The Mind as a Market of Minds

Abstract

There is substantial evidence from neuroscience and psychology that what we perceive as a unified mind is actually a vast collection of relatively autonomous minds. A decentralized brain appeals from an evolutionary perspective too, as it requires less interconnected wiring than a fully centralized brain, can more easily evolve through replication and can more easily adjust to damage. The main downside is potentially suboptimal learning relative to a truly unified brain. This paper suggests that the downside is limited. Indeed, under some conditions decentralized learning can match the optimal unified solution. These conditions liken the decentralized mind to a competitive capital market and its neural components to wealth-seeking gamblers. The very possibility lends support to the Thousand Brains theory of Hawkins (2021), by showing how the Thousand might behave like a rational One.

Keywords: decision making, probability, uncertainty, surprise, price formation, quasi-markets, neuroeconomics (D8320), neural networks, capital market, Kelly criterion, gambling

JEL Classification: C450, D870, G100