Human Brain Cells Are More Powerful Than Scientists Thought, Study Finds

How does the human brain produce abilities such as language, imagination, mathematics, and invention?

For decades, scientists have primarily focused on brain size and connectivity to answer this question. The human brain contains nearly 100 billion neurons connected through an extensive network. However, new research indicates that an important part of human cognitive ability may exist at a much smaller scale: the remarkable computational power of individual brain cells.

Human neurons are highly sophisticated information processors

Researchers found that neurons in the human cerebral cortex function as considerably more complex information-processing devices than comparable neurons in other mammals. These findings suggest that individual components of the human brain may be unusually powerful, potentially helping explain how humans developed advanced cognitive abilities.

The study was led by researchers from the Hebrew University, including Idan Segev and Mickey London of the Edmund and Lily Safra Center for Brain Sciences (ELSC), along with doctoral students Ido Eisenbad and Daniela Yoeli. The research team also collaborated with Professor Chris de Kock of Vrije Universiteit Amsterdam.

“People tend to think of a neuron as a simple switch that turns on or off,” Segev says. “What we are showing is that a single human neuron is itself a highly sophisticated computing device.”

How scientists measured the computational power of neurons

To determine how many calculations an individual neuron can perform, the researchers developed a new method for measuring the computational complexity of brain cells. They combined advanced computer modeling with artificial intelligence to assess how difficult it was for artificial neural networks (ANNs) to learn and reproduce the relationship between a neuron’s inputs and its responses.

The underlying principle is straightforward: if an artificial “twin” requires a more complex model to accurately imitate a biological neuron, the biological cell itself is likely capable of more sophisticated computations.

The results highlighted the computational advantages of neurons in the human cortex. Their extensively branched dendritic structures and distinctive electrical properties allow them to process incoming information in complex ways, including visual signals such as those used to distinguish between images of cats and dogs.

In other words, individual human cortical neurons are far more than basic on-and-off components. Each cell can perform computations comparable to those carried out by a deep neural network, effectively operating as an advanced processing unit within the brain.

A new perspective on human intelligence

The discovery challenges the long-standing view that intelligence depends mainly on the brain’s enormous number of neurons and its extensive network of connections. Instead, the study suggests that the computational complexity of individual neurons may also have played an important role in the evolution of human cognition.

The researchers also created a systematic framework for linking the physical structure of brain cells with the computations they perform. This approach could help scientists better understand how the human brain generates thought, learning, memory, and other forms of cognition.

Could complex human neurons inspire new AI systems?

The findings may also influence the future development of artificial intelligence. Most modern machine-learning systems rely on highly simplified artificial units. The research points toward another possibility: brain-inspired AI built from artificial units with greater computational power, more closely reflecting the complexity and capabilities of biological neurons.