David Guetta makes potent anti-A.I. case by creating a song with an Eminem deepfake
It’s weird to compose songs using A.I. generations of other artists without their permission! Stop it!
As the internet gets flooded with A.I.-generated content created by tools like ChatGPT and DALL-E, the conversation around them can feel like a debate over the precise path of huge boulder that’s rolling downhill toward us. Questions of personal responsibility among the powerful seem futile. In creative fields, it can feel naive to rely on major artists — those with the ubiquity and quality control of Kentucky Fried Chicken — to use these tools responsibly. If you need a case study in how a lack of restraint with A.I. can lead to upsetting results, French mega DJ David Guetta helpfully provided one yesterday.
In the clip, shared on Wednesday (February 8), Guetta gushes over a new song he made with the help of A.I. websites that are able to produce content in the style of Eminem. "I discovered those websites that are about A.I," Guetta says. “Basically, you can write lyrics in the style of any artist you like. So I typed, ‘Write a verse in the style of Eminem about Future Rave [a subgenre of electronic music coined by Guetta and Morten Breum].’” Guetta then used a different website to generate a facsimile of Eminem’s voice, the final A.I. element in the song, played at the beginning of his video during a live set for thousands of fans. "People went nuts," Guetta says of the crowd's reaction, which is unfortunately accurate.
It doesn’t take an unusually empathetic person to wonder if Eminem doesn’t feel a little bit violated by this. In an age when anyone with an internet browser can create convincing duplicates of art, it’s imperative to consider the ethical and legal responsibilities we have both to the artists who are subject to the copying, and the millions of others whose online creations are vacuumed up without permission by the A.I. to train itself and make its products more convincing.
It’s shocking how much further along the technology is from just two years ago. In February of 2020, the internet caught a glimpse of the future with “Travisbott,” an A.I. trained on rapper Travis Scott's musical output without his knowledge or permission to create a shoddy soundalike. That same month I spoke about Travisbott with the Berlin-based composer Holly Herndon. She’s used A.I. in the creation of her own art, including 2019's Proto, which was made with Spawn, an A.I. "baby" with a distinct and limited musical ability. Set on creating tools to empower artists for a future that the Guetta video suggests is already here, Herndon introduced Holly + in 2020, a meticulously programmed and highly advanced deepfake of Herndon's own voice.
Herndon was one of the first people to zero in on Travisbott’s ethical quagmire. "I think the headline is that we have now reached a point where we feel entitled to sample," she said. “Whatever the logical framework that sampling began a hundred years ago or whatever, it has reached a logical conclusion that we now feel entitled to sample human beings and their likeness without anyone questioning that."
That entitlement is seeping through Guetta's video. It doesn't matter that he won't release it commercially — he stands to gain from its creation whether it's through fawning publicity or crowds at his shows enjoying a song that they may or may believe is a genuine Eminem collaboration. Even if Eminem did approve of Guetta doing this — and there’s a very small chance that he did — videos like this put us one step closer to a future where artist autonomy and respect are secondary to cheap thrills and self-publicity. And that’s definitely one of the worst of all possible boulder paths.