From Human-in-the-Loop to Human-in-the-Design: Friction, Choke Points, and the Control of Fast Work Systems

This conceptual paper argues that human oversight in AI-enabled work cannot be established merely by placing a person “in the loop.” A work system remains controllable only while it can detect and correct deviations before they compound beyond the available response. When AI accelerates the production or propagation of work beyond the work system’s capacity to verify it, the appearance of human oversight may remain after meaningful verification has become structurally unavailable. The paper develops Human-in-the-Design (HITD) as an engineered condition of the workflow rather than a person’s nominal position within it. Building on the author's prior work defining Verification Overrun, the paper introduces Verification Requirement (VReq), Verification Capacity (VCap), the Verification Window, Verification Gates, Verification States, and the Workflow Profile - integrating these concepts into a broader architecture for workflow control. These concepts are organized around three design obligations: Provision, which requires the organization to supply the capacity needed for verification under the operating conditions in which it must function; Preservation, which protects friction, slack, and constraints as conditions of control; and Placement, which concentrates control at consequential transitions while verification and intervention remain possible. The paper establishes the conceptual and design foundation for HITD within the broader Human, AI, and Organizational Performance (HAOP) framework. A subsequent HAOP Workbook will translate the framework into a practitioner method for examining and redesigning specific work systems. Version 1.1: Clarifies the boundary of the Verification Window by including "or materially difficult to correct" after "irreversible." Added description of correction paths into the Workflow Profile section. Other minor editing changes made, as well. The paper’s central argument and framework remain unchanged.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-07-27
DOI
https://doi.org/10.5281/zenodo.21631544
Primary Topic
Human-Automation Interaction and Safety
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

From Human-in-the-Loop to Human-in-the-Design: Friction, Choke Points, and the Control of Fast Work Systems

Jaina Ko
Zenodo (CERN European Organization for Nuclear Research)
Human-Automation Interaction and Safety
preprint

From Human-in-the-Loop to Human-in-the-Design: Friction, Choke Points, and the Control of Fast Work Systems

Jaina Ko
preprint en

Abstract

This conceptual paper argues that human oversight in AI-enabled work cannot be established merely by placing a person “in the loop.” A work system remains controllable only while it can detect and correct deviations before they compound beyond the available response. When AI accelerates the production or propagation of work beyond the work system’s capacity to verify it, the appearance of human oversight may remain after meaningful verification has become structurally unavailable. The paper develops Human-in-the-Design (HITD) as an engineered condition of the workflow rather than a person’s nominal position within it. Building on the author's prior work defining Verification Overrun, the paper introduces Verification Requirement (VReq), Verification Capacity (VCap), the Verification Window, Verification Gates, Verification States, and the Workflow Profile - integrating these concepts into a broader architecture for workflow control. These concepts are organized around three design obligations: Provision, which requires the organization to supply the capacity needed for verification under the operating conditions in which it must function; Preservation, which protects friction, slack, and constraints as conditions of control; and Placement, which concentrates control at consequential transitions while verification and intervention remain possible. The paper establishes the conceptual and design foundation for HITD within the broader Human, AI, and Organizational Performance (HAOP) framework. A subsequent HAOP Workbook will translate the framework into a practitioner method for examining and redesigning specific work systems. Version 1.1: Clarifies the boundary of the Verification Window by including "or materially difficult to correct" after "irreversible." Added description of correction paths into the Workflow Profile section. Other minor editing changes made, as well. The paper’s central argument and framework remain unchanged.

Zenodo (CERN European Organization for Nuclear Research)
Human-Automation Interaction and Safety
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.