaluation, positioning critique as the central mechanism driving higher-order cognition". The cycle of generation, evaluation, and revision is aligned with Bloom's taxonomy, shifting emphasis from knowledge and application toward analysis, evaluation, and creation, and fostering metacognitive engagement.

Design Implications

Context

Requirements

• Instruction must extend beyond domain knowledge to include the ability to evaluate, contextualize, and revise AI-generated outputs
• Structured opportunities for evaluation and revision with imperfect responses

Constraints

• Over-reliance on AI can impede the development of core cognitive skills, so use must be scaffolded

Target Learners

• undergraduate computer science students

Learning Goals

• higher-order thinking (analysis, evaluation, creation)
• metacognitive skills including self-monitoring and error detection
• critical evaluation of AI-generated solutions

Claims

• Ai Integrated Course Large Prepost Db Gains [+M]
• Perceived Ai Ability Misaligns Objective Competency [+M]

Related Designs

• Four pedagogical principles for AI companion design derived from the HELF

Examples

-

Key Sources

• Hosseini, H. (2026). The Pedagogy of AI Mistakes: Fostering Higher-Order Thinking. https://arxiv.org/abs/2605.05472. doi:10.48550/arxiv.2605.05472

Page metadata

Field	Value
type	design
id	ai-learning-companion-critique-refinement-framework
title	AI-as-learning-companion: leveraging LLM fallibility through a critique-refinement cycle to develop higher-order thinking
description	The paper's central framework reframes generative AI errors and hallucinations as a pedagogical resource rather than a limitation.
status	review
generated	by process:settle-candidates on 2026-10-11
trust	unverified — no one has confirmed this page's content
sources	1 — hosseini-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.