Process update

After about a week in which I spent much of my time messing around with AI voices, I got back into composing – which, you know, is actually the thing to do.

Because this show won't exist unless I write it, and it won't be any good unless I think about it more than I think about the possibilities of AI, and fortunately there are so many people thinking about the possibilities of AI that I don't have to.

And, after exploring the possibilities as they currently stand, I came all the way back around to the decision I made before I started mucking around with any of this:

The concept album needs to be made by human voices and human instrumentalists.

First of all, and I know this is a bit of a tangent, but isn't it interesting that no artificial sound generation program has successfully created a new instrument? There is, of course, the classic "synth sound" that could be classified as a unique instrument (it would probably end up in the percussion family), but Garritan et al are just replicating oboes and French horns and pianos and voices the same way that the AI art programs replicate Studio Ghibli, and with about the same level of nuance.

It's sort of like Pantone not precisely being able to create a new color. Sound is made by striking things and blowing into things and increasing or reducing the length of things and causing things to vibrate against each other, and although I wouldn't say that we've reached the limit of the kinds of instruments we can build, the only new sound a computer has ever created is the sound of electric pulses smacking against a plastic speaker.

There's a larger metaphor to be drawn from this, and as always I trust you can draw it on your own (and if you can't, ChatGPT might be able to draw it for you).

But the point I'm trying to make here is that I want my musical to be performed by humans. I don't want to make a concept album with Cantai or any of the other AIs.

This isn't to say that we can't use technology as a tool. I'm still interested in potentially using Cantamus to teach people their parts, especially if they're not skilled at reading music, and the fact that Cantamus makes no attempt at artistic expression means that we might be able to pull this off without anybody trying to imitate the robots.

Speaking of which, I could go on an entirely new tangent about the whole "creating musicals that can only be effectively rendered with prerecorded backing tracks" thing, the wall of sound that can't ever be replicated by an orchestra, the ensemble singing along with synthesized voices and so on. I was trying, once again, to listen to The Great Gatsby – and once again it made me so angry with the state of everything, the anodyne adaptation, the generic techno-pop orchestration, the ridiculous and predictable rhymes, that I immediately flipped it off and switched over to Falsettos because I wanted to hear real music written for (and performed by) real musicians.

And since then I have been scavenging every Falsettos slime tutorial I can find on YouTube because they're all so different.

They have to be, because they are being created in real time by people who are making unique, individual, in-the-moment choices around a series of predetermined constraints.

Even the same staging is different when you put different actors into it; nobody's trying to mimic a specific performer or to create a "living movie," which is what I used to call it in college.

And –

and this may be the most important part –

you can see just how much information each performer is able to communicate, and you can see that communicating more information is more valuable.

I've been watching community theater productions, for example, where people are putting in and/or giving out a lot of information – they are precise, insightful, connected to their part and to the ensemble – but it's clear, for example, that people like Christian Borle or Mandy Patinkin are still able to communicate more.

Which is good.

IT IS GOOD.

Because that kind of standard gives us something to work towards –

and even if we never get there, we can get a little bit closer than we are right now.