Neuro-inspired cognitive architectures for artificial metacognition and social learning

Since artificial intelligence will play a crucial role in our daily lives, it is important that, along with increased task performance, socially intelligent agents will be able to cooperate efficiently with humans. However, human social cognition is a recent evolutionary event that relies on an advanced cognitive architecture. While the unconscious brain, also referred to as system 1, aligns well with the current successful applications of deep learning (statistical invariance and value-based decision-making), the conscious brain, also referred to as system 2, involves higher-order cognitive abilities that perform much more complex computations than what deep learning can currently do (compositionality and metacognition). This project will investigate how AI can benefit from system 2 as well as its relationship with sociality and it will provide new applications of AI for testing player-vs-player games.

Samuele Bolotta
Superviseur universitaire: 
Guillaume Dumas
Partner University: