Can AI Write Better Songs Than Taylor Swift?
A new study by Stephen Arnold Music (SAM) and SoundOut, conducted with thousands of respondents across the US, compares music from AI and Taylor Swift in addressing just how far AI generated music has come in the pop genre…and how far it will go. Can AI write better songs than today’s biggest pop stars? The results are surprising and suggest intriguing creative opportunities that extend beyond the world of popular music.
As in many other fields, generative AI is having a massive impact on the music industry. It’s changed the way labels manage their catalogs and how streaming services recommend songs. It’s provided record producers with new tools for mixing and mastering. Controversially, AI is also being used to create music from scratch.
How far will it go? Will AI be used by composers as a source of inspiration, or will it eliminate their jobs? Will AI generated music be cold and soulless, or will it move us to tears? More to the point, “Can AI write better songs than Taylor Swift?”
Earlier this year, Stephen Arnold Music (SAM), the world leader in sonic branding, and SoundOut, a leading sonic testing agency, conducted an experiment pitting AI music generating software against a professional “human” composer in responding to a specific music brief. The results were fascinating, and telling. While AI proved very good at producing music that was subjectively appealing, it often missed the mark for emotional accuracy — producing a specified emotional impact — something that human composers are very good at.
Now, the two companies have completed a follow up experiment testing AI’s ability in the pop arena. Their purpose was to see if generative AI could create songs as heartfelt, catchy and memorable as the world’s biggest pop star — none other than Taylor Swift herself. You’ll have to read the study itself to get the complete results but suffice to say emotional connection was once again the differentiating factor.
So, no, AI bots won’t be playing sold-out stadium shows any time soon, but that doesn’t mean that the technology won’t have a significant impact on popular music. AI has proven that it can produce music that is broadly engaging, even if machines lack the charisma of a star performer. “Consumers seem willing to embrace AI music if the song is sufficiently appealing, even if it lacks an artist connection,” says SoundOut CEO David Courtier-Dutton. “It’s not that the artist is unimportant, just that the song is much more important.”
The implications of the study go beyond pop music. The strengths and weaknesses of AI apply to all forms of music and sonic media. AI is an intriguing tool with great potential, but it’s not a substitute for human creativity and emotional insight. “With sonic branding, our composers combine analysis, artistry and subjectivity to get the right musical ‘feel,’” explains SAM president and creative director Chad Cook. “We’re not magicians, but the ability to relate to emotions is an intrinsic part of the composing process. AI is a great inspirational tool and can produce music we like, but it has not demonstrated the emotional complexity to dependably produce impactful brand music.”
The takeaway from this fascinating study is that humans and AI aren’t adversaries, they’re collaborators. “AI is here to stay,” the study concludes. “Musicians, music agencies, and leading brands are best served by exploring how to embrace it.”
