Human Evaluation of Interpretability: The Case of AI-Generated Music Knowledge

TL;DR

This paper explores how humans interpret AI-generated music knowledge, proposing an experimental method to evaluate interpretability and uncover challenges in decoding AI-discovered artistic rules.

Contribution

It introduces a novel experimental procedure for assessing human understanding of AI-generated music knowledge, advancing interpretability evaluation methods in arts and humanities.

Findings

Insights into human interpretation of AI-generated music rules

Identification of challenges in decoding AI-discovered knowledge

Proposed methodology for evaluating interpretability

Abstract

Interpretability of machine learning models has gained more and more attention among researchers in the artificial intelligence (AI) and human-computer interaction (HCI) communities. Most existing work focuses on decision making, whereas we consider knowledge discovery. In particular, we focus on evaluating AI-discovered knowledge/rules in the arts and humanities. From a specific scenario, we present an experimental procedure to collect and assess human-generated verbal interpretations of AI-generated music theory/rules rendered as sophisticated symbolic/numeric objects. Our goal is to reveal both the possibilities and the challenges in such a process of decoding expressive messages from AI sources. We treat this as a first step towards 1) better design of AI representations that are human interpretable and 2) a general methodology to evaluate interpretability of AI-discovered knowledge representations.