Brain HealthSingle Human Brain Cells Compute Like Deep Neural Networks Scientists Discover
A new study from Hebrew University published in PNAS found that individual human cortical neurons are dramatically more powerful computational units than neurons in other mammals. Using AI-based modeling, researchers measured how complex an artificial neural network would need to be to replicate a single biological neuron's behavior. The more complex the required AI twin, the more powerful the biological neuron. Human neurons, with their richly branching dendritic trees and unique electrical properties, could perform computations rivaling entire deep neural networks. This challenges the idea that human intelligence stems purely from brain size and connectivity, suggesting the processing power of individual cells plays a critical role. For longevity researchers, understanding neuronal computation capacity has direct implications for brain health preservation and cognitive resilience as we age.
