In a songwriting showdown, fans prefer the AI version of Akimoto’s style over a human-written track
What Happened
In a televised competition titled “Akimoto Yasushi vs AI Akimoto Yasushi: AKB48 New Song Production Contest,” Japan’s top girl group AKB48 released a new single composed by artificial intelligence. The AI—based on Google’s Gemini model—was trained on Akimoto’s vast catalog of lyrics and other creative outputs. Its song won more fan votes than the competing human-written track by Yasushi Akimoto himself.
Details of the Contest
The two songs—one by the AI and one by Akimoto—were released for fan voting. The AI-written song reportedly resonated more, showing how machine learning mimicking a well-known style can compete with original human creativity. The human-written song, while praised, did not garner as many votes.
Implications for Music Industry & AI
This result underscores growing experiments in the music business: AI as creator, not just a tool. It raises questions about authorship, copyright, creative authenticity, and how audiences engage with “traditional vs. tech” in art. If fans favor stylistically accurate AI compositions, what does that mean for human creators?
Broader Cultural Context and Reactions
Some listeners feel the AI song captures a polished homage, while others argue it lacks the emotional nuance of a human-written work. Critics have begun urging clear labeling of AI content, ethical AI training practices (especially using past works), and preserving originality. Meanwhile, this could mark the beginning of more contests or releases testing AI vs human creativity directly.
