An international research team announced on Apr. 13 that the human brain relies on both cooperation and competition among its circuits to function, according to a study published in Nature Neuroscience. The study involved researchers from the University of Oxford, the University of Cambridge, Pompeu Fabra University, and the Montreal Neurological Institute.
The findings highlight that intelligent behavior emerges from a balance between specialized regions working together internally and competing with each other over long distances within the brain. This interplay is seen not only in humans but also in macaques and mice, suggesting it is a fundamental feature of mammalian brains.
Using advanced computer modeling based on whole-brain data, the researchers demonstrated that models incorporating competitive interactions outperformed those relying solely on cooperation. These models better reflected real cognitive processes such as attention and memory by allowing different networks to take turns influencing overall brain activity. "Competition between circuits allows certain networks to take priority over others depending on what is relevant at any given moment, which explains phenomena such as decision-making," said Gustavo Deco, ICREA research professor at Pompeu Fabra University and one of the study's senior authors.
The analysis included more than 14,000 neuroimaging studies. The results showed that excess cooperation could lead to unrealistic synchronization within the brain's activity patterns while competition acted as a stabilizing force. Dr Andrea Luppi of the University of Oxford explained that this approach enables creation of digital twins—realistic digital copies—of individual brains: "a realistic digital twin of a given brain: one that matches your brain better than any other brain".
Deco added that these new models provide improved information for predicting diseases and symptoms compared to traditional measures. Luppi said these models could simulate an individual's response to treatments or disease progression, potentially enabling therapies tailored specifically for each person.
Researchers concluded that this cooperative-competitive architecture might reflect broader principles underlying intelligent systems beyond biology. They found it also enhances computational capacity in neuromorphic computing (brain-inspired artificial intelligence), supporting more effective processing and integration of information.
