AI Art as a Conceptual Art

 

To be a non-musician or non-artist makes it easy to place a philosophical position before craft. I am of two minds about it since I am a trained musician, as well as an untrained visual artist. Everything I do in art is essentially postmodernist, and I love working in this way. But I tend to get homesick for both: if I've left craft for a while, I tend to miss it. I see the possibility of a happy medium using both as hybrids. 
Anyone who attempts to make art, especially with no training, or is doing it digitally, will do it however they can. They won't be taking lessons--they'll jump right in. I can't draw very well and certainly can't paint portraits or landscapes, but it doesn't interest me. But if I could, I'd sometimes want to do it just because I have that skill. If I were a tennis pro at 25, I'd still want to play tennis at 65. It's the sheer desire that inspires all kinds of activities--not just art-making. People don't feel like they need a practice routine because it simply takes too long. On that score, I do "maintenance practice" or woodshed on specific techniques or skills I need for a specific idea or concept. AI-generation now obviates this as well, and we can cut to the chase without the woodshedding. But again, while we can "finesse"prompt-writing, it is not a form of practice as much as it is a form of editing of an input to achieve a desired output. That not uninteresting in itself, but in terms of philosophy, you're using the hands-off/phone-it-in approach and moves it to a form of conceptual art, or Level 4. 

"Discuss the 4 Levels of Creativity" 

Using AI is essentially using the internet as an instrument, just as we have used the recording studio as an instrument. That instrument will vary depending on the instruments and equipment available. If there isn't a piano or polyphonic synth in the studio, you won't be able to play major 7th chords for example. If I'm using only AI to generate music, I can't use any chord I want. I find that AI music lacks the harmonic interest that's possible by simple playing the piano or guitar and finding a melody. 

Eventually, music AI will become more sophisticated, and users will be able to select more complex chord changes, keys, and meters, and so on. But if we can simply select those things from a drop-down, is there a difference between the simple selection of options and intimately knowing something? For example, if you generate a jazz head that is composed of ii-V-I patterns through several keys, and all the dominants are altered or have complex extensions, what would be the result? Would it sound like the jazz that humans actually compose and play? Music composition with LLMs, while interesting in its syntactical connection with music and language, doesn’t rely on any “procedural knowing”; you’re simply clicking on things in an interface or dashboard, letting it generate iterations and selecting the ones you like. It’s like photography in that sense, choosing a decisive moment from 10 or 12 captures. If we look at creativity being untethered to a domain, where a domain doesn’t (always) require procedural knowing, perhaps the results can be more interesting. 

The Esperanto Blues album I generated is something that was never possible before. While it’s interesting, there are certain aspects coming from procedural knowing, or intimate knowledge or skill that create a cognitive dissonance, even frustration. Why am I stuck with cliche chord changes, or is it that ultimately it doesn’t matter to listeners? AI is already making that decision for us, where we just capitulate to the constraints of the tools we’re given rather than use all the tools (and knowledge) that were always available.    

On Prompt Writing

As AI-generated music evolves and matures, users are becoming more serious about how to use it as a rigorous creative process, centered around the writing of prompts. In my experience, prompts will not always produce the same (or even remotely similar) results. I find it to be an exhausting process that deflates a sense of flow in creativity, not unlike the feeling working with synths has been for me since the 1980s, because of the surfeit of options. Lately, I've not wanted music composition to be merely timbral, merely finding cool sounds that drive the work. I want to work not from prompts, but from musical seeds, which later can be contextualized with timbre. Whether I'm using synths or now AI is irrelevant to my primary underlying frameworks or objectives. However, I'm still open to happy accidents that emerge while working, which can sometimes change the whole direction in exciting ways. In visual art, even one mark that emerges can change everything. Sometimes it finishes the work, and nothing else needs to be done. In AI music, that "mark" can be the way a word or phrase is sung--something I wasn't looking for initially but find interesting nonetheless. 

Using a corollary from visual art, the prompt is the description that accompanies conceptual art, very often a contrived piece of writing where the artist attempts to explain the concept.  It's important to clarify things to the tiniest details so as to make its explanation spontaneous in any situation. Ironically, it needs a "script" or "prompt" when the artist is prompted to answer questions in interviews. If you've done 5 interviews about your latest album or art exhibition, your answers have to be coherent. You have to have a straight story or a solid philosophy on what you're doing. If that's the case, then it won't matter what tools you used, although it is vastly more interesting to people what instruments you used. I'm reminded of an interview with bassist Chris Squire in the early 70s, where he said that the young girls who came to their shows were interested in the kinds of strings he used--Rotosound Swing Bass (Medium Gauge). I would imagine they'd have an interest in 2025 in what your prompt was.