Artificial intelligence’s (AI) effects on the arts has been a controversial subject at the forefront of artists’ minds for several years. Recent advancements in AI and language learning models have rendered it capable of creating entire songs from simple prompts, mimicking instrumental tracks and writing lyrics that sound indistinguishable from human-made music.
In Moss Performing Arts Center (MPAC), CMU’s Assistant Director of Vocal Studies and Interim Assistant Department Head of Music Caitlin Moore’S office sits in the northwestern corner of the building. She has an orderly desk surrounded by shelves of green plastic plants. Moore sits across the room from her well-loved baby grand piano. She provided a music professor’s insight on recent advancements in AI technology.
Do you believe that AI has had a positive or negative effect on the music industry, and how could that evolve as the technology develops?
“I think that’s a really hard question to answer in a binary form, positive or negative. In my opinion, so far, I’m kind of comparing AI to the internet. When it came out, you had people very scared of the consequences, but also seeing the advancement that it can cause, and AI is the same way. Yes, there are parts of it that are scary, it’s a lot of the unknown, but at the same time, obviously there’s going to be advancements we can use it for, so I’m going to withhold my judgment about positive or negative. There is no choice but to advance culture at this point, and I think AI’s going to be part of that. How that happens is yet to be seen.”
What aspects of teaching or engaging with students have changed in music since the widespread adoption of AI?
“Some helpful things I’ve seen personally in AI are [that] it’s easier to create rehearsal tracks, which are really helpful since I teach voice lessons and commercial ensemble, I see that a lot, so that’s a helpful thing. One of the tricky things I’ve seen is students will write assignments, like program notes, for instance, I’ve seen some students take AI and use it for their program notes, and it’s pretty obvious when that happens.
The problem is there’s a lot of information in there that’s not quite correct. It’s close, but not quite correct, and it’s the same thing with diction. I teach diction for singers, and we use what’s called the International Phonetic Alphabet, and AI is actually really helpful in helping people figure out how to pronounce words in a language they don’t know.
It’s mostly right, but not always, so I think my general rule of thumb when using AI is using it for the process but not the final product. It’s a tool that you can use, but you have to make sure you’re checking your sources and doing all the stuff we used to do before we had AI.”
Is AI frequently noticeable in the student work you receive?
“No, because I don’t do a lot of written papers and things like that, and it’s really hard to have AI practice for you. Maybe when people are researching in the background on composers and stuff like that, and again, I’m fine with it, as long as you check those sources and use it as a process and not a product.”
Will embracing AI put musicians at a disadvantage in the future?
“Again, I think that’s a really tricky question, again I think it’s kind of the same. If we think about AI as a final product, like [having] AI write a song…, then yeah it probably would put musicians at a disadvantage. Maybe I’m just hopeful, but I don’t see it completely replacing musicians. I think that part of the appeal of live music is imperfection, the fact that at a live performance, you don’t know what’s going to happen next. Something could mess up; there’s that risk factor that AI takes away. Same with listening back on a live performance, there’s something beautiful in that imperfection; it’s a huge part of what we do. Maybe in some ways it could put musicians at a disadvantage, maybe there’s some jobs that can be replaced by AI, but I think that we’re smart, and we’re going to find a way to incorporate it into what we do and not have it completely replace art. Hopefully.”
Do you think AI can replicate the art of humans in a non-live setting?
“Yeah, sometimes, it is already hard to tell whether art is or isn’t AI, sometimes, and sometimes it’s very obvious. I think it’s gonna get better, but it’s the imperfection in human art that makes it beautiful.”
How do you think AI will affect the state of music schools in the next five years?
“I think that there will be more classes offered on it. From a teaching perspective, we have classes offered in AI, both on how to use it in the classroom, how to spot students using it, and how to use it in our profession, like writing lesson plans and stuff like that. The same is going to be true for music school.
I think that we should, instead of trying to just not use it, we should embrace it, because it’s probably going to be around for a while. We’ll have classes, especially in production and stuff, that teach students how to use it in a way that’s helpful to their careers. Students will be using it, and probably already are using it, when they need help.”
