Study Reveals Human Neurons Perform More Complex Computations Than Those of Other Mammals

A new study published in the Proceedings of the National Academy of Sciences suggests that individual human neurons are capable of performing far more complex computations compared to neurons in other mammals. This discovery may provide insight into the unique cognitive abilities of the human brain.

For many years, researchers have attributed human intelligence primarily to the sheer size and connectivity of the brain, which contains approximately 100 billion neurons interconnected through an extensive network. However, recent findings indicate that the processing capabilities of individual neurons also play a crucial role.

The research, led by Profs. Idan Segev and Mickey London of Hebrew University, alongside PhD students Ido Aizenbud and Daniela Yoeli from the Edmond and Lily Safra Center for Brain Sciences (ELSC), highlights the advanced functions of human cortex neurons, which researchers describe as significantly more complex information processing units, akin to “microchips.” They worked in collaboration with Prof. Chris de Kock from the Free University in Amsterdam.

“People often think of a neuron as a simple switch that either turns on or off,” remarked Segev. “What we show is that a single human neuron is itself an extraordinarily sophisticated computing device.”

Measuring Neuronal Computational Power

To explore the computational power of individual neurons, the researchers developed a novel methodology combining advanced computer modeling and artificial intelligence. This approach tested how complex it would be for an artificial neural network (ANN) to replicate the relationship between incoming information and the responses generated by human neurons.

The results showed that human cortical neurons possess a distinct advantage, characterized by their richly branching dendritic trees and unique electrical properties. These features allow them to execute sophisticated computations on various forms of sensory input, such as differentiating between images of cats and dogs.

Each human cortical neuron is much more than merely an “on-off” switch; these cells function as intricate computing units, demonstrating computational abilities that rival those of deep neural networks.

Implications for Understanding Intelligence

This research challenges the conventional notion that cognitive abilities are primarily a byproduct of the number of neurons and the connections among them. It posits that the intricate complexity of individual neurons is also a significant factor in the evolution of human intelligence.

Furthermore, the researchers have established a systematic framework linking the physical structure of brain cells to their computational capabilities. This could pave the way for enhanced understanding of human thought, learning, and cognition.

Influence on Artificial Intelligence

The findings hold potential implications for the future of artificial intelligence. Current leading machine learning systems are developed using simplified artificial units. However, this new research points to the possibility of creating brain-inspired AI that employs artificial units with deep computational power, more closely mirroring the capabilities of biological neurons.