UVA Arts, University of Virginia

Vol 09 Winter 18 Library
Prints from Rei-RNN, a multilayer recurrent neural network - a program that emulates the neuron structure of a human brain in order to achieve pattern recognition tasks that would otherwise be impossible for a computer. IMAGE CREDIT: Cameron Mankin
Studio Art

Q&A with Cameron Mankin (College ‘16)

Tell us about your work at the University of Chicago in experimenting with the artistic potential of neural networks (AI). 

I create primitive neural nets that produce prompts for conceptual art pieces, portraits of computers, and (most recently) tarot cards. It works kind of like collage. I feed in a huge database of source material of the given type, the neural net churns up its own heuristic understanding of the material, then it spits out its best guess at what a new entry in that database might look like. The results are often wonky and misshapen - the prompts are impossible; the portraits are amorphous blobs - but eerily similar to what humans might produce. They're uncanny, filled with haunting shapes and ghosts of meaning that we want to interpret like reading tea leaves or the stars. The resulting books and prints give us a lens into how an AI might look at the world, but they give us even more insight into our own expectations and our own desire to anthropomorphize technology. 

I'm really only just getting started at UChicago, but so far, I've been building both my technical and conceptual repertoire. Whereas before I was leaning heavily on neural nets designed by really brilliant folks like Andrej Karpathy, lately I've been making my own forays into writing them from scratch. I've also had the opportunity to talk with my faculty advisers about the scope and focus of my work, which has been invaluable as I start new projects. I'm also about to start teaching an undergraduate seminar, so that puts an exciting new twist on my day to day life.

In what ways did your UVA residency prepare you for what you are doing now?

I taught myself to code during my Aunspaugh Fellowship. I was already loosely familiar with programming logic, thanks to a hobbyist's interest in video games, but during that year I really took the critical leap of picking up Python and CUDA (a platform used for interfacing with graphics cards that has some great neural net functionalities). This really enabled me to even attempt the kind of work that I make now. 

Perhaps more important, however, are the conceptual advances I made during this residency. When I graduated from UVA, I was making these collages and relief prints based on cut-out comic book imagery. A lot of the parts of my current work were there - a pool of source imagery, procedural composition based on a set of pre-written rules, a sense of experimental iteration - but I was burned out on the comic book imagery and I felt like I was just repeating the motions by collaging anything else. The fellowship gave me an opportunity reflect on my previous work in the Distinguished Majors Seminar and the time and studio resources necessary to actually make something new.

Talk about some of the experiences and some of the mentors you had at UVA that inspired you and why?

I could go on and on about all the people I met at UVA that helped me along the way. Two specific mentors stand out to me that I'll mention here:

First, Prof. Dean Dass who was my printmaking professor and undergraduate adviser. He not only taught me the mechanical processes of printmaking, but also how to structure my ideas into projects and series of images. I have a lot of great memories of working late hours with him and a few other students on these amazingly ambitious book projects - vibrant, exuberant things filled with last minute complications and daring new ideas. One wild time he nominated me to fly to Finland on his behalf, carrying a bundle of pages for a collaborative book project we were working on with the Academy of Fine Arts, Helsinki. I stayed there for 3 weeks, teaching Dean's paper lamination technique and meeting all sorts of great artists before returning to the states with their contributions to the book.

Second, Prof. Lydia Moyer who was my adviser during my Aunspaugh Fifth Year Fellowship. She never hesitated to ask me hard questions about my work, but, more than almost anyone I talked to, understood the opportunities and difficulties of making digital work. She really helped me hone my first neural net pieces from a gimmick into a cohesive project. She also is a wellspring of practical advice for a soon to be graduate - how to budget around art-making, what to demand from your next job or school, and what toothbrush to buy. I try to drop by for a conversation in her poster-covered office whenever I am in town.

What impact did you peers at UVA have on your development?

See for yourself...

www.cameronmankin.com

My peers at UVA were instrumental in my growth as an artist (and, frankly, as an adult). Without people like my Fourth-Year show partner Jenny Campbell (who I bounced a billion theories and contrarian ideas off of) or my roommate, Dr. Alex Sirota (who just completed his PhD in Physics and was an invaluable resource in terms of the nuts and bolts of neural network design), I don't know where I'd be. Chance interactions with other students were what made my UVA education work.  

Read the next story

Unexpected O'Keeffe