Teach your kids prompt engineering instead of the guitar
Teach your kids prompt engineering instead of the guitar
No, seriously.
I've been on a side-quest of sorts with Spectral, so haven't written about ye good ol’ music industry in a while.
Often friends and readers will suggest topics I should write about, but as someone with shiny-object syndrome, usually I try not to let those suggestions sidetrack me.
This time however, I succumbed. The rise of generative AI is set to irreversibly change the music industry, and it's hard not to have an opinion on where we're all headed.
Note: Some of the thoughts expressed here might upset music industry purists, but I encourage all of you to do one thing before you read this essay. Go to Suno.ai, and create your first song. Trust me.
Alright, assuming you just did, here are my thoughts, starting with some of the more obvious takeaways:
Generative AI will democratize music creation and blur the line separating music creators (i.e. artists) and music consumers (i.e. fans).
Like most things, this will lead to desirable outcomes and undesirable outcomes.
What are some desirable outcomes?
The technical barriers to music creation will no longer exist -- creativity (and originality, whatever that means in a post-AI world) will be the true differentiating factors. This need not mean "musical" creativity - it could mean creative ways to capture attention via memes, tiktoks, wearing a giant chicken outfit outside a taylor swift gig, whatever. There was a time in history where only scribes possessed the technical ability to write. Today, most educated humans can, and I think most of us agree that it's been a net good for humanity. More people than ever before will be able to create music, growing the overall pie for everyone involved.
As the volume of music released will exponentially increase, the value of (live) performances, social capital, genuine fan connections, etc will rise. Physical experiences and tangible things will be premium. We might see a full-fledged inversion into smaller fan communities. The value of community / connection is inversely proportional to the volume of content.
Today, anything I create using Suno's paid version is owned by me. As we see a proliferation of AI-native music creators, we'll also see the acceleration of an ongoing trend - more artists will own the rights to their music. This is huge. Beyond the obvious benefits, decentralization of music IP means other startups aren't beholden to the major labels (as much). This shift albeit small at first, could end up becoming really meaningful with time.
Let's talk about some undesirable outcomes:
With such a high volume of releases, distribution will be key. For smaller high value fan communities to form, artists need to be discovered first - and we may have gatekeepers and/or highly skewed power laws preventing that from happening. Very important to note here that I'm not advocating for equality of outcomes, just equality of opportunity. Everyone should be able to fairly compete in the attention economy.
People working in the music industry overestimate how much people care. They write eloquent essays about how sacred artist-fan connections are, and how artistic expression is the only virtuous thing worth striving for. The reality is most people don't care. Most people see music as a form of entertainment, and will be oblivious / indifferent towards how generic music sounds. These are folks who are okay being deep inside the Spotify auto-suggestion realm -- and don't mind listening to similar sounding shit over and over again.
Art and tech are going to be inextricably linked -- AI has shown us that creativity is not some weird ineffable alchemy that happens inside the creator's head. It's just math we haven't completely solved (yet). Believe me, it really hurts to say this, but I think it's best not to be in collective denial as an industry. Metaphorically speaking, the model and methodology underlying a product like Suno is like a kid in high school who has just learned basic statistics. Imagine when this same kid goes to Stanford to study advanced mathematics - it's a matter of when, not if. In fact, most mediocre human musicians will be wiped out before this kid has their first kiss on prom night. We often expect AI to be perfect, but remember - it doesn't have to be perfect. It just has to be better than most humans.
All this to say that it’s a challenging yet exciting time to be working in music right now—the industry's deep understanding of culture positions us well to leverage generative AI—only if we adopt it first.
Until next time,
Yash
Says it as it is! 👌
No, seriously don't.
We don't teach children guitar to make them future music stars, although that might be a positive side effect. We teach kids music because it is enjoyable, social, calming, challenging and more. Those things don't go away because there's now a short cut.
We're still teaching and learning music decades after the invention of the drum machine, the sequencer and the player piano. Why do we do this, if it's no longer necessary to actually play an instrument to make music? One major reason is because the actual musical output is only a small part of the human motivation to make music. See above for some others. I mean, people still play jazz (including me), and no one listens to that!
Prompt engineering, on the other hand, is a hyped up industry misnomer. It's not engineering. It's asking for things. And that's fine! Have at it.
But there's nothing anyone can teach to a child about prompt engineering that will be any use to them by the time they're 18. Maybe by that point they'll just have to think of a song to create it or something.
I find it strange that you put "memes, tiktoks, wearing a giant chicken outfit" in the desirable outcomes section. We live in a world filled with talents of all sorts - musical, analytical, comedic - but increasingly these are being "democratised" away, first by social media and now AI, in favour of the dual attractors of marketing and sales. A future in which these are the only remaining useful skills depresses me more than I can express.
Thankfully, music has been around longer than history and prompt engineering is only a couple of years old. Long term, I know where I'm putting my money.