AI Music
I'm not sure how to feel about AI music. To start off with, some of it is actually really bangin⤻. At this point, I don't think there's any denying how good it can sound. The question is more about how I feel about the process, because the outcome is enjoyable (I'll come back that idea in a moment).
I do have to say that I'm generally impressed by the fact that AI music exists at all. I'm not sure how it's even possible (and IDK if anyone really knows how it works). The ability to generate lyrics is seemingly the easiest part, but I only say that because LLMs have been producing convincing text for some time now. But to be able to generate the actual instrument sounds and moreso the ability to make convincing vocals is what I'm blown away by.
I don't actually know what goes in to creating AI music. I know you can just prompt something like Suno AI and it can generate lyrics and the music wholesale. But is that what most people are doing? Or are people feeding it their own custom lyrics? Or perhaps a short sample of a melody to base the rest of the song? Are they taking the output and mixing it themselves, re-generating sections/lyrics/etc and making adjustments?
Does the fact that the whole song was generated without any real musicians involved discredit it?
I initially want to say it does. But in the end... why does it matter how the song came into existence? If it's enjoyable to listen to, then it's 'good'. Why is something enjoyable in the first place? Human music taste is an interesting matter because it's tough to trace back to a more base instinct. Preference for food is explainable by the fact that you need to know what is good to eat so you developed a taste for things that are good sources of energy (sugar is addicting because of this). Preference for how people look makes sense as beauty could be indicative of physical health for reproductive purposes. But music isn't naturally occurring in the same way, so why do we like it at all? (I will not be able to answer this, so I guess I'll just leave it to you to ponder).
So what about music is it that we really enjoy? Well some people like songs for their lyrics. Whether they're happy or sad, angry or holiday related, they make us think of those things and feel the emotions we want. Some songs are enjoyable just because their component sounds are beautiful. An orchestra or a lovely vocalist can produce sounds unlike anything else. Some music is sought after for its sheer difficulty. 'Math Rock' is a genre that basically exists to prove that you are better at counting than anyone else. (If you've never heard of math rock before it's pretty neat⤻, but I digress.)
So in my estimation, other than the music that is enjoyed as a matter of difficulty, the AI generation of the other types of music doesn't really take away the reason you would enjoy it. Now, maybe knowing that a human put their own emotion into the lyrics/the playing of the instruments does hold some weight to you. And AI generated music does have a bit of hollowness to it for that reason. So even though music has been 'computer generated' (synthesizers, DAWs, autotune, etc) for some time now, this goes beyond the remit of a tool and into a realm of generation that requires as close to zero input as possible.
So perhaps while I do like the music that is produced by AI, I am discouraged by what it means in the broader picture. Where does the musician fit in to the future of music production? What does this mean for live music events? There's already so many genres, people hardly share a music taste with anyone around them. Is this just going to totally isolate your music intake from everyone else by generating personal songs, tailored for you by AI algorithms? What are the social effects of having one less thing to share with others?
Yours Truly,
[Redacted]