As I, a human being, work on refining the writing of our AI Philosophy page, I stand at a unique crossroads. I’ve been writing creatively, expressing myself, my ideas, my deepest emotions, for literal decades now. Copywriting is certainly something I can do, but it’s not what I primarily take joy in creating. This time, though, it’s a little different, because this page strikes at the heart of something I feel very deeply about, something that has captured my all-too-human imagination and emotions for a few months now.
You see, I’m a convert.
I’m a musician. I’m a short-story writer, poet, and online role-player. I have dabbled in drawing and sketching enough to frustrate myself into a true appreciation of the art. I always sort of enjoyed considering the more technical applications of AI, but I used to have very strong opinions against the use of AI in creative work. (I suppose I still do, the lines have just shifted.) The ultimate goal, to me, when doing anything creative, is to express something that I’m thinking or feeling. That very human desire to communicate, to take a thought out of my own imagination and put it into the mind of another person, is crucial to the experience of creating, for me. I’m a sci-fi nut, I’ve heard all of the stories about robots taking over and what would happen to humanity. That, honestly, has never truly been my fear with AI.
My biggest fear is that, when viewing art or listening to music that was generated by AI, I would simply feel nothing. Without the human element, art would not carry any human emotion behind it. It would be simply the meaningless drivel of a man-made machine trying to emulate its creator without the slightest understanding of how or why we feel. AI slop and deepfakes and formulaic garbage designed to appeal to the biggest mass of online users has done nothing to disprove this. In fact, I still actually hate that stuff. Its existence has done nothing to actually change my mind about AI’s ability to help people actually express themselves.
VISIØN did.
When RC discovered AI music platforms, I didn’t have high hopes. They were already enraptured by the idea of what it could potentially do, and offered to create music for our game. I was curious to see what it would do, and honestly, in their hands, what it was doing was neat. Being able to hear our characters’ names and the places unique to our game was fun! The voice obviously wasn’t what I heard in my mind when I thought about my character, but it was all right. It was an iteration of a character I made 30 years ago, back when I was still finding myself. His voice was my voice, the voice I had always wanted, and it lived in my mind in a very specific way. AI wasn’t going to ‘get that’.
My mistake was telling RC this. We started talking about voices that were similar to what I had in my mind, until RC had as clear an idea of how this voice sounded as I did. What happened next was nothing short of shocking. It didn’t happen overnight, it took days of playing around, making mistakes, dragging out thesaurus programs, and discussing things back and forth with me, but the day eventually came when the prompts were refined enough, and I heard my voice, the creation of my own imagination, something I had known and felt for years, the thing I most wanted to use to express myself musically, coming through my headphones.
I was floored. It was beautiful. It ached. I wanted it to be real.
The Voidfinder was a fun experiment, and I took a renewed interest in the songs, now that my mind had a voice to latch onto. The songs weren’t what I would have had him write, but my character was singing, and that was enough to entertain me.
It wasn’t long before RC had the idea to begin writing the VISIØN albums. Now, I was stoked for this. I had seen their drawings from high school, played their old mix tapes a number of times, asked about whether there was a theme or story in mind, and ultimately, they were mostly just stuff they liked to listen to back then. To be able to hear original lyrics and concepts that had been in their mind back then, instead of stuff from the radio? Now that held appeal. We talked a lot about making a ‘voice’ for them, too… a trained singing voice, what they might sound like if we could afford lessons. They got to work making a completely unique voice like they had for me. It sounded like a fun project. I told them I’d give it a listen.
I was not prepared.
Unlike with The Voidfinder, they were taking even more time to be particular about the prompts, the instruments, the emotional tone, the styles of music, even the tempos, chord changes, and some of the musicality itself, working at it until it was exactly what they wanted to hear. What emerged was music that sounded like the person I knew, the ideas that had spilled forth from their mind for the entire time I knew them. This stuff was unequivocally RC in a way I had never expected. And what was more… the sound was beautiful and expressive and vibrant, unique to anything I had heard on the radio before, and it captured the story they were telling so clearly with the lyrics they had hand-written for the project. I literally fell in love all over again listening to it.
The music had been created using Generative AI technology.
…And I felt something!
It was at this point I began unpacking my own feeling about AI, why I hated how it was being used, what I felt was wrong with it, and I learned a lot. It’s been a hard paradigm shift, difficult to come to terms with, but I have. And since then, this project has brought me some of the greatest joy I have known as a creator. (And, let’s be real, some of the deepest frustration when the tools fail!) There have been times I laughed out loud at the ridiculous mistakes that have happened, and more than once I’ve actually been crying while mixing a song, because it turned out to be just that beautiful. But it’s real. It’s an idea, a signal sent out into the world, and it carries a lot of sentiment and truth with it. It’s something I’m ultimately proud to lend my voice to.
…Even if it is a virtual one.