A non-human entity complicates common understandings of queerness and HIV. With agency over data, it distorts images from archives and makes them speak. Composed with machine learning processes for generative video and sound, including deepfakes, the piece explores uncanny relations between technology, politics, and illness.

“This video installation by Marcos Serafim engulfs visitors in an environment where things may not always seem to be what they appear. Utilizing artificial intelligence (or machine learning) processes for generative video, the artist created a multifaceted work that addresses public perception, the circulation of images, and campaigns of (mis)information specifically related to the AIDS crisis, depictions of war, and the many connotations of the term ‘virus’ today. What is the relationship between an autoimmune virus, a computer virus, and a societal sickness? The use of deepfakes—a video generation technique that combines the notions of ‘deep learning’ and ‘fakery’—allows Serafim to manipulate pre-existing audiovisual content as well as layer in his own, a strategy at the forefront of present-day concerns around the spread of false or misleading information.”
Steven Bridges, Associate Curator, MSU Broad Art Museum

“Classically, the purpose of images, narratives, metaphors, etc., has been to make sense out of sensibility by selecting among the chaos of lived experience a simplification that can be understood and communicated. This process always involves abstracting some essence out of a multitude of possible inputs, which is where aesthetic choice comes into play. This is even true of artificial intelligence systems in which the data must be ‘cleaned’ before it will yield anything comprehensible to the human mind. What Serafim has done is to reverse the usual aesthetic. Instead of trying to clean up experience by finding the narrative threads, metaphors and iconic images, he has tried to recreate the lived chaotic experience by feeding an AI system with not just unselected images but more importantly the entire range of images that bombarded a pre-ART person with AIDS. By refusing to select the narrative line, one type of metaphor, a cleaned set of images – indeed by reversing the process to mix metaphors, take out the narrative line, and ‘dirty’ the images – he recaptures the confusion, distress, fear, anger, and incomprehensibility of lived experience. What I find so exciting is that this messy, real, contradictory, insane, problematic, ‘dirty’ experience yields a completely different aesthetic that Serafim illustrates in his most abstract results.  This aesthetic is much more difficult to comprehend, unfamiliar, disconcerting, and thus so much more ‘real’ than the ‘photorealism’ questioned in the beginning of his essay precisely because it IS what it is like to experience something like AIDS. It IS the mind trying to make sense of what the mind (and body) are doing when it is impossible to comprehend what is happening to it. An art that captures that sensibility is something new.”
Robert Root-Bernstein, Ph.D.

Courant3D / Immersity, 2020
Program: Artificial Intelligence / Association Prenez du Relief, Angouleme, France
6th DOBRA – International Festival for Experimental Cinema, 2020
Program: Speculative Experimentations / Cinemateca do MAM – Museum of Modern Art of Rio de Janeiro, Brazil
Autoimmune, solo exhibition, 2020
Eli and Edythe Broad Art Museum, East Lansing, MI


United States, 2020
Video and third-order ambisonics audio, 12:25 minutes

Three other pieces derived from this project: War Face, Face Up and Interface.
Made in collaboration with Douglas McCausland (music composition, sound design and mixing) and Ian Kirkpatrick (creative coding).
Autoimmune was funded by the Collaborative Arts and Design Research Grant from the MSU College of Arts and Letters and the Department of Art, Art History, and Design; counting with the support of CCRMA- Center for Computer Research in Music and Acoustics at Stanford University and the MSU College of Communication Arts and Sciences, the College of Music, and the Department of Theater.