Process-Constituted Intelligence: A Shared Criterion for Humans and Machines

2026-08-17

AI Summary

This study addresses the limitation that generative AI merely mimics cognitive traces without genuine reasoning, resulting in weak equivalence and risks of human capability degradation. To counter this, we propose the theory that "intelligence is constituted by process." By defining seven characteristics of strong equivalence, we construct a process audit framework and design principles for human-AI cognitive alignment. This research establishes evaluation criteria for strong human-AI equivalence and proposes AI design guidelines alongside testable auditing protocols to prevent human skill atrophy. Ultimately, this work provides both a theoretical foundation and a practical pathway for mitigating cognitive degradation risks and achieving deep cognitive alignment between humans and artificial intelligence systems.

Abstract

Intelligence is constituted by \textit{process} (iterative activity through which output emerges), not in the output itself. Generative AI (GenAI) is trained on \textit{traces} (textual and visual residues of human cognitive processes), reproducing samples from a distribution of those traces. Its outputs resemble reasoning, problem-solving, and creativity, yet the activity that produces such outputs in humans remains largely absent. Current GenAI is, therefore, weakly equivalent to the cognition it imitates, matching outputs while process stays absent or opaque. The cognitive sciences have long distinguished between weak and strong equivalence. Here, we define \textit{strong} equivalence across seven process features, assessable against human and machine cognition. Our process-based account addresses a symmetric risk: GenAI tools that outsource a person's generative processes may leave critical capacities unbuilt. We specify design principles for GenAI that instantiate more process and preserve rather than erode human judgment and creativity, and outline process audits that make strong equivalence testable.

Problem

• Process-Constituted Intelligence
• Strong Equivalence
• Generative AI
• Cognitive Process
• Human Cognition

Innovation

• Process-Constituted Intelligence
• Strong Equivalence
• Generative AI
• Process Audits
• Cognitive Process