How AI evolved from quest for a mathematical theory of the mind

Fast Company Tech

by Next Big Idea Club

February 26, 2026

AI-Generated Deep Dive Summary

The quest for a mathematical theory of the mind has shaped the evolution of artificial intelligence (AI) over centuries, as highlighted by Tom Griffiths in his book *The Laws of Thought*. This journey began with Enlightenment thinkers like René Descartes and Gottfried Wilhelm Leibniz, who sought to apply mathematics not just to the physical world but also to the mental realm. Their ideas laid the groundwork for mathematical logic and early cognitive science, ultimately influencing modern AI development. Griffiths emphasizes that AI’s story extends far beyond recent breakthroughs in neural networks, tracing back to psychological theories about human cognition. The article explores how different mathematical approaches have contributed to understanding the mind. Initially, mathematical logic was used to describe thought, but its limitations became apparent when dealing with fuzzy concepts and complex relationships. This led to the development of artificial neural networks, which gained traction after psychologists demonstrated their potential for learning intricate patterns. However, challenges in scaling these networks emerged, highlighting the need for probabilistic approaches to better model human-like learning. Griffiths also underscores the importance of interdisciplinary collaboration in AI research. Unpopular ideas in one field often find new life in another, driving innovation. For instance, neural networks declined in popularity among computer scientists but were later revived by psychologists and machine learning researchers. This back-and-forth between disciplines exemplifies how diverse perspectives can lead to breakthroughs, ultimately bringing us closer to replicating human intelligence in AI. Despite significant progress, modern AI still falls short in key areas like learning efficiency. While humans master language and complex tasks with minimal exposure, AI systems require vast amounts of data and computational power. This gap underscores the ongoing need for research into how human brains differ from neural networks and what makes human cognition unique. Griffiths suggests that studying these differences could not only improve AI but also deepen our understanding of ourselves. For designers and researchers interested in AI’s potential,

Verticals

designtech

Originally published on Fast Company Tech on 2/26/2026