Directed Human-AI Coadaptation Method

This document defines the From E™ Directed Human-AI Coadaptation Method: a methodological architecture for observing, directing, and measuring the functional development of a person working with artificial intelligence systems on real problems. The method organises the process into nine phases, from an unassisted baseline measurement to the verification of persistence after a temporal interruption, with seven intermediate measurement states. It incorporates a ten-dimension metric system covering structuring, control, operational autonomy, executable complexity, adaptation in both directions, reciprocity, joint performance, transfer, and persistence, together with complementary metrics for intervention load, sustainable complexity frontier, progressive withdrawal of direction, and recovery under perturbation. The method explicitly distinguishes human learning, system personalisation, bidirectional adaptation, and coadaptation, reserving the latter term for recurrent, longitudinal reciprocal dynamics verified through contrast conditions. The unit of analysis is the observable behaviour of the interaction system — human, AI, task, tools, context, and time — rather than the participant's internal state or the model's parameters. Judgement, responsibility, and decision authority remain with the human. Restricted-access deposit for the purpose of establishing authorship and temporal priority. It does not constitute open publication nor the granting of any licence. The document deliberately excludes the From E™ architecture, its internal mechanisms, its proprietary decision criteria, and the method's core instructions.