ART-AI and AI/ML group Seminar
We are delighted to welcome Kris Lee, Founder and CEO of ArtSci AI, for a joint ART-AI and AI/ML Group seminar entitled: From Tacit to Explicit: The Art and Science of Accountable Human-Machine Teaming. Abstract and Bio below.
Date: Tuesday 27 October 2026
Time: 10:00am to 11:00am (GMT)
Location: Online via Microsoft Teams
Charles Larkin, Director of Research at the Institute for Policy Research at the University of Bath, will chair the seminar. For more information or if you would like to attend, please e-mail [email protected].
Title
From Tacit to Explicit: The Art and Science of Accountable Human-Machine Teaming
Abstract
Generative AI is exposing a longstanding weakness in expert knowledge work: much of the reasoning that gives professional analysis its value remains tacit, personal, and difficult to inspect. Intelligence analysis offers a useful case study. Analysts must work with incomplete and conflicting information, make judgements under uncertainty, communicate consequential conclusions, and increasingly do so at machine speed.
This talk examines the false choice between continuing to rely on unstructured human intuition that cannot keep pace and blindly delegating complex reasoning to automated systems. It argues that simply placing a human in, or at the end of, an AI pipeline is not the same as having humans deliberately design and orchestrate the system itself.
Accountability must extend across the full knowledge workflow: framing the problem, translating expert reasoning into structured task profiles, building and refining AI-supported workflows, and systematically validating the results for use at speed and scale.
Drawing on the concept of Knowledge Workflow Design, Kris Lee will discuss how expert heuristics, decision logic, and quality criteria can be translated into explicit, inspectable, and reusable workflow components. The goal is to make human expertise more visible, teachable, scalable, and accountable in human-AI systems.
Bio
Kris Lee is the founder and CEO of ArtSci AI, a knowledge workflow design firm specializing in AI-augmented intelligence analysis. He spent 25 years at the Central Intelligence Agency, where his work centered on open-source intelligence, analytic methodology, data-driven tradecraft, and the development of AI-supported intelligence capabilities. His final roles there were chief of data workflow innovation and senior methodologist and chief strategist for OSINT. Through ArtSci AI, Kris builds the cognitive architecture behind trustworthy human-AI analysis: expert knowledge scaffolds that make a veteran analyst’s reasoning explicit enough for AI systems to apply at scale, without breaking the chain of accountability that gives the resulting analysis its credibility.
