Artificial intelligence became Beethoven: how do listeners and music professionals perceive artificially composed music?

Abstract

Purpose Artificial intelligence (AI) has reached creative industries such as music. Algorithms now produce high-quality artistic content (e.g. original songs), for hedo consumption and utilitarian business applications. While available literature to-date focuses mainly on technological development and applications, this paper aims to address the resulting research gap by investigating listeners’ perceptions towards music composed by AI. Design/methodology/approach First, an online survey was conducted with 446 respondents and compared perceptions of music professionals (n = 72) and non-professional listeners (n = 374). Following this, a 2 × 2 laboratory experiment was conducted, where 86 participants listened to songs composed by AI but were presented different narratives regarding the composition process (human versus AI). Findings Overall, results from the online survey indicated a rather negative perception, low purchase intention for AI music and a negative credibility perception of musicians using AI. Findings from the experiment indicated no significant differences between the groups, suggesting that the awareness of the use of automation did not influence the perception towards the music. Originality/value This paper contributes by highlighting the current perception of both listeners and music professionals towards the application of artificial creativity in music composition. Furthermore, it contributes to the existing literature on artificial creativity applied in music, by providing evidence of its impact on listeners’ perception. Results reveal the importance of further investigation on the topic.

AI composer bias: Listeners like music less when they think it was composed by an AI.

“…Additionally, similar work found that there was no bias against a known AI composer when the music heard was actually composed by an AI (Tigre Moura & Maw, 2021). That is, this prior study asked participants to rate several attitudes towards musical pieces labeled as human- or AI-composed, but those pieces were actually composed by AIs (vs. our approach here, which was to use all human-composed music).…”

“…In contrast, users may be hesitant about listening to music that was created or composed by an AI. For example, recent survey research suggests that both the general public and music professionals have negative perceptions of AI-created music and express a low likelihood of purchasing music created by AIs (Tigre Moura & Maw, 2021). In this prior work, participants were asked to rate their attitudes towards AI-composed music and their intention to purchase AI-composed music in general, but this study did not consider people’s attitudes toward specific songs purportedly created by AIs.…”

Boundary crossing: an experimental study of individual perceptions toward AIGC

“…Character classification, painting effect, copyright ownership, creation ethics, etc., are the frontier issues of artificial intelligence painting. Previous studies have found that people’s low artist identity for artificial intelligence is due to the lack of emotional motivation and insufficient degree of autonomy ( Tigre Moura and Maw, 2021 : 137). But if we compare the cognitive behavior of human beings and use “S (stimulus)-O (object)-R (response)” to correspond to “input-black box-output,” then artificial intelligence can also establish its own cognition ( Gu and Wang, 2021 :102).…”

Artificial Intelligence, Creativity, and Intentionality: The Need for a Paradigm Shift

“…Surely, the manifestation of human intentionality without the support of AI agents will also continue to exist, and the choice of creative process will continue to be defined by creators (e.g., to write a poem without AI assistance, though artificial co‐creation or to fully automate the process and simply supervise the output). However, AI already generates outputs which are indistinguishable from human made ones (Hadjeres, Pachet, & Nielsen, 2017; Tigre Moura & Maw, 2021). This implies that it is and will become even more challenging to assure the nature of a creative process.…”

“…Also, the ever-increasing collaboration between humans and AI calls for new ethical and legal frameworks. For example, regarding intellectual property (Palace, 2019), authorship (Hristov, 2016), and responsibility for creative outputs (Karliuk, 2018). These issues (simply to name a few) are complex, still largely unresolved, and raises additional questions involving creativity and the authenticity of creative intentionality.…”