Explanations and sensemaking with AI and HCI

Abstract

The rapid growth of available data and computational power have brought about the wide adoption of algorithmic solutions, especially using Artificial Intelligence and Machine Learning techniques. Thus, users routinely interact with AI-based systems, even though they might not be aware of it: tasks such as searching, diagnostics, credit analysis, and movie recommendations now employ AI techniques. These techniques analyze large volumes of data, iteratively processing and combining them in multiple ways. This often makes both the techniques and the results hard to understand, despite the (frequently) good results.