We’ve talked a lot about the intrusion of artificial intelligence into almost every sector of our society and industry. And we will continue to talk about it in the future because this is not a passing trend but an overwhelming and enduring trend.
And if yesterday we reported that Spotify is going to use AI to dub podcasts into all languages, something that seems quite positive as it improves accessibility, what we bring today is less hopeful for musicians.
Daniel Ek, the CEO of Spotify, told the BBC that there are valid uses of artificial intelligence to create music. Yes, AI should not be used to replace human artists without their consent, but the debate about the technology will likely continue for many years, Ek added.
AI won’t be banned on Spotify, but there’s no carte blanche either
Ek, who rarely speaks to the media, said that the use of AI could be acceptable for auto-tuning, for example.
However, a much more complicated middle ground would be AI-created music, clearly influenced by existing artists but not directly impersonating them. Spotify has already banned its content from being used to train AI models.
“You can imagine someone uploading a song and claiming it’s Madonna, even if it’s not. We’ve seen pretty much everything in Spotify’s history at this point with people trying to game our system,” Ek told the BBC. “We have a very large team working exactly on these kinds of issues.”
As we’ve discussed before, there is growing fear that self-improving machines may eventually replace human composers and producers in the electronic music industry.
However, the creators of AI-generated music applications have insisted that “generative music” will actually lower the entry bar in the industry.
For example, in early August, Meta released an open-source music generator called MusicGen, which has been trained on 20,000 hours of licensed music and produces short samples based on textual and melodic prompts.
The current use of AI on Spotify
AI helps Spotify personalize the user experience by recommending artists, playlists, and podcasts to keep users engaged and more likely to subscribe long-term.
It also helps listeners worldwide enjoy their favorite podcasts in their native language.
However, as with anything good, there’s a downside. Machine learning technology and its efficiency may have been one of the reasons behind Spotify’s decision to reduce its workforce by 6%, or around 600 people, in early 2023.