In 1906, John Philip Sousa (the "March King" and perhaps the most famous American musician of his era) penned a scathing polemic titled "The Menace of Mechanical Music." In it, he warned that the rise of the phonograph and the player piano would lead to a "marked deterioration in American music and musical taste." He envisioned a world where the "vocal cord will be eliminated" and children would grow up as "human phonographs" without souls.
Today, as generative AI platforms like Suno, Udio, and ElevenLabs begin to churn out technically proficient songs at the push of a button, the music industry is gripped by a remarkably similar panic. Modern critics describe AI music as "soulless," "slop," and an existential threat to the "human spirit." However, a side-by-side comparison of Sousa’s 1906 grievances and today’s anti-AI arguments suggests that we are witnessing a historical rhyme: a recurring cycle of technological anxiety that has, historically, failed to kill the art it claimed to protect.
Sousa’s primary aesthetic objection was that machines lacked the "soul" of a live performer. He mocked the idea of reducing music to a "mathematical system of megaphones, wheels, cogs, disks, cylinders." To Sousa, music was a living thing that required the "personal touch" of a live human performer.
Modern critics use nearly identical language to describe AI. Contemporary artists argue that AI music is "mathematically derived" and lacks "lived experience." Just as Sousa claimed a machine could never replicate the "throbbing, pulsing" heart of a singer, today’s skeptics argue that an algorithm cannot "feel" the blues or "suffer" for its art. Yet, just as the phonograph eventually became a tool for profound human expression (from the blues to hip-hop), AI is already being integrated by artists as a sophisticated instrument.
One of Sousa’s most poignant fears was the death of musical amateurism. He worried that if music could be played by a machine, people would stop learning instruments. He lamented the loss of the "hostess" who could play the piano for her guests, fearing a future of "professionalism" where only a few made music while the masses simply consumed "canned" goods.
This mirrors the modern fear that AI will "democratize music to death." Critics today worry that when anyone can generate a "hit" without learning a chord, the "ennobling discipline" of musical education will vanish. However, history proved Sousa wrong: the phonograph didn't stop people from playing music; it gave them a library of inspiration that led to an explosion of new genres and more people picking up guitars than ever before. AI, similarly, is lowering the barrier to entry for those with "musical vision" but perhaps not "technical virtuosity," potentially expanding the pool of creators rather than shrinking it.
While Sousa’s essay is full of lofty talk about the "soul," his most practical concern was financial. At the time, copyright law did not protect composers from being recorded without permission or payment. He called the manufacturers of these machines "blind to the moral and ethical questions" of taking an artist's work and reproducing it "a thousandfold" without remuneration.
This is the exact frontline of the AI battle today. The lawsuits against AI companies for "training" on copyrighted data without consent are the modern equivalent of Sousa’s testimony before Congress. In 1909, Sousa’s activism helped lead to the "mechanical license," a system that ensured composers got paid for recordings. The lesson here is that the "menace" wasn't the technology itself, but a legal framework that hadn't caught up to it. Once the laws were adjusted, the "menace" became a gold mine for the industry.
Sousa concluded his essay by warning that the "nightingale's song is delightful because the nightingale herself gives it forth." He couldn't imagine a world where a recording could be "art." But a century later, we view 20th-century recordings as some of the highest achievements of human culture.
Today's concerns about AI-generated music are grounded in legitimate questions of labor and law, but the aesthetic doom-saying is a familiar ghost. If history is any guide, "Silicon Music" will follow the path of "Canned Music": it will be resisted as a soulless invader, then adopted as a strange new tool, and finally accepted as just another way for humans to do what they have always done—tell their stories through sound.
The machine might provide the "cogs" or the "code," but the "soul" has always been provided by the listener's response and the creator's intent. Sousa’s marches are still played by human bands today—not despite the phonograph, but because the phonograph made them immortal. AI is likely to do the same for the next generation of "accidental" composers.