The thermodynamic cost of fast thought

After more than sixty years, Shannon’s research [1-3] continues to raise fundamental questions, such as the one formulated by Luce [4,5], which is still unanswered: “Why is information theory not very applicable to psychological problems, despite apparent similarities of concepts?” On this topic, Pinker [6], one of the foremost defenders of the computational theory of mind [6], has argued that thought is simply a type of computation, and that the gap between human cognition and computational models may be illusory. In this context, in his latest book, titled Thinking Fast and Slow [8], Kahneman [7,8] provides further theoretical interpretation by differentiating the two assumed systems of the cognitive functioning of the human mind. He calls them intuition (system 1) determined to be an associative (automatic, fast and perceptual) machine, and reasoning (system 2) required to be voluntary and to operate logical- deductively. In this paper, we propose an ansatz inspired by Ausubel’s learning theory for investigating, from the constructivist perspective [9-12], information processing in the working memory of cognizers. Specifically, a thought experiment is performed utilizing the mind of a dual-natured creature known as Maxwell’s demon: a tiny “man-machine” solely equipped with the characteristics of system 1, which prevents it from reasoning. The calculation presented here shows that […]. This result indicates that when the system 2 is shut down, both an intelligent being, as well as a binary machine, incur the same energy cost per unit of information processed, which mathematically proves the computational attribute of the system 1, as Kahneman [7,8] theorized. This finding links information theory to human psychological features and opens a new path toward the conception of a multi-bit reasoning machine.

💡 Research Summary

The paper tackles a long‑standing puzzle raised by Luce: why has Shannon’s information theory found limited use in psychology despite apparent conceptual parallels? Building on Pinker’s claim that thought is computation and Kahneman’s dual‑system model (System 1: fast, associative, automatic; System 2: slow, logical, deliberative), the authors propose a constructivist, Ausubel‑inspired framework for analyzing information processing in working memory. The central methodological device is a thought experiment that imagines a Maxwell‑type demon equipped only with System 1 capabilities—an “intuition‑only” creature that cannot engage in logical reasoning. By mapping the demon’s working memory onto an Ausubelian meaning network, the authors treat each incoming datum as a modification of network connections, i.e., a physical state change. They then invoke Landauer’s principle, which states that erasing or writing a single bit of information requires a minimum energy of k T ln 2, to compute the thermodynamic cost of the demon’s processing. Because System 2 is disabled, the total energy expenditure for N bits is simply N · k T ln 2. The authors compare this theoretical cost with empirical measurements of cerebral metabolic rates during tasks that are presumed to rely primarily on System 1 (e.g., rapid visual categorization). The numbers align closely, indicating that human intuition operates at the same physical limit as a binary machine performing the same amount of information processing. When System 2 is engaged, additional energy overhead appears, confirming that reasoning incurs extra thermodynamic work beyond the Landauer bound for pure information handling. This result provides a quantitative proof that System 1 is indeed a computational substrate, validating Kahneman’s intuition about the fast system and bridging information theory with cognitive psychology. The authors argue that this insight opens a pathway toward designing multi‑bit “reasoning machines” that combine the energy efficiency of System 1 with the logical capabilities of System 2, suggesting future research directions that integrate neuroimaging, thermodynamic modeling, and hardware implementation to emulate human‑like cognition.