What deep learning can tell us about higher cognitive functions like mindreading?

TL;DR

This paper explores how deep learning can inform our understanding of higher cognitive functions like Theory of Mind, highlighting its contributions and limitations in modeling complex brain computations.

Contribution

It assesses the potential and limitations of deep learning in elucidating higher cognitive functions, emphasizing that scaling current models may not achieve human-level Theory of Mind.

Findings

DL has advanced visual recognition understanding

DL alone is unlikely to replicate human Theory of Mind

Scaling current DL models may not suffice for higher cognition

Abstract

Can deep learning (DL) guide our understanding of computations happening in biological brain? We will first briefly consider how DL has contributed to the research on visual object recognition. In the main part we will assess whether DL could also help us to clarify the computations underlying higher cognitive functions such as Theory of Mind. In addition, we will compare the objectives and learning signals of brains and machines, leading us to conclude that simply scaling up the current DL algorithms will most likely not lead to human level Theory of Mind.