Many artists have felt a sense of trepidation regarding their futures when confronted with artificial intelligence. It may pose a serious threat to sectors of the music industry. The following is a list of thoughts that guide how I counsel my students:

How does AI work? It takes the average of everything relevant to a topic and presents it in a linguistically natural form. ChatGPT does this with words, and Suno (among other music apps) does this with sound. Every AI platform uses models trained on core topics help define basic protocols. So, AI pulls from millions of rock tracks to model what rock sounds like. The result will be the average rock song. Likewise, lyrics from these songs will be average lyrics based on how rock song lyrics normally go mixed with whatever content the user desires. AI creates average music.
When do people listen to average music? This is the kind of music played in department stores, in the background of sitcoms and advertisements, and in low budget media (movies, TV shows, video games). This is also the kind of music that people fall asleep to and appears on mood playlists. These are passive listening experiences, where the attention of the listener is not on the music but elsewhere. If your music falls into one of these categories, then you might be in serious trouble within the next few years.
How does one write better than AI? Because AI writes average music in the most literal sense, musicians will be forced to write music that is not average, even above average. Music above average has depth and craft. It has the ability to set up conventions (like AI does) but then stretch them or even break them. Even more challenging, it can create a sense of anticipation and carefully increases the anticipation until a climax. AI is especially good at writing music that goes nowhere because the challenges of creating and sustaining anticipation through playing with conventions require a new solution in every case. The only drawback to such music is that it requires an active listening experience. This sort of music requires a bit more investment from the listener (or can be paired with another artform).
What if AI gets better? It will. And one day, AI will write “better” than us all. But wait a second, what does that even mean? While we can all get better at our craft, we can all write good music with some training and deliberation. There comes a point where professionals write good music, and comparison is more about taste. Then, there’s the human factor. Music written by humans is going to be about the human experience and done so in a way that is unique to the artist. Also, there is value in imperfection and spontaneity. The power of improvisation comes from its improvised nature, and every performance has a little bit of its own adventure built into it.
What if listeners don’t care about music written by composers anymore? They will. Perhaps not everyone will care. Those people who don’t care are the same people who already ignore what musicians do. The people who do care about music and invest in it are looking for something more and will find it in human music. Where will they find it?
1. Live performances. Write music in a way that gets someone off the couch, into a car, and to a venue. This requires careful concert curation, planning the concert with as much thought as one might curate a visual art gallery.
2. Artistically-minded media. Write music that is married to other artforms. Film music should be so intricately cued and orchestrated that it can’t be replaced by something else. This isn’t a new phenomena because film directors already use older music in their work. Video game music should have such a distinctive sound that it serves as the signature of the game. Again, this is already a well-known fact. People seek out media with good music and rave about it (probably in video game world even more than in films these days).
3. Improvisation. Write music that incorporates elements that change every time.
4. Site-specific music. Sound installations are powerful because they are tied to a location. The music acts as a commentary to the place or transforms an otherwise normal space into something extraordinary.
5. Anywhere, as long as it’s human. People do enjoy connecting with music, but they equally like connecting with the human behind the music. So, be that human. Be friends with the people who are interested in what you do. Focus on building good and healthy relationships with everyone in music. Don’t change who you are for others (well, unless you insult others or have a serious chip on your shoulder), and you’ll find that good things tend to happen. People are looking to hear music from good, authentic people, real people.

So, I’m actually excited about what AI means for music. It means that composing means something more than the status quo. We might be entering an age where human-made music is valued more than ever, and this means that listeners will look for artists who have compelling craft. Not everyone will care about active listening, and they never really cared before AI anyhow. The people who do care will keep on looking for it, and they’ll find what they’re looking for just as much as they did before. So, if you’re writing thousands of stock tracks for download or streaming, you might need to change your game in the near future. Otherwise, the future is bright if you’re willing to continually improve your craft.

I deactivated my Facebook account yesterday after having used it for about 14 years. Yes, I was one of those people who watched The Social Dilemma and was appalled at the evils done by the social media business model. I never quite understood how much my phone tracks about me (EVERYTHING) and how aggressive it markets based on our weaknesses. Because it is run by advertisers even more than its owners (the customer is always right…), people can sway populations with a big enough price tag. Myanmar used it for genocide against the very Rohingya population I lamented in The Story of Our Journey . Much of the violent rhetoric in the past several years only spread because of how social media targets users through high-emotion content. Political extremism and conspiracy theories have completely obliterated the confidence in truth, leading to our near inability to talk to people different than us. I can’t support such a destructive platform.

Some musicians have long felt threatened by the DIY aesthetic that rose out of YouTube. The amateur could become famous while the professional gets lost in the shadows. I believe that threat is healthy for musicians! If we can’t get people interested in our music, then we have to look at either our presentation of media or our music itself. Also, being “famous” has never been that appealing of an aspiration. A musician doesn’t need to get that one big famous video to subsist and thrive, for one-hit wonders on YouTube are not worth as much as the one-hit wonders were in the 90s. True recognition comes from years of experience, great connections, and true fans. The long-term famous YouTubers place their entire career on making videos. They live the life of the entrepreneur and put in more time than most would imagine. They aren’t my competition.

Perhaps social media’s greatest blessing and curse is its redefinition of musical choices based on emotion over genre. I’m still a bit mystified on how artificial intelligence “listens” to music and recommends artists fitting into a certain mood, but it’s actively happening and has been for years. If I listen to “Chill Thursday” or “Workout Fire,” then I’ll get a playlist that we’re told fits the mood (and we’d agree much of the time). The singular aesthetical assumption is that music’s purpose is to generate a mood that resonates with the listener. Younger acts lean into this assumption and genre-bend their work to fit into multiple categories (which is argued to naturally fit into the diverse tastes of the generation). When I write music, I think about the atmosphere or sound world, but that’s a starting point. If mood were the core of music, then it would be a shallow endeavor. Maybe that’s why we now compare music with temperatures; we regard music as important as changing the thermostat from 68 to 72 F.

But, I’ll be blunt: they have done an awful job at mood-mixing concert music idioms. The contemporary classical music is more-or-less okay, but it is mostly post-minimalist. Almost every other classical playlist has branded as Baby Mozart music. I typed in Classical on Amazon Music and received the following playlists: “Deep Sleep Music,” “Instrumental Lullabies,” “Putting the Baby to Bed,” “Bedtime Lullabies,” “Classical for Pets,” “Classical Focus,” “Classical Sleep,” “Relaxing Children’s Classical,” “Relaxing Classical,” “Classical for Meditation,” “Classical Slumber,” “Dream Time,” and, as a relief, “Fun Classical.” Ouch. Obviously the AI doesn’t even try to figure out the mood of classical music. Isn’t this supposed be the same brand of music that brought you the riots at the premiere of The Rite of Spring, the performers who painted their faces white and pretended they were possessed (Liszt, Paganini), the most psychologically disturbing works ever (Berg’s Wozzeck and Lulu), and the thrilling collection of works that has influenced every dramatic film composer from Day 1 (Holst’s The Planets, Strauss’s Salome)? Even live classical musicians have fallen into the traps set up by these “moods.” I can’t say how many times I have fallen asleep during concerts where “nice” musical interpretation supersedes the power within the harmonic structures at play. Oh, how I’d love to hear (post-COVID) something more dramatic at the orchestral stage (preferably a concert by living composers)!

As AI continues to brand people towards certain moods, I wonder how that will influence musical taste and exploration. The sudden access to everything is brilliant now, especially with the collective memory of music from the 50s on. But just like in social media, what are the consequences of AI silently driving our decisions? Will we be softly baited (even if over the next several decades) into our comfortable “Autumn Chill” niche and let the vibrant blues, reds, yellows, and greens in popular idioms, like classical music, turn into an unremarkable brown? Who actually stops to listen to a full musical piece anymore? Who actually stops for anything anymore? What caught my attention the most during The Social Dilemma was that the user thinks they are in control because the AI is programmed to hustle us unknowingly into every decision to keep us engaged. I feel like the prophetic call to avoid future tragedies, not only politically and socially but artistically, is to learn how to listen and make active decisions. Radicalism and violence will be abated as people act to serve as mediators, not instigators. Artistic depth, which we do still cherish now, will bring about some of the most fascinating music if we are willing to actively connect with it rather than push it to the back of our minds. Will you search for the most different piece of music you can find, turn off all distractions, and listen to its album (yes, the whole 45 minutes) in its entirety? As John Cage said, “If something is boring after two minutes, try it for four. If still boring, then eight. Then sixteen. Then thirty-two. Eventually one discovers that it is not boring at all.” Go for it.