The neuroeconomics of epistemic curiosity

Abstract

Epistemic curiosity is an intrinsic drive to obtain non-instrumental information. It is presumed to be adaptively useful for building mental representations where there is an exploration bonus for learning. Formally, epistemic curiosity is most commonly modelled by calculating the entropy associated with information. Human curiosity across life stages exhibits a predictable Goldilocks pattern, where maximum entropy attracts attention and arousal, thus 'piquing' curiosity. This and close variants of the curiosity 'formula' have been used to generate more efficient algorithms, clickbait headlines, and maximum-suspense thrillers. Human fMRI studies have shown activity associated with curiosity and 'fulfillment' of curiosity (e.g. learning answers), typically in basal ganglia reward areas. Understanding this near-universal human drive is important for education and many other domains.

Additional details