The Refraction Principle: How AI Bends
Source: Nick Potkalitsky Substack
Author: Nick Potkalitsky
Original source: https://nickpotkalitsky.substack.com/p/the-refraction-principle-how-ai-bends
Private backup: the full article text is archived in the private repository at archives/articles/nickpotkalitsky-substack-com-the-refraction-principle-how-ai-bends.source.md. It is not published on the public Quartz site.
Summary
Nick Potkalitsky presents, with Terry Underwood, a framework for understanding AI-mediated learning as a transformation of human intention rather than a replacement of it. The article proposes a movement from seminal intention, through AI as refractive medium, toward hybrid intention, where the learner’s purpose has been clarified and reshaped through AI interaction but remains human-owned. It also distinguishes centrifugal prompting for divergent exploration from centripetal prompting for focused convergence.
Big ideas
Claims
- AI can reshape a student’s purpose, but it should not replace it
- Students need boundaries for when to use AI and when to step back
Key evidence and examples
- The article argues that intentionality remains anchored in human consciousness through the questions asked, problems posed, and goals pursued.
- It defines AI as a refractive medium that can surface assumptions, multiply perspectives, scaffold logic, and disambiguate concepts without generating its own intentional content.
- It gives examples of AI refining broad student intentions into more precise inquiries in democracy, poetry, standardized testing, and quantum mechanics.
- It defines hybrid intention as a transformed but still human-owned intention.
- It distinguishes centrifugal AI use for brainstorming, research, and multi-perspective exploration from centripetal AI use for targeted problem-solving, precision writing, and skill development.
Education relevance
This article is directly relevant to AI literacy, writing instruction, research pedagogy, and student metacognition. It offers a way to teach students how AI can refine their purposes while still requiring them to name, steer, and own those purposes.