Constructivist psychology principles of human–AI collaboration

Generative artificial intelligence (GenAI) is becoming a part of our lives, personally and professionally. Much of the current discourse focuses on its capabilities and ethical risk, while some important questions remain under explored. For example, how should we collaborate with GenAI? Or what kind of psychological assumptions about humans and knowledge are needed to guide interaction between people and GenAI? This conceptual paper proposes that constructivist psychology, with its focus on meaning making, complexity, anticipation, and agency, offers a promising framework for human-AI collaboration. Rather than treating GenAI as only a tool for increasing efficiency, whose products we passively consume, the paper invites for rethinking the collaboration as co-construction between human and artificial meaning making. First, the paper traces the historical convergence of metaphors between computer science and psychology. What follows is an articulation of core constructivist principles for human-AI collaboration. The paper also outlines a conceptual foundation for designing AI agents that are based on the principles of constructivist psychology. Finally, implications, limitations and future studies are outlined. The general genre of the paper is not deconstructive in terms of unravelling biases or positioning that GenAI may perform, but an exploration of how we can redirect existing biases and positioning with principles of constructivist psychology.