Kyle Chayka: The algorithm as your best friend, as the intimate knower of your innermost secrets, is definitely what Spotify would love for it to be.

Mark Henry Phillips: Guys, is this how it ends? Is this the sign of the death of music reviews?

Brooke Gladstone: As the year comes to a close, amid a barrage of best music lists, we reflect on the state of music criticism.

Anne Powers: The critic is in a parallel relationship with the musician, who also is revered and scorned in our society. [chuckles] All of this says something about how we treat culture.

Brooke Gladstone: From WNYC in New York, this is On the Media, I'm Brooke Gladstone.

Micah Loewinger: I'm Micah Loewinger. Also on this week's show, an AI music generator gives a composer a run for his money.

Mark Henry Phillips: I don't want to admit it, but it's good. Like really good. It doesn't feel like it was created by AI and a data center. It feels real.

Brooke Gladstone: It's all coming up after this. From WNYC in New York, this is On the Media. I'm Brooke Gladstone.

Micah Loewinger: I'm Micah Loewinger. It's December, and that means it's listicle season, when critics like to rank their favorite things from the past 12 months. The best movies, best TV shows, best books, and super popular in the genre with an assist from Spotify is the favorite track list.

Speaker 6: What time is it? December. Which means we're ending the year with Spotify Wrapped.

Speaker 7: Your 2024. Wrapped is here.

Speaker 8: Oh my God, I'm so excited.

Speaker 9: What were the best albums of 2024? Let's get into it, shall we?

Speaker 10: 2024 has been the best year for music this decade, hands down.

Micah Loewinger: One of the most vaunted roundups of music making once belonged to the iconic online music publication Pitchfork, which was founded in 1996. They still make waves with their rankings and year end roundups, but back in January, Pitchfork shrank considerably when it was folded into GQ by Condé Nast and most of its staff were cut. Earlier this year, I spoke to NPR music critic Anne Powers about the role Pitchfork has played in shaping tastes and music criticism today.

Anne Powers: The very opinionated reviews, the scoring system that became notorious where they give a number score to new releases, and that blog-like constant stream of content is what made Pitchfork so important.

Micah Loewinger: Tell me a little bit about some of the bands and genres that were elevated because of Pitchfork. Or at least were the favorites of the cast of writers in its heyday.

Anne Powers: Well, Micah, I'm curious, what bands do you associate with Pitchfork? I bet you're a Pitchfork reader.

Micah Loewinger: I was a big Pitchfork reader when I was in high school and in college, the bands I associated with it were Spoon or Phoenix or Radiohead, Animal Collective.

Anne Powers: Broken Social Scene?

Micah Loewinger: Yes, Broken Social Scene.

Anne Powers: Arcade Fire would be another one. These are really the leading "indie" rock bands of the late 2000s and early 2010s, coming out of the indie rock tradition that in previous generations had given us bands ranging from R.E.M. to Nirvana. So really, Pitchfork was the 21st century flag bearer for the kind of music that Rolling Stone would have covered in the '60s and '70s and Spin magazine would have covered in the '80s and '90s. Pitchfork's strong association with that kind of music in the early 2000s is what made it such a potent brand and gave it that foothold that allowed it to truly show its influence.

Micah Loewinger: Yes, and to put even just a finer point on it, there was a 2006 piece in Wired titled the Pitchfork Effect.

Anne Powers: Right.

Micah Loewinger: That described the make or break power it had over that era of bands.

Anne Powers: Yes. Let's put this in the historical context of music magazines going back even before the rock era. In the jazz world, for example, you had downbeat its annual polls, its reviews, its articles had the similar effect on the jazz world.

Micah Loewinger: I want to ask you, though, about the style of writing the voice, because it wasn't just their curation, it was something else about how they wrote. Can you describe it?

Anne Powers: The basic unit of Pitchfork is the album review. And what is an album review? This is a philosophical question, Micah. [chuckles]

Micah Loewinger: What is an album review, Ann?

Anne Powers: It can be something as small and simple as a little blurb that says, "Hey, you're going to like this." Or it can be as long as a cover and really become an essay that considers music in many different contexts. Or possibly it could be a vehicle for personal expression, for memoir, for expressing ideas from a very opinionated place. I think one of the things that makes Pitchfork so important is that Schreiber and then the other editors there have allowed writers to really develop a voice.

Micah Loewinger: Sometimes that voice was mean.

Anne Powers: [laughs] Yes, negative reviews, they're kind of like the Thor's Hammer of Pitchfork, writers there and the editors there would wield those negative reviews as a way of proving their influence and a way of generating discussion.

Micah Loewinger: Some of those 0 out of 10 reviews haven't aged so well. Matt LeMay had given Liz Phair's 2003 self titled album a zero at the time he, I think, was maybe 18 or 19, and he didn't like that Liz Phair, who was an indie musician, had come out with a more pop radio friendly album.

Anne Powers: Right.

Micah Loewinger: But then in 2019, LeMay returned to his review and described it as condescending and cringy. He tweeted, "The idea that indie rock and radio pop are both cultural constructs, language to play with, masks for an artist to try on, yes, I certainly didn't get that. Liz Phair did get that way before many of us did."

Anne Powers: Kudos to Matt LeMay for engaging in some very constructive self criticism and kudos to Pitchfork, because I think one of the best things that Pitchfork has done in recent years and under the guidance of Puja Patel, the editor in chief, who was let go as part of these layoffs, is they have revisited old reviews, they have created a whole feature that allowed for them to review records they'd ignored from genres they weren't as interested in. This self examination and confrontation with the limits and problems of the Pitchfork approach, to me, that is one of the most inspiring aspects of what's been happening in music writing in the past decade or so.

We have a much more diverse field of music writers now. Many more women, many more people of color, many more LGBTQIA people writing about music. That diversity has totally changed what we do. I think it's great, and Pitchfork has been a huge part of that.

Micah Loewinger: Yes, and you've mentioned that as the publication's aperture has expanded, they now cover a much more varied range of musicians. How has that affected the editorial experience, you think?

Anne Powers: Yes, Pitchfork covers pop, but they mostly cover pop and varying mainstream artists in their news section. Yes, they will review a Taylor Swift record, for example, and certainly a Beyonce record, as any of us would. They are also covering, you know, very obscure electronic music, Avant Garde jazz or Avant Garde or classical music or Americana music. I think it's that diversity that makes the review section particularly so valuable because it is like going into a great huge record store where you could happen upon something you didn't expect.

Let's be real, there's so much more to music than the cool new bunch of dudes in tight pants or whatever.

Micah Loewinger: Playing some growling guitars typically.

Anne Powers: Yes. I love that stuff, but I think it's great that Pitchfork grew up. So many of my favorite writers who've come through Pitchfork in recent years, they're diversifying the field. That, to me is so crucial.

Micah Loewinger: Not everyone has been on board with the changes at the site. Writing in the Guardian this week, Laura Snapes responded to critics of Pitchfork who have, "lamented Pitchfork's poptimist shift over the past decade". Poptimist. What is she referring to there?

Anne Powers: I'm glad you brought up that term, Micah, because it's one that drives me crazy. The word poptimism originated in response to an essay that Kelefa Sanneh wrote in the New York Times called the Rap Against Rockism. Kelefa's criticism of the world of music writing was that it was dominated by straight white men who like guitar based rock music made by straight white men, and that this had created a hierarchy within the music industry. Quickly, this critique created a space for some of us to say, "Hey, let's also take mainstream pop music seriously.

Let's take dance music seriously. Let's take these fields that happen to be dominated by African American artists, by women. Let's make a space for that." Carl Wilson also wrote a really important book. It was about Celine Dion, but it was really about how our tastes form.

Micah Loewinger: Oh, yes. This is A Journey to the End of Taste. Is that what it is?

Anne Powers: Yes.

Micah Loewinger: It's a fascinating premise. He basically says, "Celine Dion, one of the best selling artists of our time, I hate her music. Why?"

[laughter]

Anne Powers: He was saying, "Okay, from my standpoint as a fan of indie rock, as a white guy, et cetera, what am I bringing to the table when I listen to a Celine Dion record? Why do I think this is "bad music"? Why do so many other people think it's great music? That's the essence of what the poptimist project really was. It was not to promote mainstream music. It was to take seriously music that is very popular music that rock critics scorned historically. I think Pitchfork's evolution from a site that sort of embodied that scorn to one that was fighting against it was one of the most beautiful things that's happened in media in the past few decades.

Micah Loewinger: It's been interesting to see the evolution of Pitchfork land between all of these competing interests. I'm thinking of the 2020 review of Taylor Swift's surprise indie folk album Folklore, written by Jillian Mapes, who wrote a pretty positive review of the album, but ultimately the site only gave the album an eight.

Anne Powers: Right.

Micah Loewinger: She was sent death threats, constant harassment online. It's just fascinating to me that on one hand you have people who bristle at the very fact that Pitchfork is reviewing Taylor Swift, and on the other, [laughs] fans of Taylor Swift aren't happy with a critical review about her.

Anne Powers: The critic has always been an embattled figure in our society, both revered and utterly disrespected, both considered a nothing who only lives through the works of others, and someone that supposedly makes people tremble when they walk in the room. Right? In a strange way, the critic is in a parallel relationship with the musician, who also is revered and scorned in our society. All of this says something about how we treat culture. On the one hand, there are these attempts to sportsify it, to quantify it, to make hierarchies, which always, inevitably fail because encounters with art are personal.

On the other hand, there is such a thing as aesthetic judgment. I think that that combination of stepping back and being close at the same time, it's a complicated way to talk about culture, and it can be upsetting to some.

Micah Loewinger: Ann, thanks so much.

Anne Powers: It was great talking to you.

Micah Loewinger: Ann Powers is a critic and correspondent for NPR Music. She's the author of several books, including the biography Traveling: On the Path of Joni Mitchell. This is On the Media.

Brooke Gladstone: This is On the Media. I'm Brooke Gladstone.

Micah Loewinger: I'm Micah Loewinger. Before I landed a job at this show, I worked for a few years on and off at a couple record stores around New York City. Some of my favorite albums to this day were recommended to me by my co workers, men and women who I consider to be archivists not just of old formats like vinyl records, cassettes and CDs, but of underappreciated artists and niche genres, a knowledge of music history that can only come from a lifetime of obsessive listening, research and curation. Nowadays, I pay for Spotify.

I try to learn about music off the app and then save it for later listening on Spotify. Sometimes I find myself just letting its recommendation algorithm queue up the next track and the next, and it definitely works. Spotify has helped me discover great music, but it's never been as revelatory as a personal recommendation from a friend or an expert at a record store or an independent radio station. This feeling that I've traded convenience for something deeper is what made me want to read Filterworld: How Algorithms Flattened Culture by Kyle Chayka, a staff writer at the New Yorker.

Chayka says apps like Spotify and TikTok are great at studying user behavior, but that we should be suspicious of the idea that they can really know your taste.

Kyle Chayka: You are not aware of every time you click into a Spotify track. You're not aware of when you favorite an album on TikTok. You're absolutely not aware of every microsecond that you flip up a video or what you pay attention to a tiny bit longer than Something else. It does know exactly what you're doing and it doesn't forget, like that one time that you lingered too long on a shower tiling video. [laughter] It's like, "You remember those shower tiling videos. Let me give you some more."

Micah Loewinger: It would be good to define our terms a little bit. I know Spotify's algorithm is a trade secret. We don't know exactly how it works, but as you write in your book, there are clues based on literature about the development of recommendation algorithms that might tell us how it likely works.

Kyle Chayka: Most recommendation algorithms are black boxes because the company itself does not want you to know how it works because you might game it and that would ruin their product. A lot of them work along the same lines, essentially measuring a bunch of variables about the content that's on their platform, how many times people have clicked it, what the faves are, what the retweets are, what the time watched is, and then using that to figure out what to promote more and what to kind of push off to one side or another.

Micah Loewinger: Spotify doesn't know what's good or bad.

Kyle Chayka: Right? That's a fundamental thing. Algorithmic recommendation is not about quality. There is no essential metric of quality. There is only attention. It can do thumbs up, thumbs down. But it can't be like, "Oh, Bach is better than Mozart."

Micah Loewinger: Spotify's recommendation algorithm is just one part of what you call filter world. It's the name of your book, but it's also a concept. Can you describe it?

Kyle Chayka: Yes. Filter World, for me was this single term to describe the entire ecosystem of algorithmic feeds that we exist in. When we're on the Internet today, we are moving across all of these different platforms, whether Facebook or TikTok or Instagram, that are all driven by algorithmic recommendations that are constantly trying to guess what we might like and put the next piece of content in front of us based on what we've consumed before. I personally felt totally enclosed by this kind of sphere almost of algorithms, and I couldn't find something or listen to something without facing that surveillance and recommendation of what I was doing.

Micah Loewinger: I want to dig into some examples of this feeling of being boxed in by algorithms at the same time as feeling that they provide us with the things that will fill our time and our hearts, TV shows, movies, albums. Let's talk about Netflix, for instance. When I open up the app on my TV or laptop, it feels like I'm being given a wide range of shows and movies tailored to me, but what's really happening there.

Kyle Chayka: The homepage is supposed to be a thing that reflects your taste and filters through things that you're going to like. More often these categories are so broad and the kind of labels are so vague that they don't actually promise personalization. There's like a top 10 or there's a popular right now, and those shows are just what's convenient for Netflix to promote at a given time, what's popular with a certain segment of the users, and what they can most conveniently convince you to watch in a way. Like Netflix has this algorithmic system to change the thumbnail of a show.

Micah Loewinger: Yes, this is so creepy.

Kyle Chayka: [laughs] When you go on Netflix, the images of every TV show and movie are tailored to your preferences.

Micah Loewinger: For instance, in 2018, there was like a controversy where a bunch of people were being promoted the film Love actually, a pretty safe film to promote to a lot of people. Very popular. It turns out some people were being recommended with the prominent imagery of the Black actor Chiwetel Ejiofor, who plays only a minor part in the film.

Kyle Chayka: It's so manipulative. If I, the Netflix algorithm, know that you watch a lot of movies with Black actors, then I am going to present every movie as if it focuses on Black actors. In the case of Love Actually, which absolutely does not focus on Black actors, I will highlight one of the few scenes that has this man in it in an effort to get you to watch it. Not because you definitely like Love Actually. Not because you're gonna love Hugh Grant dancing through the halls of the government or whatever, but because it would be convenient for Netflix if you watched this movie.

Micah Loewinger: That's an egregious example of the bait and switch. I want to talk about another theme in your book. You're talking about something slightly more pernicious, which is a recommendation algorithm like Spotify's, that in maybe the largest library of legal music ever created, I am subtly encouraged to listen to the same stuff that I like over and over. How is that happening?

Kyle Chayka: There are a lot of knobs and variables that can change in these formulas. For Spotify, if you put on an album and then let it go, usually within a few songs, I think it serves you up. Something that you listen to constantly that it knows you are not going to turn off in order to lull you into that hypnotic state of just listening to the infinite playlist and not thinking about it.

Micah Loewinger: This gets into what you want out of a listening experience or what you want out of a library of culture. My own sort of leaps forward in music curiosity has come from listening to the radio. Big shock. I work for a public radio show and I really like radio, but I think of WFMU, the independent radio station in New Jersey, or WKCR, Columbia's radio station. I remember for the first time hearing the Indian music show and hearing a 30-minute raga, and then somebody explained why it was interesting at the end.

I had never encountered that kind of music before. There is something that you argue in your book that is lost when we take curation away from humans.

Kyle Chayka: Yes, human curation. And that idea of a DJ, a human person who has selected this raga, and even though it's 30 minutes, that person is like, "You are going to like this. It's important." That's such a different encounter with a piece of culture than what you get on Spotify or what you get in a YouTube recommendation. The job of human curators, like a DJ, like a museum curator or a librarian, is to build meaning through juxtaposition and then guide the consumer into it in a way that helps them kind of broaden their own horizons, as you said, or brings them to a new place of taste or thought.

We just don't get that from a machine.

Micah Loewinger: That said, I'm sure listeners right now are like, "But Discover Weekly has delivered some great stuff to me or I keep a close eye on some of the high profile Spotify playlists that are curated by humans." This is not a pure either or. Right?

Kyle Chayka: The Internet is not the same thing as algorithms. There are many digital platforms that are not algorithmic. There are also ways of using Spotify that are not guided by algorithms. We can't blame algorithms so much, like they fulfill a really important function in sorting information but I think we can take back some of our agency in comparing past.

Micah Loewinger: Ways of consuming music, say like through the radio, to what you call filter world, we do run the risk of being overly nostalgic. The tastemakers of old, the radio DJs, the record store clerks, the critics, they had their own blind spots and biases. DJs of top 40 radio stations were swayed by money, pressure from labels, whatever the public at large they thought would respond to, that's not exactly for the pure love of music. Right?

Kyle Chayka: The old algorithms were human gatekeepers who made decisions about what culture should be promoted and what shouldn't. Magazine editors, record label executives, the DJs who might be influenced by payola. I don't think that's inherently good. I do think in the best examples like an indie radio DJ who's not overseen by corporate overlords, that can create really beautiful moments of curation and the transmission of culture but so can a YouTube recommendation. I've gotten really interesting stuff from a YouTube recommendation that I wouldn't have known a person who could give me and I wouldn't have known to seek it out.

Micah Loewinger: Give me an example of the algorithm serving up something that got around the calcified biases of the old gatekeepers.

Kyle Chayka: An example of something that I personally like is the Japanese genre of city pop, which was this kind of music that was made in the '70s and '80s mostly. It's this very ebullient R&B, big orchestra, propulsive beats, big bold, crazy music, and it's really fantastic. It was hidden away for a long time. Japanese people were not listening to it much after the '80s, and then in the 2000s some record DJs brought it up and then it hit YouTube where it just blew up because for some reason it worked for the recommendation engine.

A lot of people were listening to this music, they were liking it, they were engaging with it, they were seeking out more of it. YouTube registered that this music was getting popular with an American audience long before a record label executive could do anything or even a radio DJ. It was a kind of democratic revival of the genre of music online, which I think is really cool.

Micah Loewinger: With the so-called democratization promised by social media, is amplification and all of the problems that it introduces. Algorithms picking things up to go viral that otherwise might not have, and that any regulation of algorithms which you explore in your book should mandate greater transparency around what gets pushed into people's feeds. Tell me a little bit about why we should regulate algorithms and what you see as the potential avenues for that.

Kyle Chayka: Right now there are essentially no rules about what an algorithmic feed can recommend to you or how it can interact with you. You can regulate what kinds of content gets algorithmically recommended. You could say that problematic content that promotes violence or self harm cannot be subject to an algorithmic recommendation. And if that was blanket illegal, as it may soon be in the European Union, then social networks would be much less likely to even touch that kind of material in its feed. All of a sudden you could only find that stuff if you opted into it, it would not get pushed out to more people.

There's regulation about what kinds of content can be recommended or promoted. There's regulation around transparency for algorithmic feeds, which means that we could see how something works and know what variables are being taken into account when something is promoted to us and there's regulation that mandates you be able to opt out of recommendation.

Micah Loewinger: Are they likely to pass and be implemented?

Kyle Chayka: In the European Union, they have passed the General Data Protection Regulation, which has caused that wave of pop ups that say, "Please let me give you cookies." The Digital Services act more recently, which does mandate things like algorithmic transparency and opting out of feeds. In the US we're way, way, way behind that. Some of these companies like Facebook are changing how their feeds work based on the European regulations but in the US we don't actually have any of those rules. And the few efforts that have been made in government have just not gotten very far at all.

Micah Loewinger: In your conclusion, you acknowledged that the intersection between art and culture and technology has always been fraught. Cameras and radios sparked fear, so did the telephone. In fact, you quoted one Japanese novelist who was lamenting what was lost when street lights were introduced in Tokyo at the late 1800s and early 1900s. Are algorithms fundamentally upending how our world works? Or is this part of a larger fear we have about change?

Kyle Chayka: We do always fear what technology does to culture. Culture is threatened by a thing like the camera, like recorded music, like the radio. Then artists find a way to carry on and make great things, and then we adapt and reframe our idea. No one's going to say that recorded music is a sin or that we should go back to only live music because that would be more authentic. There's no pure culture but I do think pendulums swing. We've gone so far into this algorithmic ecosystem that I think we desire to retreat from it a little bit.

The same way that we had to make up regulations for seat belts and car safety rather than people flooring it down the road and having no safety checks in place. The financial exchange is so different to how art is sustained. Before, a musician would sell you their album and they would make money. Now it's mediated by this huge algorithmic platform of Spotify and they only make money based on certain metrics, based on streams. I think one way to retreat from that and to go back a little ways is just finding ways to directly support the voices that you like.

A designer, even a curator or a DJ who makes cool playlists. The best way we can ensure the survival of those kinds of relationships is just to pay them money. It's more expensive than a Spotify subscription. You're not going to get an infinity of music, but getting that infinity of music for $10 a month means that musicians have a really hard time making a living. Even though it's nice to be on the TikTok feed and see who you like to see, that's ultimately a hard way for your favorite creators of whatever to make a living.

Micah Loewinger: Kyle, thank you very much.

Kyle Chayka: Thank you for having me.

Micah Loewinger: Kyle Chayka is a staff writer for the New Yorker. His latest book is Filter World: How Algorithms Flattened Culture.

Brooke Gladstone: Coming up, the AI robots are coming for your music.

Micah Loewinger: This is On the Media. This is On the Media. I'm Micah Loewinger.

Brooke Gladstone: I'm Brooke Gladstone. We're sticking with the music theme for the rest of the show, but we're turning the focus away from listeners to makers. This comes to us from a former OTM producer, Mark Henry Phillips. You may not remember his name, but you'll probably be familiar with his work.

[MUSIC]

Brooke Gladstone: Yep, that infamous jingle was a Mark creation. When he left the show 14 years ago, it was to pursue a career in music, and it kind of worked out. He started scoring films and making commercials for clients like Google, PayPal, and Ford. He's also made music for podcasts like Serial Startup, This American Life, and Homecoming. As he says, he's not a rock star, but he's made a decent living making music, until now. Over the past few months, Mark has become terrified of and fascinated by AI music generators, particularly one that launched in April of this year.

When he first encountered it, Mark realized his work would never be the same.

Mark Henry Phillips: I know this sounds dramatic, but the first time I played around with this AI music generator, it caused an existential crisis. I literally lost sleep staring at my ceiling in the middle of the night wondering, "Will this be my last year making money as a musician?" Okay, what happened? I stumbled across a track by a user named Man or Monster. He was trying to recreate a Toots and the Maytals track and had typed in some lyrics along with Soulful Reggae, Ska, 1969 Hammond organ male vocalist and out popped this.

[MUSIC]

Mark Henry Phillips: I don't want to admit it, but it's good. Like, really good. It doesn't feel like it was created by AI in a data center somewhere. It feels like it was made in a makeshift studio in Jamaica on a hot summer night in 1969. It feels real. Of course, this brings up all sorts of copyright issues because this sounds a lot like Toots, also known as Frederick Hibbert, also known as a real live human being. I would wager to bet that sounds that good because Toot's entire catalog was used in this AI music generator's training data.

Let's ignore that for now. Instead, let's just focus on how good this sounds.

[MUSIC]

Mark Henry Phillips: Girl is not a good one for you. I know that you must feel it too. AI generated images often have glitches like pictures of people often have extra fingers. In AI text, it also has its subtle tells that make it feel not quite human, but this track, it might not be perfect, but it feels like real music. If it feels real, that has big implications. Like what, you ask? Let me give you a hypothetical. You're an ad exec making a beer commercial. You want a track that sounds like it was recorded in Jamaica in 1969, but you don't want to deal with all the money and legal back and forth that comes with light licensing a vintage track.

So you turn to a composer like me, you pay me. Not as much as you'd pay Toots or Desmond Dekker, but still enough that it'd be worth my while. Boom. You get a commercial and I have a job. Years ago, that's exactly what happened to me. I was hired to make a song that sounds like an early rock steady tune from 1969 Jamaica. Essentially, I was given the same prompt as that AI track, and here's what I came up with.

[MUSIC]

Mark Henry Phillips: It took me a couple days, and it's pretty good, fine, mediocre. It's a little embarrassing to play it now, and I was actually going to take the vocals out because that part is super embarrassing, but if we're going to do a comparison, it's kind of relevant. So here they are.

[MUSIC]

Whatever you think of my track, the vocals on the AI version just sound so much better, in that legally and morally dubious type of way. Don't get me wrong, there are many musicians out there who could have produced something way better than me or the AI but the vocals, here's that AI generated track again.

[MUSIC]

Mark Henry Phillips: To get something like this, you'd kind of have to hire or be an amazing Jamaican legend with a group of amazing backup singers. Even if you were, you know, Toots himself, you couldn't write, produce, record, and mix a track in under a minute. That's the crazy thing. Our hypothetical ad exec can now make this track. Hell, 10 tracks like this in under five minutes, and it's free. See why I was having that existential crisis? A lot has changed over the past decade, and the money has been on a steady trajectory downward, but now, post AI, I think a lot of composers are going to be without a job, including me.

If you've messed around with these services, you might be thinking, "They're interesting, but they're not that good without, what's this dude talking about?" but you're probably not using it to make the same stuff I make. AI isn't going to replace Billie Eilish or Radiohead, at least not for a while, but my bread and butter theme songs, commercial music score, yes, it could replace me, like today. I don't think that's necessarily true yet with other AI products like ChatGPT or Dall-E. They're definitely cool or scary, depending on your orientation, but they're not really replacing professional humans quite yet.

ChatGPT isn't writing big ad campaigns, Midjourney isn't replacing photo spreads in magazines, but music, it's different from those other mediums. I think it's interesting to think about why. I was always taught that the golden rule of music is if it sounds good, it's good. Sounds pretty basic, but it kind of has big implications. For the listener, that means you don't have to understand how a song was made to enjoy it. It's a black box process. Most people have absolutely no idea how their favorite songs were made, and it doesn't matter.

If it turns out good and it's good, but here's the weirdest part. For the musician, the process of writing music is also a little bit of a black box. A musician is never that conscious of what they're doing. You have a vague idea or a goal or a feeling, but you just mess around and discover the song. At a certain point, even if you're super cognizant of what scale and mode you're in and what the chords are doing and what is common in this style of music, at a certain point, you really just have to take a guess at what the next note is.

When I have a job, sometimes I'm really explicit with what I need to do. I'll say to myself, "Okay, this is a scene where a couple is getting to know each other. This needs to be something like a Jon Brion Score from 1997. Something a little goofy and a little romantic." Jon Brion, if you don't know, is an amazing composer and producer who scored films like Magnolia, Eternal Sunshine,and I Heart Huckabees. Anyway, I don't actually go back and listen to one of Jon Brion's pieces and dissect the chord progression or the instrumentation, and I don't find just one piece to rip off.

I just get a vague notion in my head and I start playing around and then something like this comes out.

[MUSIC]

Mark Henry Phillips: It might not be amazing, but it hit the mark for the show I was working on, and importantly, I don't think it's a ripoff of Jon Brion. I had a good target in my head and I didn't get a perfect bullseye but that's a good thing because it means I came up with something new and that's really how music works. I think this is a really key point. Take the Beatles, they were trying to sound just like Buddy Holly and Elvis and their failed attempts became the Beatles. That's just how music evolves. The point is, I can't tell you exactly how I wrote this song.

Even though this was a pretty conscious for hire process. Even this writing and recording a song for a show, it's a black box process for me. That's why these AI music services are so good. It approaches music creation in the exact same way. It doesn't consciously say, "Write a four chord progression in a melody in a mixolydian mode." No. Instead it creates a Jon Brion-esque track just how I would. It's listened to all of his music and then using neural networks, whatever those are, it has a fuzzy image of a Jon Brion vibe. I don't know exactly how it works, but I think it's just always guessing what the next sound's going to be.

On one level that means guessing the next note in the melody, but on another level that means guessing what instrument should come in next or guessing what the next cool production trick should be to keep the track interesting. In other words, it's approaching it just how I would. Here's the AI music service doing it for a Jon Brion track.

[MUSIC]

Its process is remarkably similar to a human musician. And that's why the results are too. When ChatGPT is "writing", it's also just guessing what should come next, but that's very different from actual writing. A writer knows what they're writing, but musicians, we don't fully know what we're creating, so the process is much easier for AI to replicate. This is why I think the next version of one of these AI music generators will replace me very soon. Let's take a project I'm working on right now. I'm scoring a show that's Hitchcock ass. Hitchcockian.

The show involves trains and even though Strangers on a Train was scored by Dimitri Tiomkin, I obviously thought of Bernard Herrmann, who scored a bunch of other Hitchcock movies like Psycho and Vertigo, and who was friggin awesome. Before I sat down at the piano and tried to figure something out, I thought I'd do a little experiment. I logged on to the AI music generator, typed in Bernard Herman theme, Alfred Hitchcock film Train Mysterious, and this is what I got.

[MUSIC]

To my ear, that's really good. Scary good. I keep waiting for it to break down and do something uncanny and weird, but it never does. Of course I guided it as it made the track, but I really didn't have to do much. It was kind of a choose your own adventure between really good options. This is the track that really scared the crap out of me. I could never in a million years write and record something this good. Even if I could, I would need to spend three times the entire budget of the project to hire a musical arranger, a real orchestra and a recording studio.

All which would be necessary to create something that sounds this close to a Bernard Herrmann score but AI it did it in five minutes for free. And this is where we flesh puppet musicians just can't compete. None of us are experts in every style. I might be able to beat it at an indie pop tune, but I can't do a better debut piece and a better medieval choral hymn, and certainly can't do a better Bernard Herrmann track. Plus it's a virtuoso at every instrument. Violin, piano, vocals, guitar. Yes, in the vibes department, I might have an edge, but also maybe not.

Those examples were really good. That's what's so unmooring about this AI thing for me. It's not just the loss of work. It's part of my identity. It was my thing. You could give me a commercial or a film or a podcast and I could make a song for it. It's the thing that made me a little bit special. When I tell someone I'm a musician, a lot of times their eyes light up. They think it's cool. Some even ask me, "How do you just write a song?" But now with AI, anyone can write a song. My special skill just isn't that special anymore. From a musical and economic point of view, AI just has me beat.

This is where I was this summer, freaking out. I picked up mountain biking to distract myself. I thought about starting a granola business but then something happened. For me, it changed everything. The AI music generator I'd been using added a new feature allowing you to upload your own music. The idea was that it could listen to what you were doing and just extend it. To test it out, I uploaded a 12 second jingle I made a decade ago for a commercial I pitched and didn't land. This is what I made back then.

[MUSIC]

Nothing special, but I always thought that could have been the start of something.

[MUSIC]

I uploaded it and with a little prompting, it turned into this. It starts exactly the same, but then completely, seamlessly, it keeps going. This is way more mind blowing than the tracks I heard it make from scratch, because it doesn't feel like some other musician, it feels like me. This track isn't mine in the traditional sense, but because it grew out of a seed I planted, it really does feel like something I would have done if I wanted to turn this track into a full length song. Like the percussion that just came in, I totally was imagining a vintage CR78 drum machine loop.

That's exactly what the AI put in. I know, I know. This is like someone saying they could have invented Instagram or could have invested in Nvidia before the stock skyrocketed. Yes, dude, you could have, but you didn't. So maybe this is just me tricking myself into thinking this is my music, even though it isn't. The more I played with it, the more exciting it was. Take this track.

[MUSIC]

This was a little demo I made to test out a new guitar amp. I don't even remember making it, but I stumbled across it and thought, "It sounds pretty cool." What you're hearing right now is still just me, what I originally recorded, but I uploaded it to the AI music generator and then prompted it to do a horn arrangement. Of course it obliged and it's switching to the AI version right now.

[MUSIC]

But why stop there? I messed around with the prompting and had it come up with another version, and it popped that out in just a few seconds. That version starts now. Also amazing. Again, it really sounds like where I was imagining taking the song next, and yet I would never feel comfortable using the audio it's generating as my own work. That feels like a bridge too far, but I could take bits and pieces of what it's done as jumping off points. This is where it gets so exciting. If I'm stuck on a song, let's say I can't figure out a new section.

I could have the AI music generator come up with 10 new choruses for me. I could take elements from different versions. A tempo change from one, a chord change from another, a bass line part from yet another version. Of course, I would play it all my myself. Since I'm just taking bits and pieces from different versions and putting my own spin on it while I play it, it wouldn't sound like any one version that the AI created. When it comes to ownership, this way of doing things would make it much, much murkier. It's not an AI creation, but it's not entirely mine either.

As murky as it is, it's the really appealing way to bring AI into my workflow. It'd be like having the best music writing partner ever. They're always awake. They're fast, enthusiastic and good. I have tons of unfinished songs. Some are produced tracks I couldn't quite finish. Some are like these 15 second samples I did for commercials and didn't get made. Others are voice memos. With all of them, I didn't finish the track because I got stuck. I think every musician, maybe every creative person, can relate to this.

It's easy to come up with the germ of something cool, but it's so hard to get it from 70% done to 100% finished, but now I have no excuse for leaving a song unfinished. I've never been more unsure of my future as a professional musician but as much as it pains me to say this, I haven't been so excited to make music in a long, long time. I have so many songs, so many projects that I feel like I can finish now, and that's really exciting. As weird as it feels to me using AI as a co-writer, I think young musicians coming up now could lean on these tools.

Just like I grew up with spell checks baked into Microsoft Word. Yes, there will definitely be holdouts, but as a society, it'll become the norm, both for musicians and the listeners. I can't help but see this as a weird fork in the evolution of music. Like music was pure up until 2023, but from here on out, it won't be because AI music is already in the water. All future models will will be trained on music that was made with AI this could cause feedback loops in the algorithms, a proverbial AI snake eating its own tail. It could get really weird.

This might be the last year I make money as a musician. If that's the case, it'll free up time for me to finish the half dozen unfinished albums I have. I'm excited to actually do them, using AI as a writing partner. Yes, that brings up all sorts of moral and legal issues, so I probably won't release them. I'll just make them because it's fun. Isn't that what making music is all about? For On the Media, I'm Mark Henry Phillips.

Micah Loewinger: That's it for this week's show. On the Media is produced by Molly Rosen, Rebecca Clark-Callender, Candice Wang, and Katerina Barton.

Brooke Gladstone: Our technical director is Jennifer Munson. Our engineer is Brendan Dalton. Eloise Blondiau is our senior producer, and our executive producer is Katya Rogers. On the Media is a production of WNYC Studios. I'm Brooke Gladstone.

Micah Loewinger: I'm Micah Loewinger.

 

Copyright © 2024 New York Public Radio. All rights reserved. Visit our website terms of use at www.wnyc.org for further information.

New York Public Radio transcripts are created on a rush deadline, often by contractors. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of New York Public Radio’s programming is the audio record.