Reclaiming AI as a theoretical tool for cognitive science

Abstract

The idea that human cognition is, or can be understood as, a form of computation is a useful conceptual tool for cognitive science. It was a foundational assumption during the birth of cognitive science as a multidisciplinary field, with Artificial Intelligence (AI) as one of its contributing fields. One conception of AI in this context is as a provider of computational tools (frameworks, concepts, formalisms, models, proofs, simulations, etc.) that support theory building in cognitive science. The contemporary field of AI, however, has taken the theoretical possibility of explaining human cognition as a form of computation to imply the practical feasibility of realising human(-like or -level) cognition in factual computational systems; and, the field frames this realisation as a short-term inevitability. Yet, as we formally prove herein, creating systems with human(-like or -level) cognition is intrinsically computationally intractable. This means that any factual AI systems created in the short-run are at best decoys. When we think these systems capture something deep about ourselves and our thinking, we induce distorted and impoverished images of ourselves and our cognition. In other words, AI in current practice is deteriorating our theoretical understanding of cognition rather than advancing and enhancing it. The situation could be remediated by releasing the grip of the currently dominant view on AI and by returning to the idea of AI as a theoretical tool for cognitive science. In reclaiming this older idea of AI, however, it is important not to repeat conceptual mistakes of the past (and present) that brought us to where we are today.

A Scholar's Year in Review: Navigating the Convergence of AI, Economics, and Physics in 2023

“…The argument centers on the nature of both our communication languages and our internal language of thought. If we view such languages as carrying only form, rather than function and meaning, then it can be argued that transformer models cannot act as more than a poor facsimile [16]. However, I argue that both communication languages and internal languages of thought carry both form and function/meaning.…”

What Makes a Good Theory, and How Do We Make a Theory Good?

“…promoting dehumanising and contorted self-images, which have only recently been discussed by practitioners themselves from the inside of these disciplines (viz. Birhane and Guest, 2021;van Rooij et al, 2023;Spanton and Guest, 2022;Forbes et al, 2022;van der Gun and Guest, 2023;Erscoi et al, 2023;Barlas et al, 2021). Such an emphasis is not only scientifically advantageous, but also an ethically unavoidable facet of our work.…”

Specifying precision in visual-orthographic prediction error representations for a better understanding of efficient reading

“…The high flexibility of DNN architectures often prevents concrete insights into the processes underlying its decisions, thus creating a new researchable artifact (Ma & Peters, 2020). This characteristic stands in steep contrast to the transparency of the implementations of the oPE models described above (van Rooij et al, 2023).…”