The Music & AI Journey: An Update

Someone actually got inspired by AI? How can that be?

If you’re a regular reader here you’ll know I’ve been dabbling with the musicAI tools in Suno. Some of the results from that process are on YouTube here, and some I’ve yet to share more widely, although you can listen to some of that second-tier collection in this post.

Today I’d like to offer results from my intention for this music.

My Intention

I’m pulled in many directions with the musicAI thing. My primary intention (from the post cited first above) was this:

---

The hoped-for result I have in mind is in two parts:

1. Musicians and creators have new grist for their mills;

2. Non-musicians discover how they, too, can write music that’s pretty good and definitely satisfying!

If either of those things happen, the ripples created could have a profound effect on music as we know it.

---

To directly address the 800-pound gorilla in the room, here are a few of the louder ways that this intention encounters resistance in real life:

MusicAI takes jobs from musicians

Yes: this is one effect that musicAI can have, and there is evidence that music streaming services are favoring AI-created music over organic music. Since I’m not doing that — all “my” published AI music is in the creative commons (more about that in a bit) which means anyone is free to use any or all of it without attribution or payment to me — I want to address everyone’s concern about AI stealing jobs from real musicians and songwriters.

Those of us who can’t afford to hire a band that can play in multiple genres instantly without rehearsal aren’t really impacted by this, are we, especially if we are using musicAI to rapid-prototype ideas that can become fodder for some human-created song? I’d love to be able to employ a team of ultra-competent musicians and singers who could quickly create eight different settings of, say, Shakespeare’s Sonnet 116, one of the greatest love poems of all time.

However, at this stage, I don’t have the resources to hire a band. I do, however, have SunoAI, and, in less time that it takes to write this, here are four completely different settings for that famous sonnet (Suno made eight of them, these are my favorites so far):

Hip-Hop

Cabaret

Country/Folk

Big Band

What will I do with these? Probably keep them under my hat for inspiration at some future date. The purpose here is to fire my imagination, not sell or license something that somehow harms working musicians — like me. Why? Because I have a commitment to authenticity — my solo recordings are one-take, unedited, and from the heart, such as this album, the actual notes of which I’ve never played again since the day they were set down for the first time.

AI Music Isn’t Any Good

This subjective statement requires the follow-up question: compared to what? There’s plenty of truly awful human-created, organic music out there, and some of it has made its way to the music streamers and actually has a paid following. I won’t bore you with examples!

There is most likely a place for AI-created music in genres where listeners may not be able to tell the difference such as ambient, screamo, or ultra-progressive jazz or rock. AI truly fails miserably in the Classical genre. One might make the case for AI being “better” than many organic songwriters in the more popular genres, however that, too, is subjective.

One way in which AI is quite useful in the music “quality” space is that it can establish a sort of “minimally good music” against which other music can be compared. Since the statement “AI is always derivative” is a subject of some debate these days, the jury on its quality within the music realm is still out. Any of you who have tried to do creative stuff with AI will most likely have encountered its limits in many ways, and those limits also exist within the music AI space — there are times when SunoAI hits the wall in ways that are actually comical.

The built-in creative problem with music these days is that a tug of war exists between music that will sell millions of copies and music that exists simply to extend the novelty of what music can do. One could say that, were Taylor Swift to write another hit song, it could be very much like what’s come before in her catalog as well as others where writing pop hits is the goal. This same argument could be transported backward in history by simply replacing Taylor’s name with Jo Bach, of Viggie Beethoven, or Mozart, or Mahler, Berlioz, Debussy, or even the Johns (Adam and Cage). No less than René Girard have studied how ideas arrive and propagate over time, and the Girardian theories of novelty are grounded in more than two millennia of civilization’s creative output.

So let’s leave the “any good” point aside for now as somewhat unsettled, and challenge creators to just make good art (as someone once said) and let listeners decide, with or without bias, until we have a more objective way to describe what it actually “good” music.

As I reflect on the clips I shared above, I don’t feel that any of them are particularly “good” — they don’t often have cohesive melodic ideas, suffer from musical limitations that make them rather trite, and need more development to make them into full-length songs. To me, that makes them less than good, but definitely not bad. What say you?

AI Music Can’t Be Sold Because It’s Not Of Human Origin

We’ve already covered this a bit based on listener response. Let’s look at it from a more granular point of view. The argument here is one that creators face daily right now. Whether it’s a speech, a song, or a video, the creative world is facing the ethical questions of how to disclose the contributions of AI to the final product.

This video is quite useful on sorting out the legalities of AI-assisted creation in light of whether the ultimate created thing can be legally sold, and with what level of disclosure.

May I share an example?

It turns out that one of the AI settings for 23rd Psalm from a month or so ago caught the ear of a friend of mine, Pastor David Brisbin of the effect faith community. Full disclosure, I play in the praise band at the effect. Dave heard this AI setting (just to put it back into your ears):

He was inspired to set 23rd Psalm to his own music, which I would love to share with you here, but it’s his music. Instead, take another listen to the above track from 2:31 to the end, because there’s a six-second melody and chord progression that Dave used in his setting. Here are a couple clips of how that sounded (live recording from the performance at church so please be kind!):

Can you hear the AI melody on the words “Surely your goodness and love will follow me” and “And I will live in the house of the Lord forever?”

Dave’s song is pretty good IMHO. He would like to try to sell it to a recording artist in the Christian music scene. Can he do that legally? How does he “credit” me or should he?

These kinds of questions are the ones we need to be asking now. Legally, the framework we have does offer protections, and it is actually up to creators (like me and Dave) to decide how to “credit” the sources of our published works. You can bet Dave and I have talked about this!

In the music space, it is now expected that composers with disclose the effect that AI has had on a particular finished work. In this case, for example, Dave might market his song — which he calls “All My Days” — with the disclosure that six seconds of the chorus melody were “created” by AI. That disclosure, however, would be novel, since many pop and (Classical!) composers “steal” melodic ideas from one another without attribution. It’s a very intriguing thing, for composers to attempt to locate the source of their musical ideas within a particular pre-existing song.

How would you respond to that challenge? There are many record producers and amateurs whose musical memory is so apt that they can listen to s jazz solo, for example, and tell you exactly whom the soloist is quoting and from what other recording. Is that legally actionable? Is it more salable because a human being did it? The debate on this issue may go on for a while.

Back to My Intention

First, I wanted to give creators a new kind of grist for the mill. With Dave, at least, I think that happened. Dave and I will collaborate on future projects, one of which is in the works right now, and I look forward to sharing that process with you here.

Second, I wanted to offer non-musicians a way to begin creating music. Why? Well, the process — even the old-fashioned way — is cathartic. I can tell whether a SunoAI song “works” for me in a very visceral way, and that kind of awe from a machine is something that I find remarkable.

Therefore, I want to encourage you to create.

For me, being able to almost-effortlessly set words to music has opened up new meaning in the words, as I’ve already shown you with 23rd Psalm. On the other hand, writing a poem or essay to process an idea and then setting that writing to music? There’s something to that! While I can offer you my experience, my hope is that you’ll try this for yourself.

Here’s a recent example of music that worked in this way for me.

The Seven-Day Poetry Challenge

Recently, I signed up for this. I had no idea what to expect or what I’d write. It turns out, perhaps not surprisingly, that I wrote lyrics. The words seem to describe a marathon runner in a race. There’s a darkness about being just another runner in the pack on a seemingly useless quest to better their time, and you can watch and listen as they try to make their way through all 26+ miles.

For me, this poem was an investigation of what it means to work very hard at something for a very long time only to see incremental improvement if any. Here’s the finished work — see how it feels to you:

Unlike journaling or writing poetry, setting the words to music unlocks insight and emotion that may not have been available in the words themselves. I don’t want to say too much more about that just now (there will be more in future posts) except to invite you to try it for yourself.

Maybe you “set” lyrics that already exist, like I did with 23rd Psalm. Maybe you choose an existing song, pull the lyrics, and see what AI does with them. Maybe you cut and paste a journal entry into Suno and see how it sounds in different musical genres. However you go about dipping a toe into this pool, I can assure you that the water’s fine, and that you may be moved by what happens when you’re in it.

Maybe you’ll get something that has commercial viability; maybe not. I encourage you to go at it without that intention, and take this invitation as an experiment in your own music-informed personal development. It costs nothing except your time to give it a try.

Let us know how it goes, especially if you feel comfortable sharing what happened (SunoAI is very capable of doing downloads, both audio and video, and I hope you’ll post yours in the comments below).

Again, because this is NOT about making money from your personal development, there are NO issues about taking jobs from musicians, making anything that’s “good,” or making something that can be sold. Think of this as fingerpainting or collaging for your own delight and edification, maybe with some emotional inspiration thrown in. If you remember that the “result” (if there is one) is all about your own personal experience, you will free yourself from all that AI anxiety and start to enjoy the process.

“Make good art” because it’s good for you, and for no other reason, and you’ll begin to taste the sweetness that composers from all genres and ages get to experience, too.

Bill Protzmann has rediscovered the fundamental nature and purpose of music and accumulated a vast awareness of anthropology and sociology, as well as the effects of music, the arts, and information technology on human beings. Bill has experimented with what he has learned through performing concerts, giving lectures, facilitating workshops, and teaching classes. He first published on the powerful extensibility of music into the business realm in 2006 (here and abstract here). Ten years later, in 2016, he consolidated his work into the Musimorphic Quest. In this guided, gamified, experiential environment, participants discover and remember their innate connection to this ancient transformative technology. The National Council for Behavioral Healthcare recognized Bill in 2014 with an Inspiring Hope award for Artistic Expression, the industry equivalent of winning an Oscar.

In addition to individuals, Musimorphic programs support personal and professional development and wellness for businesses, NPOs and at-risk populations.