---
id: "claim-expertise-redefined"
type: "claim"
source_timestamps: ["§ 3. Help your workforce harness the skills AI is unlikely to master"]
tags: ["expertise", "epistemology", "prompt-engineering"]
related: ["concept-curiosity-hacks", "concept-intellectual-slow-food"]
confidence: "high"
testable: false
speakers: ["Tomas Chamorro-Premuzic"]
sources: ["adoption"]
sourceVaultSlug: "hbr-seg-adoption"
originDay: 9
articleStem: "hbr-edu-36-team-collaborate-with-ai"
sourceUrl: "https://hbr.org/2024/11/set-your-team-up-to-collaborate-with-ai-successfully"
sourceTitle: "Set Your Team Up to Collaborate with AI Successfully"
---
# AI Has Changed the Definition of Expertise from Knowing Answers to Asking Questions

**Claim (confidence: high · testable: no).** Historically, an expert possessed a vast repository of knowledge and knew the answers. With AI now serving as an instant, comprehensive knowledge repository, expertise has shifted. It is now defined by the ability to **ask the right questions** (prompting), **vet and assess** the insights AI provides, **make smart decisions** based on them, and possess the **judgment to ignore** AI outputs when they are incorrect or hallucinated. This is why [[concept-curiosity-hacks]] matter and why the [[concept-intellectual-slow-food]] premium exists.

**Enrichment assessment — strongly supported by prompt-engineering/human–AI discourse; conceptual rather than formally empirical:** Askme360 says the most valuable professionals now interpret, challenge, and act on data rather than compile it. Balanced Scorecard Institute keeps humans central for questioning assumptions and integrating intuition. Academic human–AI cocreation work assigns humans sense-making, problem framing, and value definition; AI pattern recognition and prediction.

**Limits:** Domain experts still need **deep substantive knowledge** — asking good questions presupposes understanding. AI shifts the *emphasis* of expertise (adding questioning/judgment on top of knowledge) rather than eliminating knowledge requirements; over-reliant novices can become over-confident interpreters of flawed outputs.