Can AI Write Better Songs Than Taylor Swift?

The Experiment: AI Meets Pop Music

To answer this, SoundOut, a sonic testing agency, reverse-engineered five of Taylor Swift’s most iconic hits into creative briefs. These were then fed into the AI platform Suno, which generated new songs based on the prompts. The AI-generated tracks were further refined using Jammable to emulate Taylor’s vocal style. Tracks from Suno, Jammable, and Taylor were blind-tested with 200 listeners to evaluate their appeal and emotional resonance.

Music Appeal: AI Holds Its Own

Surprisingly, two AI-generated songs outperformed Taylor’s originals in listener appeal. “Forever Young” scored a 71% appeal rating compared to Taylor’s “So High School” at 38%. Similarly, “Winter’s Embrace” received a 65% appeal rating, beating “Evermore,” which scored 19%. However, Taylor’s other three songs, including “Anti-Hero” and “Cruel Summer,” outperformed their AI counterparts by an average of 37%.

These results sparked a broader debate: Do listeners value the artist or the song more? A survey revealed that 54% of listeners prioritize the music itself over the performer. This highlights AI’s potential to compete with even the most celebrated songwriters if it delivers compelling tracks.

Emotional Accuracy: A Gap in AI’s Arsenal

While AI demonstrated its ability to craft appealing songs, it faltered in emotional complexity. Taylor Swift’s songs consistently conveyed layered emotions, scoring above 80% in SoundOut’s emotional accuracy metrics. In contrast, AI struggled to meet these benchmarks, with its highest emotional accuracy score being 38% for the track “Broken But Dancing.” Most AI-generated tracks scored below 10% in emotional accuracy, lacking the depth and nuance required to truly connect with listeners.

The AI compositions primarily reflected simplistic emotions like “pure” or “peaceful,” often unrelated to the intended brief. This gap underscores the limitations of current AI in capturing the emotional complexity that defines Taylor’s songwriting.

Implications for the Music Industry

AI’s ability to generate hit songs raises questions about the future of music creation. While AI can produce catchy tunes, emotional authenticity remains a human forte. Industry experts believe AI is best suited as a tool for inspiration rather than a replacement for artists. Chad Cook, President of SAM Creative, states, “AI can generate music, but it lacks the emotional depth and complexity that only human experience can provide”.

For brands, AI offers potential cost savings and efficiency in creating sonic branding. However, the emotional inaccuracy of AI-generated music could hinder its effectiveness in campaigns that rely on specific emotional resonance.

Can AI Top Taylor Swift?

AI has proven capable of crafting appealing music, even outperforming Taylor Swift in some cases. Yet, when it comes to emotional complexity and storytelling, AI still falls short. Taylor’s songs resonate because they weave intricate narratives with genuine emotions—qualities that AI has yet to master.

As generative AI continues to evolve, it may soon augment human creativity, serving as a collaborative partner for musicians. For now, the answer to whether AI can write better songs than Taylor Swift is clear: not yet, but it’s getting closer.