AI and Citizen Science for Serendipity

TL;DR

This paper explores how AI integrated with citizen science can foster serendipity and creative discovery through system design that emphasizes task environment, participant traits, and handling anomalies.

Contribution

It introduces a framework for designing hybrid citizen science systems that promote serendipity by leveraging complex interactions between AI and human participants.

Findings

Designing for task environment enhances serendipitous discoveries.

Citizen scientist characteristics influence creative exploration.

Managing anomalies and errors can facilitate unexpected insights.

Abstract

It has been argued that introducing AI to creative practices destroys spontaneity, intuition and serendipity. However, the design of systems that leverage complex interactions between citizen scientists (members of the public engaged in research tasks) and computational AI methods have the potential to facilitate creative exploration and chance encounters. Drawing from theories and literature about serendipity and computation, this article points to three interrelated aspects that support the emergence of serendipity in hybrid citizen science systems: the task environment; the characteristics of citizen scientists; and anomalies and errors.