Interviewed by Michael Laskow
Let’s talk about the elephant in the room that I’m sure everybody is aware of, which is AI, or Artificial Intelligence.
I recently asked ChatGPT this question: “How can copyright holders and music creators protect themselves from AI Music overtaking the market for original music in media?” And here’s what ChatGPT had to say: “As of my last knowledge update, September 20th, 2021, AI-generated music was certainly gaining attention but has not completely overtaken a market for original music and media. However, it’s essential for both copyright holders and music creators to be proactive and adaptable to protect their interests in the face of technological advancements like AI-generated music.” Here’s the part that I love: “Remember that while AI can generate music, it often lacks the emotional depth, creativity and human touch that comes from human composers and musicians. Music created by humans will always have the unique value that AI cannot completely replicate. By staying informed, protecting your rights and embracing technology as a tool rather than a threat, you can continue to thrive in the evolving landscape of the music industry.”
So, my question to you, Adam, is, “Do you think AI music creation will kill off the need, and the use of, music created by the people sitting in this ballroom right now?”
Well, I can’t see the people in the ballroom [Adam joined us by zoom], so I don’t… [laughter]
“I don’t think that AI would have created Rock from everything that came before. I don’t think that it would have created Hip-Hop from everything that came before. And even if it had, would it catch on?”
If I didn’t have so many wires and microphones on my laptop right now, I’d turn it around so you could.
No, I really don’t. I don’t have a crystal ball; I don’t know where all of it’s gonna go and where we’re gonna be 50 or 100 years from now, or whatever. But I think in our lifetimes—yours and mine and younger people in the room—I think we’re relatively safe. I think that there is always going to be a unique value that human beings give. I don’t think that AI would have created Rock from everything that came before. I don’t think that it would have created Hip-Hop from everything that came before. And even if it had, would it catch on? Would it become a social, cultural movement? Music is a cultural vehicle; it’s a crucible in which all of our emotions, our passions, our cultures mix to create something new. And that’s going to be a human endeavor and that’s the important part of music. You could look at library music or music for programming, maybe because people like to think it’s a commodity or whatever, and certainly some of these royalty-free and crowdsource sites tend to reinforce the commoditization of music. But nonetheless, the quality of the music that is produced by human beings that has been and that will continue to be produced is extremely important to the culture as a whole. It’s incredibly important for storytelling, and I just don’t think that AI-generative music is going to be able to take its place. I also think there will probably be a pushback. I’ve already seen it, people not wanting to use it. And for us, we don’t accept AI music, so there’s a pushback. [applause]
That was real applause.
What was that noise? It seems like there’s some kind of artifact going on. It must be AI-generated, I don’t know. [laughter]
I think that using AI to assist to come up with ideas… There are all kinds of things; AI is just advanced software. It will help stimulate ideas, like somebody might use a beat library or something. But I think that how that is all assembled has to be your own voice, and I think that is going to remain. That’s my personal opinion, and it’s the opinion I brought into APM and it’s the model that we are following. We have not slowed down the amount that we’re investing in the music and the technology and everything in order to continue to market human creative music.
Great. I am really glad to hear that. I worried… I think the day is not that far away when an editor in an edit bay, rather than having to search catalogs or genres within a catalog, they know they need dramedy for—my favorite example is when Kim Kardashian breaks a fingernail trying to open a milk container—that the editor will simply be able to say, “Alexa, find me music for the scene,” and AI will already know the dialogue in the scene and maybe the general emotional tenor of the scene and suggest music without an editor having to think, “Wow, I need dramedy for this,” or search through a catalog.
However, there are so many other complications and obstacles in the path between that, which I believe technologically could happen at some point, but I think that… and you’re going to know far more about this than I will. Networks. How do the networks feel about AI? Have they made any proclamations yet? Are they not going to want to use music that’s not copyrighted, because you can’t copyright something that a robot makes?
The first thing I just want to say is that these editors or producers better be pretty careful, because at some point you could just ask AI, “Just edit music into the video that does the following,” and create the video, and you don’t need the cameraman and you don’t need the editor. And I agree with you; I think that at a certain level some people are going to be doing that. They will do that for kind of inexpensive stuff, and maybe these subscription models. If you can get music for $5 or $10 a month, then you’re probably not caring that much what the music is; you just want something very rapid. So, I think that maybe there is more of a risk from these tools on that level than there is on the kind of music that the people who are TAXI members produce.
In terms of the networks, we haven’t heard a network or a studio take a position [on AI-generated music] at this point, because they were all and are still involved in the Writer’s Guild strike, the WGA strike, which was a lot about AI. My wife is a TV writer, so I’m very familiar; she was on the board of the WGA for a long time. They were wanting to be able reduce the size of writers’ rooms. Instead of having five writers in a room, they wanted to be able to have one writer in a room and then have an AI tool—somebody working with the AI tool. And they ultimately decided in the settlement that they are not going to do that.
But they are not speaking publicly about their position, and they are still of course in the negotiations with SAG, AFTRA for actors, and they want to be able to use their likenesses to be able to create other scenes and things like that. However, again, they did sign the WGA agreement, and so they have accepted for now that they’re not going to be using AI for that purpose, and they are protecting authorship by humans.
“In U.S. Copyright law you can’t sue for copyright infringement unless you are the owner or an exclusive representative.”
From the legal perspective, you can’t copyright, as you said, and therefore you can’t indemnify. In U.S. Copyright law you can’t sue for copyright infringement unless you are the owner or an exclusive representative. A non-exclusive owner of something or representative or agent of something can’t protect the copyright. That’s another problem with the non-exclusive sites is that it’s very difficult to protect a copyright. If somebody were to infringe, you’d have to go back to the individual who wrote the song.
So, I think that every single one of our contracts has a thing that we own or control the copyright and that we fully indemnify them against any potential copyright infringement. Thankfully, there have been very few that have been an issue, but there have been some.
“What happens if a computer infringes on something?”
I just got an email yesterday from my litigation counsel, who wanted a favor. He said that a friend of his was a lawyer at another firm. He’s representing a different music library, not APM, and they have gotten a cease-and-desist and a demand letter from one of the big acts out there saying that that library track infringes on their copyright, and they want $325,000 dollars. So, he asked me a question about it, which I answered. But these things do happen. What happens if a computer infringes on something?
I think that the networks and the serious professional companies are not going to be using AI music. So, keep writing, everybody.
I’m comforted by that, because, honestly, people always ask me, “Why doesn’t TAXI work with some of the big royalty-free libraries?” It’s because I’m trying to elevate the craft and the use of music and not go for the lowest common denominator. But I do understand that there is a need for user-generated content—wedding videos are my default example for that. So maybe both things can co-exist in their own place and appropriately so over time.
On a global basis—because this is not just an American problem—do you think that the various PROs or societies on a global basis are meeting behind closed doors? I’m hoping they are trying to figure out how they can get involved in lobbying, because they don’t exist without you, without the people in this room. So, I kind of feel like they need… You know, they’ve got strong political lobbies; they know their way around D.C. Have you gotten wind of anything that you can divulge that the PROs are doing on behalf of the musicians in this room?
Sure. As you can imagine, there are a lot of conversations going on around the world with the PROs and with various governments. In the United States I met not too long ago with Shira Perlmutter, who’s the registrar for the U.S. Copyright office, and she told me and a number of other people—there was a meeting over at BMI’s offices here in Los Angeles—that they will not be accepting AI-generative music for copyright registration. So, it was a very strong statement that she made, and they’ve come out with that publicly. And I think that that is going to remain for a long time. I tell people that the countries around the world can’t even agree on the length of a copyright, so how are they going to agree on this in every country in the world what is acceptable and what isn’t acceptable? We are decades away from that, in my opinion.
However, governments are passing guidelines. I know these two countries are very different, Spain or Japan—one of the two—has already passed a law about this. And then there were disclosure laws, the things that are written by AI have to be disclosed and other regulations. Some PROs… pretty sure that SACEM in France and other ones have already publicly released statements saying that they are not accepting AI music for registration. And I know that ASCAP and BMI are in those discussions right now. So, I think that you’re going to see more and more of that.
But I also think that, as I mentioned before, I think there is going to be a pushback by the culture at large. And I’m not trying to be a Luddite; software has its purpose, AI has its purpose and it’ll be used for a lot of different things. I personally feel that the best work will come from composers who use all of the tools that are available to them, and AI certainly is a tool and that’s how I think it should be looked at. But I think from a pushback perspective… We all remember what happened with the Google Glasses, or Google Glass, whatever it was called, with people walking around. It was too early, and there was such a pushback from the culture that they abandoned it. I don’t think that AI music is going to get abandoned, but I think that it’ll find its level, but I don’t think that it’s going to undermine the ability of composers to earn a living.
So there’s already been a couple of large royalty-free libraries where the quality isn’t very good, in my opinion. Some of them have already allowed AI, and others have AI to scrape their catalog for data. For anybody in the room that’s not familiar, they are not sampling the music, they are not listening to a string sound and taking an actual digital recording of it, it’s zeroes and ones. It’s looking at all these different digital aspects of the piece of music, but it’s not sampling it.
There have been catalogs that had contracts [with their composers and writers] that allowed them to let the AI companies scrape the catalog for data, and then they send checks to the composers in the catalog. My question is, is there a way to fairly, adequately and thoroughly compensate musicians who are in a catalog whose music gets used as source data for AI, eventually. Like how the hell do you know if a song is one one-millionth of a percent of your guitar playing and somebody else’s bass playing? It’s not sampling, and it’s all these little micro components that are all kind of reassembled back as more zeroes and ones. Will there ever be a way that they can fairly and thoroughly compensate people whose music is in a catalog where it’s legal to scrape it for data?
First, let me say that the libraries and the companies that are doing this, to me, it’s not a good idea, even if they are getting compensated, even if the composers are getting compensated. It’s very short-sighted in my opinion, and it’s just focused on getting some money in, it’s not protecting the value of the creative endeavor—and so it has nothing to do with protecting composers or anything like that. And I have no problem naming the companies that are doing this, you know.
“Even if there is a mechanism for compensation, if you can’t measure it and track it, it’s never going to be fair.”
Be my guest.
Adobe, Shutterstock, Pond5, and others have announced it with pride. So, I think it’s hubris, and I don’t think there’s a really equitable way to do it. I think that if you take a defined set of music and you get X amount of dollars, and you say, “Okay, I’m gonna give X percent of that to those composers,” that one [the first] generation of people, then maybe that’s fair, trackable, but after that, what happens? All of that information, all of the learning that comes from that music is then no longer associated with that music, and there’s no way for the next generation of learning to be able to adequately or fairly compensate [the original creators]. And even if there is a mechanism for compensation, if you can’t measure it and track it, it’s never going to be fair. It’s not transparent, and I think we all know that systems of distribution of revenue or income or advantage are never fair and never even, and if you can’t track it, you’re not going to get your fair share. So, I think it is a false flag to say, “Yeah, we’re compensating composers for taking your music and allowing a computer to create new stuff.” It’s a false flag; it’s just to build their business and reduce their costs. That’s my opinion.
Don’t miss Part Three of this interview in next month’s TAXI Transmitter!