How did you get into AI art?

If we’re talking generative AI in the current sense, my friend Joe showed me how to run Jukebox, a music-generation AI, in 2021. That was fun, and I sampled the outputs in some of my own music. But at the time, I was already writing my own algorithmic techniques for composing music in Max/MSP. And even when I was a kid, maybe 15, I was using algorithms to generate fractal patterns in my visual art. So I have been interfacing with algorithmic systems to make art for more than half my life. I don’t see myself as an “AI artist”, since AI is a marketing buzzword that means very little, and this relatively new conception of AI isn’t central to how I think about artmaking. Put a tool in front of me, and I will learn it.

What tools do you use?

I don’t want to promote any particular tool. It’s not so much about the tool. If you’re really sure you want to interface with AI systems to create art in this way, I say: learn Python, dig deep into something open-source, and modify it yourself. Be sceptical of tools that make things too easy for you, that you have to pay for. Really learn about how these things work, and exploit that knowledge. Be thoughtful, be critical. There is enough thoughtless, meaningless slop out there.

What is your process?

I think I invent a new process for every project I embark on. I am always learning something completely new. I always try to have an artistic line of inquiry, even when screenwriting, like I’ve been doing for the past year. The process exists to find the answer to the question.

What are the conditions necessary for your creativity to flourish? 

Conditions that are mostly incongruous with the artistic profession, like being able to live comfortably and not worry about money.

Are there certain themes running through your work? If so, what?

Perhaps that’s not so much for me to say. People tend to comment on the emotionality of my work, which I think surprises them, since they may have come for commentary on technology. I experiment with different mediums (the materials and substances used to create works of art) and forms, mixing them in ways I haven’t seen before, and I love to mix reality with fiction. The group Negativland, whose work I look up to immensely, wrote that my film Suicide III “blurs the lines of what the hell it actually is,” which is maybe the nicest compliment I’ve ever received.

Do you have a vision for the finished piece, or do you let the process and the collaboration with AI take the lead?

I am constantly beset by visions, but unfortunately, I find that AI-based tools are inadequate for realising visions. They strip you of a lot of control and funnel you toward a predetermined algorithmic centre, aesthetically. Unless you break the tool, perhaps. I find myself always fighting against the tool.

How many creative ideas are you normally working on at the one time? 

Too many.

How do you prioritise them?

Well, ones that I get paid for are the first priority. Second priority, I try to collaborate with my friends. That’s part of why I find myself at home with filmmaking. I like getting together and making things with my friends.

Who is Loab, and how did you create/summon her?

An emotionally devastated-looking woman who haunts AI images. People call her a demon or cryptid. I summoned her with negative prompt weights, which give you the thing farthest away from your prompt in the AI’s latent space, instead of the thing closest to it.

Why did the public react so strongly to Loab?

Horror fans love her, it’s a new technology people are worried about, and she’s a character with a four-letter name. But I certainly didn’t expect the level of viral popularity she achieved.

What kind of reactions did you get?

Vastly, overwhelmingly positive reactions. Podcasts, fan art, tribute songs, and even academic research articles. But I did get death threats. It turns out that when you summon a demon, some people don’t like that.

What is the role of the grotesque in your art? 

Well, I think generative AI is not a technology that creates – it is a technology that distorts. ChatGPT and similar chatbots are the worst offenders; they’re trying to distort the basic processes of human cognition and make us reliant on them. Similarly, image-generation AI has ingested the entire human corpus of images, and primarily just distorts and regurgitates it. These processes are grotesque. This isn’t what we thought would happen with our data.

Last year, I was asked to curate a selection of short films for Uppsala Short Film Festival in a program I titled “Algorithmic Grotesque”. I included stuff from Negativland, Conner O’Malley, Ada Ada Ada, and my collaborator Eryk Salvaggio. Eryk just curated the exhibition “Signal to Noise” at the National Communication Museum in Melbourne, which I highly encourage everyone to check out!

How is AI influencing human creativity?

AI is a deskilling technology. ChatGPT asks us to forget how to write and think, so we become dependent on it to do so for us. Tools are being developed to replace many artistic trades, and this is especially visible in an industrial medium such as film. In 10 years, will anyone know how to light a scene, or will everything be filtered through AI lighting and colour-grading tools? The way capital promises to implement AI means we’ll suffer an immense net loss of human creativity and craftsmanship.

AI companies seem to be monopolising creativity. How do you work against this?

Anything I can do is very small compared to the amount of capital being poured into making this technology inevitable. For art and craftsmanship to survive in the face of the profit motive, we need large-scale efforts and organisation to secure a place for artmaking in society. Some small things that can be done, such as securing a living wage stipend for artists. But ultimately, the systemic issue is that capital is concentrated in the hands of so few people. Because it’s not just art that will die if this current system is allowed to run its course fully, the earth will be made uninhabitable.

What is the role of accident in your art?

Accidents play a big role. Francis Bacon said that all his brushwork was accidental. You just have to choose which accidents to keep. I feel similarly.

Do you dream about your art or come up with new ideas in your dreams?

Yes, I am a big dreamer, and it’s a big source of inspiration. But I don’t want to spoil any projects I’m working on.

What is your earliest memory of being creative?

I was maybe 4 or 5. I had drawn a picture in crayon on the dining room wall. My mother found it, and hit me – which is legal where I grew up in the US. I didn’t understand what I had done wrong.

A big part of being an artist is navigating power structures that want to dictate how to make art and what art is allowed to show, but most of these rules go totally unspoken. Now I’m trying to get funding for my next short film project, which is partially pornographic, so I expect to hear some no’s from financiers based on hidden criteria against certain types of image-making. In a way, I still feel like that child, just trying to get away with expressing myself in a world whose rules I don’t fully understand.

Who are your biggest influences?

David Cronenberg, Autechre, JG Ballard, Nam June Paik, Maya Deren, Peter Tscherkassky, Kiyoshi Kurosawa, Hito Steyerl, Lynch, Bacon, and my late collaborator, the cartoonist and video artist Morgan Vogel.

What is your favourite work of art?

Lynch’s Inland Empire was the most pivotal for me. There is a before-and-after I saw that film. I think it convinced me I’d actually have something to say with images, and by putting images together.

What are you currently working on?

I’m directing that erotic short film I mentioned. It’s about cameras and the algorithms that control what we see. I’m also in the process of writing several longer screenplays.

Matthew T Phillips