Work

The Delegation Map: Locating Wants Evaporation, Will Chilling, and Skill Evaporation in Human–AI Work

Year: 2026
Type: preprint
Abstract: If AI systems compose better music, write more engaging novels, and solve and pose mathematical problems faster than any person, people may come to find it pointless to do these things themselves and stop thinking. This note provides an observation framework that researchers in any field can use to examine that possibility, together with its method. The Delegation Map follows one activity from WANT (wanting to do something) through I think I can (the person's own prediction of capability), WILL (deciding to begin), and QUESTION (a question arises) to ten origin-tracking indicators, and then along four routes to a completed product: I don't do (delegate and execute), I don't verify (execute without verification), and I can control, which divides into I actually can and I cannot do by myself. Across this map the note separates three layers of change. Wants Evaporation is the loss of a want the person once had. Will Chilling is the repeated choice not to begin while the want and the skill remain. Skill Evaporation is the unnoticed loss or non-formation of the capacity to perform without AI, which the person discovers only when they try to use it. For each layer the note states what must be observed, what must be kept apart from it, and three levels of evidence: a difference observed at one time, a change that persists across repeated measurement, and a causal test. A record of delegation then becomes a route map that shows at which stage a risk arises, and later researchers can ask at which point a question, a criterion, or a choice was delegated and how closely high delegation is linked to failures in practice.
Language: English
Open Access status: green
OpenAlex ID: W7237350545
Cites:
Cited by:
Related to: