MBW Views is a series of exclusive articles by some of the biggest names in the music industry.It comes from Andy Chatterley (inset), founder and CEO of MUSO, a London-based technology company that provides anti-piracy services and market analysis to music and entertainment companies.
Interesting how much heat the Ghostwriter track is generating this week.
This Lady Gaga and Lana Del Rey AI mashup has been available online for some time. Artist’s blessing.
I have a few questions:
- How can you be sure that the “Fake Drake” truck is an AI and not a clever marketing tool? There is only the word “ghostwriter” and no evidence of what program was used. The source data is self-explanatory, but that’s all we know.
- If this is indeed AI, it leads to some very interesting philosophical and, inevitably, legal arguments. should it be?
The problem, of course, is that millions of individual musical splices are already being used to hone AI music modeling.
As a creator, how do you prove that your work is being used as source material for AI, unless it is overtly or self-disclosed for marketing purposes, as is the case with ghostwriter tracks?
Regarding the possibility that AI-generated music is legally infringing on human copyrights: As far as I know, currently a similar real human being produced in another artist’s style is completely different. There is no law prohibiting the creation of new tracks.
Indeed, the industry has been booming for some time, with covers that are nearly identical to current hits being used in shops, restaurants, etc., but with lower licensing costs.
The music industry is fundamentally a business built on influence.
A modern example: Celeste sounds strangely (and wonderfully) like a cross between Billie Holiday and Nina Simone, and even samples Simone on her hits. stop this flameWhat’s the difference between Celeste’s work gaining influence and AI literally doing the same thing?
Furthermore, who owns the AI in a given case, and how do you sue someone with no name, no social security number, no company number?
Do you want to sue the prompt engineer who typed in the command and made the track? Should every human being have some sort of guardianship for every computer they own?
Either something is artificial or it is not.
There may be gray areas here that a competent lawyer can address. It’s a matter of intent. Can you claim that something was created via AI with the express intention of depriving the “source” artist of a legitimate income?
At MUSO, we frequently deal with the “artist hijacking” situation in today’s music business. In this situation, it’s clear that the bad guys are trying to impersonate the official artists and divert the streams and revenues of the digital service.
Fake advertising (“hijacking”) of artist names on DSP accounts, whether by AI or humans, constitutes an infringement and/or trademark issue.
Then there is the current trend of remixers speeding up or slowing down famous tracks and uploading them online as new copyrights (especially on TikTok).
Again, these are simple violations of copyright and there are quick ways to fix them.
The AI problem I see now is a familiar story in the music industry. The stable door is closed long after the horse is gone.
Trying to ban AI services from accessing music on DSPs for modeling will not address the wide availability of music outside of major platforms. Not to mention how AI works in practice.
The millions of records, songs, symphonies, and more that have already been fed into data-hungry AI models are not publicly identified. This process is shrouded in mystery, fully understood only by behind-the-scenes programmers.
“No one can compile an exhaustive list of artists or the things that influence their art. It’s just that it’s from.”
I don’t know who’s work was used and how much.We probably never know (even if we all somehow that is this data).
But how is this AI replication process different from how human creators produce their work?
No one can compile an exhaustive list of artists and what influences their art. All we know is that something is really connected because the sum is greater than its parts.
Therefore, AI duplication in music is not a Cut and Dry copyright issue. Rather, it is a new form of interpolation. But just like sampling, it may eventually require some form of rights holder clearance.
One might argue that AI tools have democratized music, making weekday drafting tools for track creation accessible to more people than ever before at a fraction of the cost.
But it can also be argued that this leads to a crisis of homogeneity in the creative arts.
(Some say we are already there, with or without artificial intelligence.)
Things are moving at an unprecedented pace in the world of AI.
Still, anyone who’s seen it will tell you ABBA Voyage is an extraordinary live experience, but they usually add, “I can’t believe it’s all done in Avatar.”
Meanwhile, Taylor Swift (who, as far as I know, is still human) is selling a ton of shows in record time.
Humans will still be humans.
Younger generations may be more inclined to the relationship between AI and music as technology improves.
As an anecdote, my 12-year-old hip-hop-loving son listened to Ghostwriter’s track and said, “The beat was great, but the compression was off. Drake’s verse was hard. So Metro Boomin’s producer said the tag was 2016 old Life of Pablo”.
At 12, that means pretty good, but not incredible.
I asked him if he thought AI would kill his productivity and creativity. Whether people care about the drudgery of making music.
He was adamant that the opposite would happen – the fun and magic of making music was in tinkering with sounds and beats and bringing things to life. Hopefully AI will facilitate this further.
“Can I copyright the sound of my voice? Or condition of voice? How do we legally define this difference? ”
No one can predict the future, but don’t panic.
AI is still in its infancy, and how we use it will fundamentally affect how it uses us. From a production standpoint, it’s very similar to 1980s MIDI and sampling, but with the potential to do much more.
From a copyright perspective, labels should embrace AI and be proactive rather than reactive. Instead of trying to kill the AI, we have to find a way for the AI’s music to coexist with the existing catalog.
We expect to see many more acquisitions in the future, with major music companies buying AI companies to take control of their technology, much like they invested in Spotify.
But it is questionable whether the music industry can actually monopolize and dominate music AI modeling companies.
Ultimately, AI will exist in some form in the same way that Auto-Tune is ubiquitous in modern music.
In many ways this is a matter of metadata and can be enforced as such.Can voice sounds be copyrighted? or copyright condition of voice? How do you define the difference legally?
Right now, someone might be typing in ChatGPT like this: global music business On the impact of AI on the music industry. ”
how do i know How can you know that what you are reading right now is not what it is?
The music industry has found a way to monetize Crazy Frog. They figure out how to monetize AI.
The question is how long it will take them to catch up with technology and how much money will they leave on the table.global music business
