Will AI kill music making, or could it save creatives?

BY Neal Sawyer

5th Jul 2023 Music

Will AI kill music making, or could it save creatives?

Video killed the radio star—will artificial intelligence finish the job, or could it help musicians in a struggling industry?

What is AI?

AI, short for Artificial Intelligence, is a computer simulation of human intelligence. As Phil Tee, CEO of AIOps company Moogsoft highlights, we already use it in our daily lives. "You have AI on every device you own, from your Apple watch to your smartphone."

Truth is, for many years, the music industry has put AI to use. Ever asked a smart speaker to play a song? Or ever streamed Spotify's new DJ AI, which plays music catered to your exact taste? This is AI at work.

So maybe Terminator 2 was catastrophising AI in true Hollywood fashion. Tee continues: "The truth is very mundane. It has absolutely no self-awareness or intent, it’s no more malicious than the car in your garage. And, at the end of the day, every computer has an off switch!”.

Technology drives musical innovation

Paul McCartney smiling and playing keyboardCredit: Everett Collection Inc / Alamy Stock Photo. Technology and musical innovation have long gone hand in hand, with the invention of the sampled keyboard birthing artists like The Beatles 

So, in the words of an AI expert, there's nothing to fear just yet. What's more, we need to remember that technology in music is nothing new. In fact, the past has looked kindly at the early innovators.

Born out of Birmingham in 1963, the Mellotron was one of the first mass-produced sampled keyboards. The Beatles adopted the revolutionary sound in no time.

A mere four years later, in 1967, the whimsical, dreamy tones were prominent on “Strawberry Fields Forever”, a song that helped to spearhead a psychedelic movement.

"Learn to use the tech or get left behind"

Talking of The Beatles, music writer Rob Bowman said "one of the two most influential bands in the history of postwar popular music" alongside the Fab Four was Kraftwerk—a band who embraced a full electric soundscape, changing the course of pop music forever.

Multi-platinum-selling music Producer TommyD is open to AI and believes it’s the next step in the ongoing relationship between tech and music.

“Electric guitars have been on sale for nearly 100 years, but not everyone is Jimi Hendrix. When the hard disk recording came in, I told all my engineer mates to buy systems or get left behind.

"Those who did are still in the business, those that didn’t are selling insurance. Learn to use the tech or get left behind,” he states.

AI to the rescue

Taylor Swift in gold dress performing musicCredit: Eva Rinaldi, CC BY-SA 2.0, via Wikimedia Commons. Artists like Taylor Swift have been hit by plagiarism claims, which AI could help artists avoid

Is it high time for musicians to consider the benefits of AI? Above all, the latest tech boom could offer a helping hand for time-consuming tasks.

There’s so much more than music-making needed to get your name heard. Crafting social media content to keep an online persona is a job in its own right—not forgetting the creation of music videos, artwork and gig posters thrown onto the to-do list.

With streaming royalty rates too low to supply vital funds, the money isn’t available to outsource the workload. Perhaps it’s time to lean on AI to help.

Truth is, AI can complete these tasks in seconds. This should appeal to musicians who can then concentrate on what's most important: the music.

When probed on the use of AI in music, Tee cited plagiarism as a key area. “ChatGPT is already very good at spotting when two pieces of text are the same or from the same author and it is a short step to do the same in music."

Songwriters are no strangers to lawsuits. Even one of the world's most streamed artists, Taylor Swift, found herself embroiled in a dispute over lyrics at the turn of the year.

Writers Sean Hall and Nate Butler brought the case to court having written the 3LW song “Playas Gon' Play” containing the lyrics "Playas, they gon' play, And haters, they gonna hate." Sound familiar?

Swift denied lifting lyrics for her smash hit "Shake It Off". That could well be the case. It’s impossible for a songwriter to know the lyrics of every song ever written.

With that in mind, there's a chance of imitating phrases unknowingly when scribbling ideas in the studio.

A plagiarism-detecting tool could become a vital part of a songwriter's arsenal. It would make sure new tracks are original, plus prevent legal fees and stress from future lawsuits. If the tool could alert plagiarised melodies, songwriters would embrace AI.

Overstepping the lines?

Speaking of lyrics, they're resistant to generative AI, right? Well, it turns out not. AI can write lyrics. We prompted ChatGPT to "write me song lyrics about lost love in the style of Leonard Cohen."

Reams of original lyrics rolled down the screen in a matter of seconds. Verses, choruses, an outro, the whole nine yards. Evidently, AI can write lyrics. Moreover, write lyrics in a similar vein to one of the world's favourite songwriters.

Here’s a sample of the results…

"I met her in the autumn
With leaves of gold and red
Her eyes were like the ocean
Her hair, a crown on her head
I try to fill the void
With cheap whiskey and cigarettes
But nothing can replace
The love I can't forget"

Songwriters use words to release their own emotions. When heartbroken, they’ll write about their vulnerabilities and pain. Once on record or played live, it’s an insightful glimpse into their soul.

"Songwriting is spiritual. You put pen to paper and write what the world tells you"

The listener who shares the same sentiments takes solace, and that’s one of the many beauties of music. The purity of human connection.

But if AI writes the lyrics, then surely that’s at risk. Furthermore, as AI becomes wiser, it'll become harder to distinguish words that come from human emotion or not.

It's a concern that resonates with Leon Luis, owner of AltarBoy Music Publishing.

“It’s a loss of humanity. Songwriting is spiritual. You put pen to paper and write what the world tells you. You need heartbreak before you can write a song about lost love. You can’t substitute human experience with code.”

The rise of the deep fake

David Guetta DJing onstageCredit: David Jensen / Alamy Stock Photo. David Guetta has embraced AI as a collaborator onstage

Artificial lyrics are the tip of the iceberg. In February this year, The (Not So) Real Slim Shady featured on a song with DJ David Guetta in front of thousands of ravers. As punters danced and rejoiced at hearing the new collaboration, most were unaware it wasn't Eminem at all.

Having asked AI to write a verse in the style of Eminem, Guetta then put those lyrics into another generative AI program to recreate the voice. The result was a close likeness, indistinguishable to most.

Now, the floodgates are open and AI can rap and sing in the style of your favourite artists.

Later, an AI duet featuring Drake and The Weeknd landed on streaming platforms. It soon went viral, rattling the music industry at the same time. Drake, the golden goose of the Universal record label, had been cloned.

Universal responded by banishing the song off Spotify and others. A statement by the label called the song "a breach of our agreements and a violation of copyright law".

It's not only the digitised pop genre that has fallen victim to AI deep fakery. A set of songs by AIsis, an AI Oasis, complete with fake Liam Gallagher, took the internet by storm in late April. It poses an enormous question about copyright.

Luis, who deals with music law, had this to say: "With AI, we're in completely uncharted territory. We're talking Wild West. Like Arizona in the 1880s during the Gold Rush."

AI has stepped into murky, dangerous waters. It’s safe to say, Universal has nailed their colours to their mast when it comes to AI deep fakes.

Their statement continued with defiant wording: “Which side of history [do] all stakeholders in the music ecosystem want to be on? The side of artists, fans and human creative expression, or on the side of deep fakes, fraud and denying artists their due compensation?”.

The death of the songwriter?

When doomsday merchants warned of an AI influx, we thought the creatives would be the last to fall. But the warning signs are there. The sophisticated nature of modern AI has the potential to take the job of songwriters.

Yet, as it stands, even AI has its limitations. Phil Tee concludes, "For all the hype surrounding ChatGPT and AI, what it is incapable of is novelty. Having the sum total of all knowledge and an excellent pattern repetition algorithm just condemns you to repeat the past.

"We may have a more efficient production process for music…but making the next Bob Dylan or predicting the 'Whispering Grass' hit is a long stretch."

"For all the hype surrounding ChatGPT and AI, what it is incapable of is novelty"

With no law to abide by, the next few months will determine the future of AI in music. Questions loom over the intellectual property of AI-written lyrics while, for all intents and purposes, deep fake AI tracks appear to be on the wrong side of the law.

Moves are being made to protect the songwriter. The Human Artistry Campaign is a petition calling for the responsible use of AI in music. It has the support of powerful organisations in music and sets out seven principles where AI and human creativity can live side by side.

Because, let's face it, while AI has its place in the music industry, in the wrong hands, it can do damage.

I guess we need to hear from one more "person" in this conversation. Over to you ChatGPT…

"No, AI will not destroy music making. While AI technology has advanced significantly in recent years, it is still not capable of replicating the full range of human creativity."

Maybe it's not quite “hasta la vista” for the songwriter after all.

Keep up with the top stories from Reader's Digest by subscribing to our weekly newsletter

This post contains affiliate links, so we may earn a small commission when you make a purchase through links on our site at no additional cost to you. Read our disclaimer

Loading up next...