Machine Psychology: Integrating Operant Conditioning with the Non-Axiomatic Reasoning System for Advancing Artificial General Intelligence Research

TL;DR

This work addresses the lack of a coherent theoretical framework for artificial general intelligence by proposing Machine Psychology, which unifies operant conditioning with the Non-Axiomatic Reasoning System (NARS) to enable adaptive learning in AGI. The authors evaluate this framework using three operant-learning tasks (simple discrimination, changing contingencies, and conditional discrimination) implemented with OpenNARS for Applications (ONA) to demonstrate rapid learning, real-time adaptability, and complex hypothesis formation under resource constraints. Results show that NARS-ONA can learn from feedback, adapt when contingencies change, and form higher-order conditional relations, supporting operant conditioning as a viable guiding principle for AGI and highlighting NARS’ robustness in uncertain, data-scarce environments. The work suggests that integrating sensorimotor reasoning with operant psychology provides a scalable, flexible pathway for real-world AGI deployment and points to future extensions into broader cognitive frameworks and tasks.

Abstract

This paper introduces an interdisciplinary framework called Machine Psychology, which merges principles from operant learning psychology with a specific Artificial Intelligence model, the Non-Axiomatic Reasoning System (NARS), to enhance Artificial General Intelligence (AGI) research. The core premise of this framework is that adaptation is crucial to both biological and artificial intelligence and can be understood through operant conditioning principles. The study assesses this approach via three operant learning tasks using OpenNARS for Applications (ONA): simple discrimination, changing contingencies, and conditional discrimination tasks. In the simple discrimination task, NARS demonstrated rapid learning, achieving perfect accuracy during both training and testing phases. The changing contingencies task showcased NARS's adaptability, as it successfully adjusted its behavior when task conditions were reversed. In the conditional discrimination task, NARS handled complex learning scenarios effectively, achieving high accuracy by forming and utilizing intricate hypotheses based on conditional cues. These findings support the application of operant conditioning as a framework for creating adaptive AGI systems. NARS's ability to operate under conditions of insufficient knowledge and resources, coupled with its sensorimotor reasoning capabilities, establishes it as a robust model for AGI. The Machine Psychology framework, by incorporating elements of natural intelligence such as continuous learning and goal-driven behavior, offers a scalable and flexible approach for real-world applications. Future research should investigate using enhanced NARS systems, more advanced tasks, and applying this framework to diverse, complex challenges to further progress the development of human-level AI.

Figures (5)

• Figure 1: An overview of the architecture in OpenNARS for Applications (ONA). Reprinted with permission from Patrick Hammer, the author of ONA.
• Figure 2: Examples from the three experimental tasks investigated.
• Figure 3: Operant conditioning. Dots illustrate the percent of correct in blocks of 12 trials. The solid line shows the mean NARS confidence value for hypotheses.
• Figure 4: Operant conditioning with changing contingencies. Dots illustrate the percent of correct in blocks of 12 trials. The solid lines show the mean NARS frequency values for the respective hypotheses.
• Figure 5: Conditional discriminations. Dots illustrate the percent of correct in blocks of 12 trials. The solid line shows the mean NARS confidence value for hypotheses.