Thanks to Hanna Wallach for pointed to some articles about autism and mirror neurons .

In the field of agent simulation in a VR world, some AI method give to an agent a representation model of its world where it can simulate approximate behaviors of other agents before to take a decision of its action in its real (virtual) world: simulation of a virtual world inside a virtual world…

But the first time I discovered this idea of “interpreting a model of other” was with Hercule Poirot who try to imitate the person he want to understand, and even some small movements (like the position of hands or the smile…) can help him to “take the place of other”.