The Great Pretenders

I was reading an article this morning about art fraud, or more specifically, how your own feeling about what you were looking at can change after you found out it was a forgery even though it’s perfect to the untrained eye. The sentence in the article actually ran like this: “A 2014 study published in the journal Leonardo tested how the belief in authenticity of art shapes our perception of it.”

Yeah - that’s the kind of topic I like to wax lyrical about. Let’s shift the focus to the left a little and point the torch at music. Last year, whilst intending to make a couple of demos to check it out, I ended up taking every song I’d ever written from 1992, right up to May 2024 and reworked them all with an AI robot (at Suno) not because I wanted to ‘disrupt the industry’ (God, I hate that phrase), but simply because you can’t have an honest opinion on it unless you’ve got your hands dirty. I was really curious as to how it worked, how to make it work and more importantly, how far could you go with it?

Turns out, very far indeed.

I didn’t intend to do quite so much but it was a bit like buying a family pack of Haribo and just taking one out of the bag-i.e., nigh on impossible. A few songs to experiment with turned into five albums with a sixth well underway - in five months! It seemed wrong to produce quite so much (even though the quality was high), so I stopped and even though I’d released them all to streaming services because I thought it was good enough for others to get a kick out of, I hit the pause button.

Well, actually, I walked away for a couple of months to get some perspective and got on with something else.

There’s two things worth noting here. The first is that the response to the work was mixed. Some songs went through dozens and dozens and dozens of iterations before they were finished so it’s not like you simply turn up and say “Hey, Suno, write me a song” - you could do that but you’d get shit handed back to you. To make it work, you do need some kind of experience of a song. It absolutely sucks diesel at writing lyrics too, which was fine by me because when it comes to songwriting, that’s my thing. But the fact remains: on Suno, you can upload a song you wrote and recorded on your phone, give it some direction and turn it into something that more or less sounds like you spent a digital fortune on.

Anyway, some of my friends dug in and could see that what I was trying to do was create huge songs that could, in the real world, be a huge theatrical production, because that is indeed how I think. Others heard it and thought, “Well that’s not real, is it?” As if the act of creation has to come with sweat and broken strings or it doesn’t count.

Both responses are fair.

And despite personally being very much into the fruits of my labour, it all felt more than a little bit hollow-which brings us back to the initial idea…

FAKE IT TIL YOU MAKE IT

There’s a point to an art forgery. It has a purpose whether that be to deceive a buyer and make yourself some money or simply to prove you can do it. There’s intent behind it. Same with this stuff—nobody’s trying to pass AI music off as Springsteen (well, some are), but if the only goal is to see if you can do it, then what?

I pulled my albums off the streaming services - there are many, many people doing the same thing who don’t find it hollow, but to what end I don’t know. Maybe they’ve figured out how to monetise it, which is as good a reason as any to keep going. I know exactly how it’s done and I’m as good as anybody with the tech available but what intrigues me more, is the thought patterns of the people who are listening to it.

Do they enjoy it as much as any other genre? ‘Robot music’ has been around since the first computer ever got plugged in. Has it even become a genre all of its own - to be enjoyed on its own terms in the same way you learn to live with CGI dinosaurs in Jurassic Park? Watch enough of them and you forget they’re not real in the end and I would venture a guess that this is what’s happening here, even though there is very much a divide between generations of people and what’s acceptable.

(For the record, I didn’t totally abandon the work. I recollected it as some kind of greatest hits compilation and stuck it back out there for posterity - it’s called Dig Out Your Own Pirate Grave and you can find it on Apple Music, Spotify and wherever else you might look).

All of which brings me to a far bigger question. Does the creation of the art matter if the audience enjoys it?

I mean, a covers band is fraudulent but people still go and have a good time. The remake of Planet of the Apes is not an original idea but it got better and better until the sequels eclipsed the original sequels (the original first movie is exempt from this train of thought).

What actually matters to me appears to be soul. I love art that has it visibly on display. I can take mistakes in my stride easily, which is probably why I always think a band’s first album is the best they will ever do because: there comes a point at which you get so good at something, such as writing a song, that you no longer write a song from your heart but rather, know what a great song looks like and are able to replicate it.

…which is pretty much how AI works actually, but that’s a whole other box of frogs we don’t want to open.

This is a cold hard truth about anything in life. Go compare the first Harry Potter book to the last. Hitchcock’s early work to the latter. Apple’s early designs to now–brothers and sisters, I would kill for a Blueberry iMac that worked like the new MacBooks do. It doesn’t pay at a cellular level to be fantastic at something for this very reason. Ain’t nobody listening to whatever the last ABBA album was called even though their skills are tenfold what they had on Arrival.

Did you really write that book if an editor had to point out where to make it better because you sure as hell didn’t bake that cake if you opened a box with Sarah Lee written on the front.

And I’d love to know what would happen if AI were able to create a technically better version of Mona Lisa. Where would that put us all from a cultural standpoint? How are we to separate the 500 year old original from something technically better because that’s what experts look at…

Big questions from every direction, but does it matter who lit the match if the fire still warms you? I guess the vast majority of people could care less, so long as they are warm.

So... I’m not saying that people shouldn’t make AI music (and I really don’t think anybody professional should be worried by it) but it was an interesting experiment for me and also one that I moved on from. If I’d made a few grand out of it, maybe I’d feel different but money’s not really my thing in life to go questing after. Nobody gets paid to go looking for the truth–but you know where to reach me if you know something.

I don’t want to do months of work and feel hollow. I want to do months of work and feel - actually feel - that I created something worth a damn even if it was only to myself.

Just because I could doesn’t mean I should. And just because I did doesn’t mean I still will.

Changing your mind about things is actual freedom don’t you know.

The lesson here is that perhaps AI is a gateway drug to believing you’re something you’re not.

It’s not unlike passing your driving test and your old man gifting you an F1 McClaren for day to day use. You’ll get where you’re going really fast, but at some point you’ll take it too far and crunch it into a wall, laying waste to all in your path–and when you do, you’ll still have your hands on the wheel and your foot to the floor because you’re hooked on the dopamine of being better than you were before.

Fake it till you choke on it.