Beyond Quantification: Navigating Uncertainty in Professional AI Systems
Abstract The growing integration of large language models across professional domains transforms how experts make critical decisions in healthcare, education, and law. While significant research effort focuses on getting these systems to communicate their outputs with probabilistic measures of reliability, many consequential forms of uncertainty in professional contexts resist such quantification. A physician pondering the appropriateness of documenting possible domestic abuse, a teacher assessing cultural sensitivity, or a mathematician distinguishing procedural from conceptual understanding all face forms of uncertainty that cannot be reduced to percentages. This paper argues for moving beyond simple quantification toward richer expressions of uncertainty essential for beneficial AI integration. We propose participatory refinement processes through which professional communities collectively shape how different forms of uncertainty are communicated. Our approach acknowledges that uncertainty expression is a form of professional sense-making that requires collective development rather than algorithmic optimization.
Authors
- Vincent Fortuin (ORCID: https://orcid.org/0000-0002-0640-2671)
- Yulan He (ORCID: https://orcid.org/0000-0003-3948-5845)
- Umang Bhatt (ORCID: https://orcid.org/0000-0002-4611-1668)
- Jessica Montgomery
- Mennatallah El‐Assady (ORCID: https://orcid.org/0000-0001-8526-2613)
- Carl Henrik Ek (ORCID: https://orcid.org/0000-0003-1302-6309)
- Neil D. Lawrence (ORCID: https://orcid.org/0000-0001-9258-1030)
- Sylvie Delacroix (ORCID: https://orcid.org/0000-0002-8517-7782)
- Diana Robinson
- Tom Diethe (ORCID: https://orcid.org/0000-0002-0776-5407)
- Ivana Dusparić (ORCID: https://orcid.org/0000-0003-0621-5400)
- Søren Hauberg (ORCID: https://orcid.org/0000-0001-7223-877X)
- Gaël Varoquaux (ORCID: https://orcid.org/0000-0003-1076-5122)
- Neill Campbell (ORCID: https://orcid.org/0009-0000-3974-7521)
- Jacopo Domenicucci
Institutions
- King's College London (GB)
- Tohoku University (JP)
- Trinity College (CA)
- Trinity College Dublin (IE)
- Helmholtz Zentrum München (DE)
- ETH Zurich (CH)
- Bridge University (SS)
- AstraZeneca (Brazil) (BR)
- Centre Inria de Saclay (FR)
- University College London (GB)
- Technical University of Munich (DE)
- Technical University of Denmark (DK)
Publication Details
- Journal
- RSS Data Science and Artificial Intelligence
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1093/rssdat/udaf002
- Citations
- 2
- Primary Topic
- Artificial Intelligence in Healthcare and Education
- Type
- article
- Field-Weighted Citation Impact
- 1.31