VAL3R1E: The AI Assisted Music Project, and What's This All About?

Permalink

Follow me via:

---

Text-to-speech support required

So, I’m sure a lot of people, or at least some people, remain apprehensive. Some probably angry, banging their fists against a wall, plopping some deuces on the ground, or what-have-you. That’s fine, as long as I don’t have see it. But as someone who’s been making music for eleven years, and has seen some of the worst of the music scene and also the industry as a whole, and as someone who most definitely knows what I’m doing: Let’s talk about VAL3R1E.

VAL3R1E kind of started out as an experiment as I was learning more about AI. I spent about a year and a half being blindly angry about LLMs, because that was the thing to do and be in a lot of social circles circulating the Fediverse and parts of Bluesky. And, you know, you don’t want to anger and piss those people off because that could mean social ostracization! You might get digitally dragged through the mud, and mulched for doing things you wanna do, and investigating things that people are telling you to hate and be mad about!

But, then I realized, a lot of people on the third-party social networks have already socially ostracized me for much more tame things, and honestly, opinions that are correct and good that I will never budge on (and I’ve basically conceded that making new connections and potential friends on any social media in 2026 is just never going to happen anymore, people are too stuck in their own bubble).

Obviously, though, I decided to do my own work, and I discovered a lot of very complicated subjects, points, and nuance that many refuse to explore, or think about. This is why a lot of the anti-AI sentiment you see across Bluesky or the Fediverse reads like it’s still based on assumptions and data from 2022 … because almost nobody actually investigates the things they’re mad about. And now you have so-called leftists championing the idea of strict and extreme copyright laws, as if that makes any sense.

Or …

“There was a study done that AI makes you STUPID!” Because calling people stupid will surely convince them that they’re wrong about things, and side-stepping the fact that none of those studies have been peer-reviewed: This is something most people who call themselves leftists wouldn’t be doing in any other circumstance or context, correct?

Suffice to say, this is all very complicated, and a lot of sentiment is very … surface level, face-value.

But, what about VAL3R1E, “Don’t you just type a sentence, hit ‘submit’, and then upload the output to a streaming platform?”

No.

You might notice that it’s been months since I produced and released GHOSTFACE. And that’s because I’ve taken the concept and workflow to a lot of steps that I think make it actually listenable. And now, if you’re one of those aforementioned angry people, you’re probably saying, “You didn’t make anything!”

Basically, I start the way I always did when I was making music for Eyeshadow 2600 FM. I open FL Studio, I arrange a bassline, maybe drop in some kicks, snares, and hi-hats, maybe something that goes, “doot doot pish bom pish pish doot doot doot bom,” and you can totally get a bit more in-depth and complicated. But then once I’ve got that, I … give it to the AI, and describe what I’m going for. “retrosynth, cross between numetal, live snares, 152bpm,” etc, etc. Except, oh no? I don’t have lyrics.

You gotta write those, too (don’t generate lyrics, it’s almost always completely nonsensical).

So when I get to this part, I kind of start thinking of themes, or things I’m mad about, things that piss me off, or maybe I think about fictional concepts, like having a romantic partner! Or, concepts related to isolation, loneliness, or … The Complex.

Then, once you’ve gotten all that, it’s time to hit the “Submit” button.

But I generate … maybe 5-10 different times, until the original input and my ideas match up with what I’m expecting, or imagining (imagine? people who make things with AI can imagine things?! Oh FUCK!).

And it’s done, there, right?

Nope!

Once I’ve completed these steps, and have something solid, it’s time to download the track, and bring it back into FL Studio.

This is where I’ll add things. Maybe clips, quotes, more snares, more kicks, whatever. Then, of course, you have to mix and master it properly, and this is something I’ve struggled with for a long time, because it’s a complicated thing to do. To make music sound good on everything. And I think I did a good job with that on ENTITY ZERO.

Basically, it takes a lot of the cumbersome, grueling work, and switches the process around. But, it’s also about as quick as I used to produce music in FL Studio with just raw digital synthesizers, except now I can have a vocalist, and a lot of other things I would have never had under Eyeshadow 2600 FM. Because, and as I’ve stated on the socials: I don’t know anyone. I have a hard time connecting with people, and it seems most people don’t even want to talk. To add insult to injury, I know nobody offline outside of people at my job, and I’m drowning in bills and debt.

If this sounds bad to you and makes you angry, I implore you to explore the idea that all people are their own person, and everything is heavily nuanced, and you cannot escape that fact. You never will.

And I know, that even after reading this, there will still be people who are like, “Slop! AI bad! NO AI!” Most definitely the asshole synthwave scene that originally excommunicated me for thinking transgender women should be treated fairly by them and the scene in-general.

That’s okay, though. If that’s what sparks joy for you, be all about it, and keep on truckin’.