Like many people, I was, and probably still am, nervous about the changes AI will bring to musical creativity, but after speaking yesterday with a leading executive with extensive knowledge of the business, law and talent aspects of the music industry, I’m less convinced.
This person made some pretty controversial views about generative AI. They almost scoffed, saying everyone was overreacting, and predicted that it would eventually become something of a ringtone. This is his mid-00s trend of using parts of songs as cell phone ringtones. , for those who don't remember (or don't want to). Ringtones, usually fully licensed by copyright holders, were a lucrative business. Lil Wayne's “Lollipop” still holds the sales record, having sold over 5 million units at a retail price of $2.99 each. But with rare exceptions, ringtones have always been a nuisance.
Of course, the potential for creative uses of AI in music goes far beyond ringtones, but the real question isn't what to do with it. did it Not just anyone can do it, but AI. will do.
For example, let's say it's possible to license Frank Sinatra to rap on a Notorious BIG track or Beyoncé to sing a D'Angelo song. The results may be great, but how many times will you or anyone play it? If it was pressed on vinyl, would you buy it? And even if a few songs like that go viral on TikTok, which is every legendary artist's dream, how long did it take before it became a nuisance?
This does not mean that it is legal or should be acceptable to reproduce someone's voice or music without permission or compensation (except in parody). It's not possible to copyright the human voice at this point, but at least he has two legal actions moving closer to the concept, and lawyers being lawyers, they're finding other ways to stamp out copycats.
Last year, Universal Music reported that a ghostwriter used AI to create the song “Heart on My Sleeve,” about Drake and The Weeknd, by using copyrighted material without the owner's permission. They succeeded in disabling the song by claiming that the AI had been trained to do so. In the 1990s, Tom Waits and Bette Midler won a false advertising lawsuit against a company that used copycat singers in TV ads (after Waits and Midler refused). Advertising is effective, but advertising is ineffective. Commercially released songs. But as a friend of Ringtone Caller said yesterday, it's really just a virtual impersonator or copy band. (Note: I'm talking about music here, not the legitimately dangerous use of AI, such as deepfakes of political leaders exhorting their followers to action or violence. Not that we need deepfakes. there is no.)
AI is already being used for good in the music world. A combination of AI and a similar-sounding singer gave the great country singer Randy Travis a new voice after he lost his singing ability in 2013 after a severe stroke. Such technology can be good. But it won't be long before AI is used for not-so-good purposes. It's only a matter of time (if it's not already the case) before healthy singers simply let AI synthesize their voices for them, rather than actually singing new songs. It's hard to imagine producers and labels saving money by AI-izing multiple voices, instead of paying for backing singers (a Beyoncé with a billion, an Adele with just a few keystrokes), not to mention engineers and other technical staff. The real threat of AI to the music world is in the many jobs it will replace, but the industry's top people rarely make a big fuss about it.
Even before Chat GPT was announced in late 2022 and generative AI entered the public consciousness, we were already well down this path. For years, even bad singers have been made to sound almost good in the studio through the magical effects of autotune. At many major concerts, vocalists sing to their own prerecorded backing tracks. Some people are very good at lip-syncing, and live sound technicians can adjust volume very strategically, so it takes a well-trained eye and ear to detect it. How long will it take until I don't understand it at all? For that matter, how long will it be before ABBA's Abba-tar technology becomes practical enough to make actual tours a useless luxury?
Why stop there? AI-created fictional pop stars already exist, and it's only a matter of time before biopics and alternate history stories become virtual reality. If all of John Lennon's audio recordings were fed into AI, it might be possible to convincingly recreate what a series of albums from 1980 onwards sounded like. Avatars could go on tour with live musicians, similar to hologram concerts. And it won't be long before we can use AI to create Zoom calls with historical figures and deceased loved ones. I feel sick just thinking about it, but will it last forever? If that fake world was nicer than the real world, we might not want to leave it.
Randy Travis' “new” songs were created with his full approval and possible participation. Although he didn't write any songs that feature his AI voice, it's probably already possible for him and other singers to write songs on a keyboard using his AI replica of their own voice. And it's probably only a matter of time before we won't be able to tell the difference.
During the pandemic, many of us expected livestreams to not be Exchange There will be concerts in the future, but to make up for it, if live tickets are sold out, you can purchase virtual tickets and see the show that way. However, with some exceptions, there is little interest in this concept. People want real shows and real performances, even though the meanings of adjectives are becoming more vague every day.
I could wax lyrical about the sanctity of genuine human expression, but you've heard and read it all before and I'll refrain from the obvious, genie-out-of-the-bottle cliches. What is legal and what is illegal is for lawyers to decide, and what is artistically acceptable and unacceptable is up to the individual. AI could cause many dangerous things, but with strong safeguards and laws, the loss of human musical creativity won't be one of them.