Why people like Corey Feldman are a treasure to humankind

The unfortunately titled “Jump Back Black Cat (Gimme That Rock & Roll)” is, as I write this, the newest addition to the oeuvre of the notorious, the famed, the controversial, the almost universally panned musician – the one, the only: The Corey Feldman. I know this because I’m on a binge of Feldman’s music, and I’m having a fantastic time. Much, much better than you’d imagine if you’ve heard this stuff before. I mean, seriously.

You can argue that CF’s work is derivative. You can argue that it’s annoying. You can argue that it’s simply not pleasant to listen to. You can argue that it gives you severe, inexplicable, worryingly sharp and medical-feeling internal pain.

And even if I were to give you all of this as true,¹ I would still argue right back to you that Corey’s songs are a gift to the world. An invaluable addition to true reality.

Now I’ve got the even more haphazardly titled Angelic 2 The Core (Angelic Funkadelic / Angelic Rockadelic) playing front to back. It sounds like fucking shit. The man cannot hit a note to save his fucking life, and it sounds like he arranged the twenty-five-plus vocal layers as he sang them.² The songwriting is weak as a hungover Michael Cera. The backing tracks are limp as overcooked spaghetti. The message is inscrutable. By all accounts, it’s pretty fucking bad music.

But It’s Music.

Yes. That is my argument. This is very real music made by a very real person. He may not be very good at it, but he did it.

It may be bursting with flaws, but real people made it.

I am so, so, so fucking glad that Corey Feldman’s music exists. I hope he makes fifty more albums.

And I think everyone should agree with me.

Listen: We’re in an age where anyone, anywhere in the world with internet access, can type into their browser bar one of several web addresses, input a few words, and be rewarded with a full, professionally-arranged piece of music. This song may fool even avid music listeners³ as being of human provenance. Some may even be classified as “bops.” Some will rise to the top of the Sp*tify charts. Some will end up on Billboard. Someday, one may even win a Grammy. Who fucking knows?

And yet, AI music will always be a disappointment.

Why? Because we care about the people who make music. Think about it: Women passed out in Michael Jackson’s presence – but only when he was there. In their presence. Did they faint when they heard “Smooth Criminal” at the club?⁴ No? Why is that?

Because they were attracted – in some of several senses – to the person who made the music. The music was the beginning, but the music could not exist without the person, and the person would be – again, in some sense – irrelevant without the music.

I feel that some general theory of music enjoyment is in order. Allow me to philosophize:

How We Enjoy Music:

1. We hear music.
2. We feel the music.
3. We choose to engage psychologically with the music (listening to lyrics, analyzing guitar solos, etc.)
4. We re-listen to the music.
5. We choose to explore where the music came from.
6. We engage emotionally with the source of the music.
7. Depending on what we find in the source, our interest in the music is either a) amplified, or b) stymied.

AI music can fairly easily pass steps (1) and (2). It is simply sound, and humans are predisposed to feeling sound. It can, if it’s a particularly convincing generation, even pass steps (3) and (4).

The issues inherent in this medium, however, become apparent at (5). The direct source of the music is simply a website which created the song wholesale. The source is a block of code.

Now, some of the more skeptical amongst you may say, Then this is the end of the line. There’s no way anyone could ever engage with AI music in step (6) if it did not come from a source which can be engaged with emotionally, and therefore their interest cannot be amplified at step (7). This is where the listener drops off. And you have a pretty good argument there: There’s no human touch to analyze, no emotional stake by which to feel excited.

But that line of thinking is not comprehensive; it’s incomplete; it misses a major plot point of 2020s Earth.

The missing piece is: AI is novel. It’s A New Thing. And any New Thing holds emotional stakes for at least some people. After all, this is history in the making – it’s never happened before. If any given AI song could not have happened before the advent of this New Thing, then the music is, by definition, A New Thing itself.

This does hold emotional salience for many people – especially those who’ve always wanted to make music, but either couldn’t or didn’t put the effort in to create it themselves – and now see that anyone, anywhere can “make” music in just a few clicks.

What that mindset means is that the listener has a baked-in emotional investment to the person on the other side of all that code, the one who opened the website and typed in the prompts and uploaded the result via DSP to be consumed by the aforementioned listener. That person, the prompter, has something in common with the listener who views AI music as A (positive) New Thing.

From there, we can see an indirect link to steps (6) and (7). The listener who feels – via step (2) – the crux of the AI song can transport themselves to a mindstate in which the prompter is the one they engage with emotionally. In their minds, the prompter is just as responsible for the AI song as is the AI itself, if not moreso.

And yet, objectively, something is missing: The Human Component.

Yes, the prompter has some modicum of control over the AI music she⁵ creates: She has editorial control – she can choose to discard any of her numerous generations, she can tweak her prompt until something arises which is to her taste.

Yet re: the creation of the music, her fingerprints, are distinctly absent. She does not sing the melodies, nor does she choose the cadence to which the words are sung; she does not struggle to play a tricky guitar part, and she does not compromise the drum rhythm to something her skills encompass; she does not fight with the snare drum compression to find the groove she hears in her head; she does not have the chance to reject a mastering engineer’s work to preserve the dynamics she wants . . .

She is the reason the song exists, yet she did not create the song.

What we lose here is not simply authorship, but stakes. If our prompter generates a bad song, the prompter is not the one who messed up the song: The AI is.

Which brings us back to Corey Feldman.

The Stakes of Making Music

CF is blamed for the shittiness of his songs. No matter how many other people work on these singles and albums, the common factor is: Corey Feldman’s name is emblazoned on the cover. His voice is the one delivering the words he himself wrote. His taste determines the final product of the music-making process in more than just his choices, but his abilities.

Music makers – of the non-AI variety – are limited by their personal capabilities⁶ in a much more meaningful way than prompters are limited by their vocabulary. Some hurdles in humans-made music are simply insurmountable in such a way that the creator is severely restricted.

While these limits can lead to shitty music, they are also necessary to create good music. A completely unlimited musician would make uninteresting music.⁷ Listen again to your favorite musicians: Their music is constrained by their abilities; they do not have complete creative freedom. Their limits are what cause them to focus their energy on a limited amount of subject matter, sonic territory, and level of musicianship.

So when a human makes and releases music, what they are really saying is:

“Here are what my limits are.“

AI music shows no such limits.⁸ It shows only a product with very little work behind its origin.⁹

And that is why Corey Feldman’s Angelic 2 The Core [ . . . etc.] feels like an actual, joyful experience to listen to. It is a celebration of Corey’s limitations, of his feelings, of his view of the world – every note was sung for a reason, every word was written consciously, every arrangement decision was made because of someone’s taste and limitations.

This is a good thing.

Now, you could argue that Feldman & Co.¹⁰ have put in the bare minimum. Maybe that’s true. Maybe they didn’t give it their all. Maybe they didn’t dig deep enough into their creative well . . .

. . . Or maybe their taste just doesn’t align with yours. Maybe Feldman’s limitations are not what you want out of a person and therefore not what you want out of his music.

Because recording an album requires much more than you think it does.

I would know. I have recorded five of them. Every time, it’s a difficult, straining experience. At times you aren’t sure you can finish it, but you must know within yourself that you will finish it. It can be a grueling effort. It will reveal to you your own personal blocks and incapabilities. This is painful. This is not easy to countenance. By the end, you are aware of exactly what you can and cannot do.

But more importantly, you are capable of more than you were when you started. This is a powerful thing: The act of creation increases ability and pushes your limitations back, like a powerful offensive line shoving a running back headfirst through a defensive line. Like the Philadelphia Eagles’ “Brotherly Shove,” the making of an album will necessarily and inescapably push you forward at least as far as you need to go.¹¹

That is joyful.

That is real. That is human. That is important. Any piece of human art – no matter how much you like it – pushes all of us forward by pushing some of us forward. Humanity is a throng of individuals attempting to move itself towards something far in the distance. This is why life is beautiful; this is why we choose not to give up.

No matter how much you hate something made by a person, that Something is a part of the human experience; and the human experience is sacred, it’s infinitely valuable, it’s a part of all of us.

AI music does not do this. AI music leaves us exactly where we are. It has no true impact, because it was not really created. It was generated, and generation is an entirely different ball game. If you take away one thing from this tirade, I think it should be this: Something which was created is always, always more valuable and worthwhile than something which was generated. Humans need to do something to produce value. AI music is simply an idea that is generated by code; there was no human effort put in to create the product of that generated. It is only an idea. Real, human music is ideas + effort.

Effort is a good thing.

So I say, thank God for Corey Feldman. Thank the Stars for Kid Rock. “Poland” by Lil Yachty is a gift to the world.

Human music is necessary for human life. Let’s keep it fucking going.¹²

1. And there’s a good chance it is. ↩︎
2. Probably while high on something very, very strong. ↩︎
3. (Though most will deny this.) ↩︎
4. Excluding, of course, drug- and alcohol-induced collapses. ↩︎
5. Obviously, gender is not the important thing here. The Prompter comes in many and sundry forms. I simply needed to choose a pronoun, and this is what came out. ↩︎
6. Obviously, things like vocal alignment tools, tuning software, and quantizing can mask a good amount of flaws – but their usage, if the performance is bad enough, is apparent to the trained ear. Plus, if the song is bad, it doesn’t matter how good the sound is. ↩︎
7. See, e.g., Jacob Collier: One of the most talented multi-instrumentalists and composers and music theorists in documented history, who has maybe four interesting songs. ↩︎
8. There’s a reason I’m not putting music in quotes. I do think it is actual, real music. As real as anything. It’s just that underneath that realness is fakeness. That distinction feels crucial to me. ↩︎
9. Aside, of course, from the fact that the AI itself is trained on almost the whole of the human effort in music. This is a tragedy and a travesty, not something to be celebrated. ↩︎
10. And & Co. is a serious Co.: Fred Durst and Snoop Dogg feature on this album, among many others; some tracks were even recorded and mixed by the legendary go-to engineer of Tupac Shakur: Tommy D. Daugherty. If you listen to “Wanna Break Free,” you will hear his fingerprints. This is not an amateur-studded release. Real music-industry figures worked on this. ↩︎
11. “As far as you need to go” is not up to you. It’s up to the creative Gods. I’ll explain this more elsewhere, but in my experience, it’s the truth. ↩︎
12. One last thing: If you search “Corey Feldman AI,” you will find that he used AI to duplicate himself in a music video. This is unfortunate and regrettable. But the music itself – for now – is all human. That’s the last asterisk. ↩︎