The AI industry is building products on a flawed model of how the human mind works, according to a growing critique that draws on neuroscience and evolutionary history. The industry's guiding assumption, that the brain is essentially a computer, has been endorsed by figures including Google's Demis Hassabis and Elon Musk. That comparison understates human complexity, and the mismatch helps explain why AI products are disrupting culture as they spread.
The computer model of the mind traces back to Alan Turing and treats thought as a three-step algorithm: input, computation, output. Technologists have used that model productively, advancing computing from adding machines to artificial neural networks and generative AI. But the image it offers is warped and limited. John von Neumann, a central figure in computer science, doubted the model could capture the exceptional complexity of the human nervous system.
Neuroscientist Paul Cisek of the University of Montreal has developed a biological alternative spanning millions of years of evolution. In his account, brains are feedback-control systems rather than information processors: bodies take action to adjust the input they receive, depending on available options. He traces the idea to philosopher John Dewey, who described the mind as an organic circuit in which motor response determines the stimulus as much as sensory stimulus determines movement.
Cisek's example is catching a fly ball. The computational model requires an outfielder to estimate velocity, calculate gravity and perform other hidden math. The feedback-control approach requires only a simple heuristic: keep the ball in the same position in your visual field and move to maintain that. This account also fits the observable architecture of brains as they evolved from ancient fish to amphibians, mammals, primates and modern humans, with new behaviors emerging as environments offered new possibilities.
The computational model does not sync with observable neuroscientific structures and obscures how much brains do to control organisms' interactions with their environments. It also fails to account for human sociality. Our ancestors learned to imitate gestures, pass along practices such as stone toolmaking, and coordinate through shared ritual, then moved to imitating sounds, oral language and settled agricultural communities.
More AI news from TechManNews.
