Field Dispatch
Say No to SUNO | Matt Connarton Unleashed
Speaker 1: I do want to We've got a little bit of
Speaker 1: music news. This particularly caught my eye, so we have
Speaker 1: we do have a few minutes before we uh before
Speaker 1: we hit the top of the hour, So I wanted
Speaker 1: to talk about this because this well, in a sense
Speaker 1: it reflects or effects. I should say something that I've used.
Speaker 1: This is from Music Businessworldwide dot com artist representatives launch
Speaker 1: say No to Suno campaign Now Suno if you're not familiar,
Speaker 1: it's an app. It's that uses AI. You can actually
Speaker 1: create songs using AI. You just put in some simple prompts,
Speaker 1: tell it what you want, and it will create a
Speaker 1: song for you. Jenny and I it was over a
Speaker 1: year ago now, maybe a year and a half ago,
Speaker 1: the first time we ever used Suno. We did an
Speaker 1: experiment with it, kind of live on the air, just
Speaker 1: to see what would happen. It's very easy to use,
Speaker 1: and it creates some very convincing you know, like you
Speaker 1: wouldn't know just from listening to what it creates that
Speaker 1: it's AI generated necessarily. I've used it quite a bit.
Speaker 1: We've done some really fun things with it. But of
Speaker 1: course we talk about AI a lot on the show,
Speaker 1: and it's an effect on the music industry and how
Speaker 1: artists feel about it, And so now there is this
Speaker 1: say No to Suno campaign. Now, I have no intentions
Speaker 1: of involving myself in any way in the campaign because
Speaker 1: I actually like Suno and it would be hypocritical of
Speaker 1: me to do so, even though my position ultimately is
Speaker 1: and I think everyone knows this because of what I do,
Speaker 1: what I'm involved in, the people I know, et cetera,
Speaker 1: and the various you know, in terms of careers, the
Speaker 1: various things I've done, I will always at the end
Speaker 1: of the day, I will always always advocate on behalf
Speaker 1: of artists. That's what I care about the most. But
Speaker 1: that doesn't mean I won't use an AI app like
Speaker 1: Suno to create a fun, silly song, because I do
Speaker 1: use it for that quite a bit. So. Artists representatives, though,
Speaker 1: have launched the say No to Suno campaign, saying AI
Speaker 1: slop dilutes the royalty pools of legitimate artists from whose
Speaker 1: music this slop is derived. So let's take a look
Speaker 1: at this because again, and who knows, who knows? Maybe
Speaker 1: looking at this article, maybe there's information in here that
Speaker 1: will change my mind and make me decide to join
Speaker 1: the say No to Suno campaign. I doubt it, but
Speaker 1: let's see so it says here again. This is from
Speaker 1: Music Businessworldwide dot com. A coalition of artist representatives has
Speaker 1: published an open letter calling on the music community to
Speaker 1: reject AI music generator Suno. In an open letter titled
Speaker 1: say No to Suno, the artist reps described the company
Speaker 1: as a brazen smash and grab platform, accusing it of
Speaker 1: using unauthorized AI platform machinery trained on human artists. Work
Speaker 1: published Monday, the twenty third, So this just happened this
Speaker 1: past week on the Music Technology Policy Blog. The letter
Speaker 1: was signed by figures including Ray Gubitz, executive director of
Speaker 1: the Music Artists Coalition, Helene Lindvall, songwriter and president of
Speaker 1: the European Composer and Songwriter Alliance, and Chris Castle of
Speaker 1: the Artist Rights Institute. Other signatories included artist David C. Lowry,
Speaker 1: artist and Artist Rights Alliance board member Tift Merritt, Blake Morgan,
Speaker 1: artist producer and president of VCR Music Group, and Abbey North,
Speaker 1: president of North Music Group. The letter arrives as Suno
Speaker 1: continues to face copyright infringement lawsuits from major music companies
Speaker 1: and rights organizations and by the way. By the way
Speaker 1: I tell everybody, I say, if you're enjoying these apps
Speaker 1: like Suno or Udio is another one, a very popular
Speaker 1: AI music generator. If you enjoy these apps and use
Speaker 1: these apps, enjoy them while we have them. Because the
Speaker 1: sharks are circling. Okay, there's all kinds of litigation and
Speaker 1: all kinds of stuff happening to try to get rid
Speaker 1: of these apps permanently or well, probably no getting rid
Speaker 1: of anything permanently, and there's probably no getting rid of
Speaker 1: these apps period. But I'm just saying, don't be shocked
Speaker 1: if one day, all of a sudden, you can't use
Speaker 1: these anymore because they've been shut down. Okay, they've been
Speaker 1: bankrupted by lawsuits, whatever whatever may happen. But right now,
Speaker 1: this is what's going on. Okay, so so uh. The
Speaker 1: RIAA filed suit against both Suno and rival Udo in
Speaker 1: mid twenty twenty four, acting on behalf of all three majors,
Speaker 1: alleging mass infringement of copyright. Udio has since reached settlements
Speaker 1: with both Universal Music Group and Warner Music Group, signing
Speaker 1: licensing agreements with each for a new AI music platform
Speaker 1: expected to launch this year. Interesting WMG Music Group also
Speaker 1: settled separately with Suno in November, but the AI company
Speaker 1: remains locked in legal battles with UMG and Sony Music Entertainment,
Speaker 1: as well as European music rights holders, including Denmark's CODA
Speaker 1: and Germany's Gema Gema. I've not used either of those.
Speaker 1: In fact, I've never used Udio either. The only one
Speaker 1: I've used as Souno Okay, it says here. The open
Speaker 1: letter took aim at the broader concept of AI in
Speaker 1: the music business, arguing that unauthorized generative AI platforms are
Speaker 1: fundamentally different from previous technological disruptions in the music industry.
Speaker 1: And by the way, as we've discussed many times on
Speaker 1: the show, Jenny and I've talked about this, the music
Speaker 1: industry has had many technological disruptions over the years, you know,
Speaker 1: transitioning from vinyl and cassettes to CDs, napster. You know,
Speaker 1: when Napster first came in and file sharing became a thing,
Speaker 1: and all of a sudden, music was free. I was
Speaker 1: working for trans World Entertainment at the time, which owned Fye,
Speaker 1: Which and Strawberries and everything. When I was working for
Speaker 1: them as a store manager in a music store, and
Speaker 1: you know, we saw firsthand what was happening. All of
Speaker 1: a sudden, people aren't buying CDs because they're getting the
Speaker 1: music for free. And then there was all these debates
Speaker 1: about while is file sharing stealing or are you just
Speaker 1: taking advantage of what's available and so forth. We won't
Speaker 1: get into all of that, but anyway, there's been many
Speaker 1: That's probably the single biggest technological disruption that's happened in
Speaker 1: the music industry is the Internet and the ability to
Speaker 1: access music easily and without paying for it. That's probably
Speaker 1: the biggest one, at least leading up to the advent
Speaker 1: of AI. Well, not the advent of it, because AI
Speaker 1: has been with us for a long time, but now
Speaker 1: with these apps, the advent of these apps like Pseudo
Speaker 1: and Udio. Okay, so it says here again, this is
Speaker 1: from Music Business Worldwide dot com. The open letter took
Speaker 1: aim Actually, well, I read this already, but this paragraph
Speaker 1: is worth reading again. The open letter took aim at
Speaker 1: the broader concept of AI and the music business, arguing
Speaker 1: that unauthorized generative AI platforms are fundamentally different from previous
Speaker 1: technological disruptions in the industry. The letter argued that the
Speaker 1: quote hijacking of the world's entire treasure trove of music,
Speaker 1: floods platforms with AI slop, and dilutes the royalty pools
Speaker 1: of legitimate artists from whose music the slop is derived unquote. Right,
Speaker 1: because when you create something using these apps, you're creating
Speaker 1: something that derives its content from all all the content
Speaker 1: that it has learned. You know, we call them large
Speaker 1: learning models. You know, it's taking that from from everything
Speaker 1: else that's available that it's learned, that it's taken in. Now,
Speaker 1: when we talk about whether or not that's okay, my
Speaker 1: argument has always been if I am arguing on the
Speaker 1: side of not having a problem with any of this,
Speaker 1: which I really don't, because you can't hold back technology.
Speaker 1: It never works out. You have to adapt. You can't,
Speaker 1: but you can't stop it. So you have to adapt.
Speaker 1: When I let's say I write a song, and I've
Speaker 1: written songs, I'm a musician. If I write a song,
Speaker 1: anything that I write is somehow informed and influenced and
Speaker 1: therefore is a product of all the music that I've
Speaker 1: ever heard over the course of my life. Right, anything
Speaker 1: that I create is somehow derived from everything that I've heard.
Speaker 1: So how is that different from when a large learning
Speaker 1: model learns music and then creates something from that. Now
Speaker 1: it is different in fundamental ways. But I'm just offering
Speaker 1: as a counter argument to you know, everyone who says,
Speaker 1: oh AI it's got to be stopped, we can't have
Speaker 1: AI generated music and so forth. Well, is it really
Speaker 1: different from when a human creates music? Now it is different. Again,
Speaker 1: I'm just playing Devil's advocate here, but that would be
Speaker 1: my counter argument to that. Well, if I write a
Speaker 1: song as a human being, my song that doesn't come
Speaker 1: out of thin air. My song is based on knowledge
Speaker 1: and information and experience and all of it from my
Speaker 1: own life and all the music that I've ever consumed
Speaker 1: over the course of my life. That's all I'm saying. Okay,
Speaker 1: getting back to the article, The letter arrives amid a
Speaker 1: growing volume of AI generated music hitting streaming platforms. Now,
Speaker 1: this is a problem that we've talked about on the
Speaker 1: show before too. Is all this AI generated music that
Speaker 1: ends up on Spotify, for example, Well Spotify already, I mean,
Speaker 1: I don't remember the numbers. We've talked about it before
Speaker 1: on the show. The amount of music that is up
Speaker 1: that is uploaded to Spotify globally on a daily basis
Speaker 1: is staggering. So then Spotify has to figure out, well
Speaker 1: what to get rid of because now that you factor
Speaker 1: in the ease with which you can create music using
Speaker 1: an app and then just upload it. Now it's it's
Speaker 1: that problem of you know, you only have so much
Speaker 1: storage and you have to keep creating more storage. And
Speaker 1: you know now that because people tend to think of
Speaker 1: it as well, it's just data. It doesn't take up
Speaker 1: any space physically, it's just data. But that's not true.
Speaker 1: Think about your own computer, your own devices, right, you
Speaker 1: only have so much memory. You know, I only have
Speaker 1: so much memory on my cell phone. Now I keep
Speaker 1: my cell phone very clean because I don't want to
Speaker 1: getting clogged up with stuff I don't need. But I
Speaker 1: only have so much memory on it. So it's not
Speaker 1: like it's no different for Spotify, for example, they only
Speaker 1: have so much memory. Now they can create more, but
Speaker 1: there is an expense to and also the larger the
Speaker 1: catalog of music that is available on the platform, doesn't
Speaker 1: it kind of dilute everything that's already there. So this
Speaker 1: is a problem for Spotify. Not that I have any
Speaker 1: great sympathy for Spotify because they don't pay very much,
Speaker 1: but you know, in terms of royalties and so forth.
Speaker 1: But you know, but now they've got a whole new problem.
Speaker 1: This is why a few years ago, before before these
Speaker 1: apps came out, even a few years ago, Spotify was
Speaker 1: struggling with this. They were having to start to go
Speaker 1: through I mean, obviously they have bots that do it,
Speaker 1: but having to go through and delete low quality demo
Speaker 1: recordings that people were just uploading to Spotify, thus diluting
Speaker 1: what's on the platform. People uploading things that probably shouldn't
Speaker 1: have been on there to begin with, because they weren't
Speaker 1: of high enough quality to even be on the platform
Speaker 1: to begin with. So now they've got a whole new problem.
Speaker 1: Let's see. According to getting back to the article, according
Speaker 1: to recent data from Deezer, proximately sixty thousand AI generated
Speaker 1: tracks are being uploaded to the Paris headquartered streaming service
Speaker 1: every day, with synthetic content now accounting for roughly thirty
Speaker 1: nine percent of all music delivered to the platform daily.
Speaker 1: Let's look at that again these numbers. According to recent
Speaker 1: data from Deezer, approximately sixty thousand, sixty thousand AI generated
Speaker 1: tracks every day are being uploaded. These are also revealed
Speaker 1: that up to eighty five percent of all streams on
Speaker 1: AI generated music on its platform were fraudulent in twenty
Speaker 1: twenty five, up from the seventy percent figure it reported
Speaker 1: in the prior year. Those streams are demonetized and removed
Speaker 1: from the royalty pool. By comparison, in streaming fraud across
Speaker 1: Caesers entire catalog accounted for eight percent of all streams
Speaker 1: in twenty twenty five. By the way, I'm not sure.
Speaker 1: I don't know what they mean by fraudulent. That is
Speaker 1: not clear to me. What do they mean by fraudulent?
Speaker 1: I'm not sure. Do they just mean anything that's created
Speaker 1: with an app or what do they mean? While AI
Speaker 1: generated music currently makes up only a small share of
Speaker 1: total streams on Deezer up to three percent, the platform
Speaker 1: maintains that generating fake streams remains the primary purpose for
Speaker 1: uploading such content. The letter also weighted into the ongoing
Speaker 1: industry debate over so called wald gardens in AI music,
Speaker 1: criticizing a recent LinkedIn post by Paul Sinclairs, who Knows
Speaker 1: chief music officer and EXWMG executive, who argued that closed
Speaker 1: AI systems limit how people engage with music Sinclair penned
Speaker 1: a lengthy post Grammy Week LinkedIn memo entitled open Studios,
Speaker 1: not wald Gardens, which took direct aim at the approach
Speaker 1: champion by universals A group in its recent AI licensing agreements.
Speaker 1: UMG was the first major to settle its litigation with
Speaker 1: Suno rival Udio in October twenty twenty five, alongside a
Speaker 1: licensing deal for a new AI platform set to launch
Speaker 1: this year. The concept of a walled garden was introduced
Speaker 1: within that announcement, a model where AI generated music cannot
Speaker 1: be downloaded or distributed outside the platform. Udio disabled downloads,
Speaker 1: with users given a forty eight hour grace period to
Speaker 1: retrieve previously created tracks before the walls went up. By
Speaker 1: the way, this is the first time hearing about this
Speaker 1: the walled garden concept. Around two months after that deal
Speaker 1: was announced, in his annual memo to staff, UMG chairman
Speaker 1: and CEO, Sir Lucian Grange appeared to warn against firms
Speaker 1: quote validating business models that failed to respect artists work
Speaker 1: and creativity and promote the exponential growth of AI slop
Speaker 1: on streaming platforms. So, in other words, he's saying, don't
Speaker 1: do this, don't make deals with AI, you're making a
Speaker 1: deal with the devil. Effectively, you're getting in bed with
Speaker 1: something you don't want. You don't want to encourage this.
Speaker 1: If you start making deals with AI, you're going to
Speaker 1: encourage this. I guess a quick counter argument might be, well,
Speaker 1: if you look back to the previous seismic disruption in
Speaker 1: the music industry, if you go back to Napster, what
Speaker 1: was happening. Well, you know, the industry finally had to
Speaker 1: start making some deals and some compromises and working some
Speaker 1: stuff out. But what he's saying is, don't do that,
Speaker 1: Do not encourage it. Michael Nash, executive vice president and
Speaker 1: chief Digital Officer at UMG, explained the company's position in
Speaker 1: further detail during an appearance on Billboards on the Record
Speaker 1: podcast last month and outline the waldgarden concept in stark terms.
Speaker 1: He said this quote, The concept of a wald garden
Speaker 1: is to set up, through AI, a component of the
Speaker 1: service for deep interaction with the artists and the content,
Speaker 1: but not to create derivatives that you then take off
Speaker 1: of the platform and post all over your socials and
Speaker 1: post on Spotify and Amazon Music and Apple unquote. He
Speaker 1: argued that by doing so you quote effectively use artists
Speaker 1: content and their brand to create derivative derivatives where you're
Speaker 1: going to compete with the artists on other platforms unquote.
Speaker 1: Beyond Sinclair, who joined Suno in twenty twenty five July
Speaker 1: twenty twenty five, the company has made other high profile
Speaker 1: hires from the music industry to support its efforts to
Speaker 1: build relationships with the sector, including former Merlin CEO Jimmy
Speaker 1: Sorry Jeremy Sirata as chief commercial Officer and former Spotify
Speaker 1: executive Sam Berger as Senior Director of Artist Partnerships. In November,
Speaker 1: Suno closed a two hundred and fifty million dollars Series
Speaker 1: C funding round. I don't want to get in all that. Okay,
Speaker 1: never mind, don't care about that part. But that's the
Speaker 1: end of the article anyway. But there you go. So again,
Speaker 1: I will not be joining any such campaign. It would
Speaker 1: be a critical of me to do so. But this
Speaker 1: is all about the say No to Suno campaign.
Podbean