This essay discusses the attempts of progressive artificial intelligence (AI) models to understand behavior. Because of their sophistication and complexity, it is difficult to understand how these models work and therefore it is difficult to make use of them to understand behavioral phenomena. This is indeed the problem with the present state of cognitive psychology that is founded on the analogy between human behaviors and computer operations: if we do not understand progressive AI models, the most successful and sophisticated programs for predicting behavior, then perhaps this should be interpreted as a warning that cognitive psychology is approaching the limits of its explanatory power. This paper develops and backs up this proposal.
|Number of pages||14|
|Journal||Journal of Mind and Behavior|
|State||Published - 1 Jun 2022|
Bibliographical notePublisher Copyright:
© 2022 The Institute of Mind and Behavior, Inc.
- cognitive psychology
- explainable AI
ASJC Scopus subject areas
- Experimental and Cognitive Psychology
- Arts and Humanities (miscellaneous)