How our brains perform “social mapping” might teach robots to generate “AI-based social maps”

https://www.cas.cn/cm/202507/t20250730_5078183.shtml

https://www.sciencedirect.com/science/article/pii/S1053811925003696

A team at the CAS Hefei Institutes of Physical Science has discovered that the human brain uses the same neural systems to understand and judge complex social hierarchies as it does to navigate and find directions. It creates an invisible “social map” within the brain to navigate complex interpersonal relationships.

In everyday situations, people can quickly judge the social position of each person using abstract symbolic labels such as titles, ranks, and generations, without having to interact with them in depth. These labels act as mental shortcuts, helping people efficiently understand interpersonal relationships. However, how does the brain “translate” these symbols into an internal, structured network of relationships that guides people’s judgments and decisions?

To address this question, the team designed and conducted a three-day functional magnetic resonance imaging experiment where  participants played the role of “hiring managers” and learned and memorized the social hierarchies of a series of “job applicants,” represented by different faces, along the dimensions of “ability” and “character.” They were then required to make rapid hiring decisions based on this information.

The results provide both behavioral and neural evidence. Behavioral evidence suggests that the greater the social rank gap between two people, the faster and more accurate the participants’ judgments. This suggests that people rely on an internal “map” when making social judgments, rather than recalling isolated pieces of information. Neurally, imaging results pinpointed the map’s creator: the hippocampus-entorhinal cortex system, a part of the brain responsible for spatial memory and navigation. Analysis revealed that the core function of the hippocampus is to encode relative “social distance” between individuals, while the entorhinal cortex goes a step further, encoding this information in a unique hexagonal grid pattern, providing direction and structure to this abstract map, similar to the coordinate grid on a physical map.

The brain operates through a mechanism called “neural reuse,” it utilizes its spatial navigation system to “translate” simple linguistic labels into a rich internal social map, supporting complex social reasoning and decision-making.

This discovery offers a new perspective for the current development of artificial intelligence. Researchers can build a dedicated “cognitive map” module for AI, enabling it to learn to more efficiently transform abstract symbolic information into structured internal models. This AI system, deeply aligned with human cognitive patterns, is expected to pave the way for efficient human-machine collaboration and advance the development of AI systems.

Most popular posts: