Synopsis
This post expands Leon Furze’s earlier “expertise problem” argument by introducing a three-dimensional model of expertise for working productively with generative AI. Drawing on Punya Mishra’s domain × technology matrix and adding insights from Dreyfus & Dreyfus, Lave & Wenger and James Paul Gee, Furze distinguishes domain expertise, technological expertise, and a newly foregrounded situated expertise—the reflexive, context-aware capacity to apply and communicate knowledge in real communities of practice.

After explaining why large-language models still hallucinate and therefore reward those who can spot errors, Furze shows how the three dimensions overlap in practice: from the “ivory-tower” academic strong in domain + technology but weak in context, to the social-media power-user rich in technology + situated skills but shallow in discipline knowledge. The ideal sweet spot lies at the centre of a Venn diagram where all three strands meet.

Originally published at: https://leonfurze.com/2025/06/13/three-dimensions-of-expertise-for-ai/

Links
https://leonfurze.com/2025/06/13/three-dimensions-of-expertise-for-ai/
https://leonfurze.com/2024/09/18/expertise-not-included/
https://punyamishra.com/2025/02/13/the-genai-and-expertise-paradox-why-it-makes-expert-work-more-important-but-harder/
https://en.wikipedia.org/wiki/Dreyfus_model_of_skill_acquisition
https://en.wikipedia.org/wiki/Community_of_practice
https://en.wikipedia.org/wiki/James_Paul_Gee
https://help.openai.com/en/articles/8313359-is-chatgpt-biased
https://leonfurze.com/2025/06/04/teaching-ai-ethics-2025-truth/
https://dl.acm.org/doi/10.1145/3442188.3445922
https://www.microsoft.com/en-us/research/publication/the-impact-of-generative-ai-on-critical-thinking-self-reported-reductions-in-cognitive-effort-and-confidence-effects/

The post Three Dimensions of Expertise for AI appeared first on Leon Furze.