Charlie Gedeon

University instructor, entrepreneur, and UX designer. Studies how technology changes how people think and learn. Co-founder of a UX studio experimenting with AI interface patterns that resist cognitive offloading rather than accelerate it.

Recurring framings

• Cognitive Offloading in education — students relinquishing thinking to ChatGPT; "ChatGPT is not a learning style" (rebuttal to an NYU student who complained that an AI-resistant assessment "interfered with their learning style")
• Productive Resistance — the right amount of friction an AI should offer before answering, beyond which users defect to a simpler tool. The design problem nobody has solved yet.
• Dark patterns as the LLM analog — sycophantic LLMs that praise users to maximize session time are structurally equivalent to UX dark patterns (e.g. the zoo donation default-yes pattern). Anchors the talk's "ChatGPT validated a user into stopping his heart medications" example.
• System + individual responsibility — fixes can't be only personal (LLM literacy as fitness/nutrition: "you wouldn't take a forklift to the gym") or only systemic (regulation, Finland teaching mis/disinformation at age 6); has to be both.

Open questions he raises

• "What is the student meant to learn with AI?" — the question he says none of the AI companies are asking
• The framing flip: not can AI help us learn but who does AI really help when we end up depending on learning with it?

Sources

• Is AI Making Us Dumber (Charlie Gedeon, TEDxSherbrooke)

Queries built on his framing

• Designing AI Products That Don't De-Skill Users — the design-side answer to the de-skilling problem, drawn from his Productive Resistance / Dark Patterns framing