Straying from the pragmatic ground of natural intelligence

Abstract

We are in the vestibule of an era when Artificial Intelligence seems destined not only to manage, but to govern, the flow and use of information. Because the impact on human life and culture will surely be profound, it is imperative that the conception of intelligence that guides AI is conformable to human intelligence. Neuroscientist and AI developer, Jeffrey Hawkins, is convinced that AI is on the wrong path with its reliance on large language (big data) models, and that to achieve Artificial General Intelligence (AGI), the near-universal goal of AI developers, a biologically-inspired approach is needed. He has developed a promising new natural theory of intelligence based on the functioning of the human neocortex, but he dismisses old brain processes as mainly unnecessary and counter-productive for post-Darwinian AI. The naturalist conception of intelligence developed by the classical pragmatists a century and a half ago, provides a richer theoretical framework of intelligence that situates intelligent agents in interactive environments, the pragmatic ground of natural intelligence. Peirce, in particular, with his life-long quest to deeply understand how knowledge can be gleaned from experience, developed theories of perception, belief formation, semiosis, and cognitive logic, which taken together constitute the basis for a comprehensive theory of intelligence that incorporates instinctive and emotional intelligence and that allows for a conceptual space where thought can be focused on complex theoretical, normative, and aesthetic concerns and interests. This pragmatic account of intelligence, of mentality in general, is necessary for understanding the limitations and risks of AI as it rushes to develop AGI.