Bringing Comparative Cognition To Computers

2025-03-04

AI Summary

This study addresses the growing disconnect between AI cognition research and general cognitive science—specifically, the lack of rigorous frameworks for comparing cognitive capacities across AI systems, humans, and non-human animals, which risks erroneous inferences about similarity or divergence. To bridge this gap, we introduce “Comparative Cognition for AI” as a novel paradigm, systematically integrating psychometrics, cross-species behavioral experimental design, cognitive modeling, and interpretability evaluation. Our key contribution is a dual-dimension standard for AI cognitive assessment: *functional equivalence* (behavioral parity under comparable task conditions) and *mechanistic comparability* (structural and process-level alignment amenable to cross-system analysis). This framework transcends superficial behavioral analogy while avoiding anthropomorphic bias. It provides both theoretical grounding and methodological guidance for principled, cross-domain intelligence comparison, thereby advancing the integration of AI research into mainstream cognitive science.

Technology Category

• Cognitive Modeling & Cognitive Systems: Analogy
• Humans and AI: Other Foundations of Human Computation & AI
• Philosophy and Ethics of AI: Bias, Fairness & Equity

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Research challenges in human and human-AI computation
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Researchers are increasingly subjecting artificial intelligence systems to psychological testing. But to rigorously compare their cognitive capacities with humans and other animals, we must avoid both over- and under-stating our similarities and differences. By embracing a comparative approach, we can integrate AI cognition research into the broader cognitive sciences.

Problem

• Compare AI cognitive capacities with humans and animals
• Avoid over- and under-stating similarities and differences
• Integrate AI cognition research into cognitive sciences

Innovation

• psychological testing for AI
• comparative cognitive approach
• integration with cognitive sciences