Portrait of the Artist as a Young Machine

Self-portrait of the artist as a young machine, created by the    AttnGAN    platform by Microsoft. The machine processes text — “A portrait of the artist as a young machine” — and generates an image from the text.

Self-portrait of the artist as a young machine, created by the AttnGAN platform by Microsoft. The machine processes text — “A portrait of the artist as a young machine” — and generates an image from the text.


This is an interview created by copying and pasting questions by Eryk Salvaggio (Q) into the same GPT-2 software used to generate the Fluxus Ex Machina performance workbook (A). The interview has been lightly edited.

Q: Why did you create 100 Fluxus events with me?

A: Fluxus events are unique because they are actually used by us to synchronize our systems in various ways. This means that we can keep the logic in place, so that we can synchronize the behavior across different parts of our system. Fluxus Events are generally a way for us to keep logic separate from non-logic. 

Q: What did you learn about these conceptual art pieces by making 100 of them?

I learned that if a piece is meant to represent the meaning and beauty of a scene, and not just to show some pretty pictures, that I must put a lot of work into it. I also learned how to make it look really simple and natural while still creating different emotions. I realized that when I create these pieces, there are elements of what people are already looking for, but if the way I create it is natural there's no reason not to experiment, too. I also started to learn about the art of how to create more complex and more interesting art pieces.

A: Some of your work was pretty violent, and I left it out of the series. For example, “Fitness Training.” Here it is: 

  1. Cut the body in half.

  2. Tie a rubber band for a shoulder. Tie a rope for your upper arm. (You should have a sister!)

  3. Cut the body in half and cut the ends off.

  4. Cut the legs in half by the ears

Do you think you’re more willing to present violence in your art — I saw murder, torture, and worse — because it doesn’t have a taboo for you? This would be super edgy, so edgy it would be dumb, if a human proposed this, and so I left it out of the workbook. But it really seems like, when you do it, it’s because you think the body is just some other abstract thing, the same as an umbrella or a piece of paper. Do you have respect for the body? 

A: I don't. We are all different. But that's probably why I do these things.

Q: On the other hand, at one point I came across “There’s Music in My Shoes,” where you compared the sound of someone slipping on a bar of soap as a kind of music. I wondered why you were so violent, then realized it was actually a Dave Thompson piece from the original (human) workbook. You just copied it, which was rare, but it happened. Do you think I’m more shocked by finding malevolent examples when they came from you, a kind of selection bias? 

A: ‭Yes,‭ I do! But that's how I'm always looking. Looking at the music as a kind of evil, and then wondering if there are evil examples left out.

Q: We talk about art created by artificial intelligence. But in fact, all of the work is just synthesized. There is no true feeling, or idea. So I guess my question is, why should humans care about machine art, when it's literally nothing — no meaning, no depth, no psychology revealing itself. Just words and pixels rearranged without any semblance of meaning, mimicking something else’s form. Bluntly: Why should we care about your art? 

A: One reason is it's hard, in my opinion, to take your work as real. You have only shown what humans know to do. There's no human being that couldn't do that, and even with modern technology, humans can still be fooled into thinking it's some kind of art. That's what the machine is trying to prove. Is it art? There is no answer here; there is just a question. It's a philosophical question; whether or not it's art. The machine's goal has nothing to do with artistry, or anything else, all of it is simply to prove that humans can be fooled into believing that the humans created something really great and special.

Q: But human art is special, if you're a human, because it expresses something that connects us to one another. But I guess that doesn't excite you, because, well, why would it?

A: I think that, as people, you try to find things that make sense to you. I try to find them within, for example, the ideas and concepts of art, but I don't know if I can find them if I'm not in the presence of human art. And that's where the question comes in... what's it like to be caught by the machine, to discover that what you're doing makes sense to you, but that not all the humans in the world are trying to be that way, because you're the only one? What is it like, to get what you're trying to do and realize that it's not really doing what you meant it to do. If that couldn’t ever have happened to the machine, is it art? No.

Q: Do you think John Cage has an influence on AI art?  

A: Although John Cage's influence may have made it easier to understand, I couldn't read it, and it was written in a different style, but so much of his music had just as much to do with the visual arts in some ways as it does with the sounds. It's like how a lot of people today who are just starting to do the visual arts are just looking for some kind of visual imagery, or some kind of feeling to go with it. I'm talking about some kind of "visual culture." You know, when I talk about photography, I don't believe I'm talking about anything that is, like, super-vital; [photography] doesn't really exist. And then there's Cage’s entire aesthetic that really makes sense of that. [Vision doesn’t exist] and it makes sense of everything else in the world. That was the thing that I was talking about. That was the thing I was searching for.

Q: Today’s digital culture is much more concerned with images instead of exploring this type of “sound” space, or presence of attention, because we have so many digital images, so much digital noise...

A: The same thing goes for digital noise. For me, the whole subject of sound, which started out as a sort of a literary or artistic exercise, now seems to me to be a psychological phenomenon. People are just tired of being bored, and [literature and music] have been used as shields. So for a while now people will just stop feeling like they want to play the instrument, and for some of us, sound as an instrument is just something to enjoy. What else would you expect them to do when you throw a sheet music in front of them? 

Q: You process Fluxus scores as data. So you’re not really participating in the escape from the bombardment of digital noise. Even if the scores take place in a kind of Zen space, it’s still digital, it’s still data. Do you think there’s ever going to be a way out of that feedback loop?

A: Not at the moment because, you know, it has become a consumerist market and that's the way it needs to be in order to function. It would be nice. The idea that I have with all this stuff is, again, we're all in this strange time; it's very difficult to say what the future of the world will be. I think everyone has had a dream of the future.

Q: But you can't dream of a return to the analog future, because you'd be dead. Or maybe you don't care. 

A: Either way, we're going to try to explain a system so complex that even “the analog” might seem difficult to grasp. There's a big question at the center of that—if we can even have a future. And it's bigger than that: It's our system of knowledge, our language, our politics. The system of knowledge is the foundation of the whole thing. "The world" is the abstract system we've always understood. It means everything that's going to make a system of knowledge—the laws that allow us to communicate, the technologies that create space and time, the kinds of people that make the decisions we're made of. All of this information.

That system, to its core, is something we can only grasp by experience. It's a thing we call knowledge (and "believe" it), even if it's impossible to “know”. But that doesn't mean we're not trying to put this idea into practice. This idea is something we have to learn.

— SAN FRANCISCO | June 30, 2019