When we look at a photograph, we can ‘read’ massive amounts of information at a glance, without trying. But what does a computer see?

In Recognition Patterns, Marcus DeSieno applies the technology of machine learning and artificial intelligence to his personal collection of found images. In this eerie meeting of archival photographs and cutting edge technology, he gives us a glimpse into computer vision, making visible the way manmade super structures look back at us.

With this work, the idea of photography as a historical document becomes a foundational layer onto which the line-work of Artificial Intelligence etches itself. The series looks at the spectrum of visual literacy and draws to attention the ways in which machines learn the human form through geometrical patterns.

Recognition Pattern #22a. Archival Pigment Print © Marcus DeSieno

Pareidolia is the phenomena of incorrectly perceiving something specific in a random pattern, such as a face in a cloud, as if the human mind is searching for familiarity in the abstract. In DeSieno’s series, it is the computer that actively sees and interprets the world — in strict and geometric terms. Found photographs from the early 19th and 20th century, which have long lost their original meaning, are transformed through the vision of the computer.

Recognition Pattern #17b. Archival Pigment Print © Marcus DeSieno

Human form and landscapes become lost, traced upon by purely mathematical analysis implemented by algorithmic systems of identification. Geometric lines create visual illusions of folds and shapes, brought upon somewhat generic archival pictures: family photos, yearbook pictures, monuments, and documents of public history. Through DeSeino’s images, we catch a glimpse of the lens through which a computer views us.

In the quiet of the picture, we are confronted by stark and unemotional lines. The implications of what this means, and the difference it has to the way we see, interpret, and independently discern choices is immediately compromised as nuanced details are meted out. These lines may seem playful, but upon closer observation, they slice up soft human figures and interrupt the relation of the face to the physique of the person in the portrait with little regard for their identity.

Recognition Pattern #179d. Archival Pigment Print © Marcus DeSieno

The human form is reduced to lines, a framework, hollowed or awaiting another series of lines to be imprinted upon the surface of the image. The loss of the individual life depicts, as DeSenio describes, the consequences of power and the farce and absurdity that underlies technological analysis.

At such a crucial juncture in the developments of technology, Recognition Patterns demands that we look at the reality of a technological future and the way it will shape humankind. What does it mean to be human? And how will that transform through artificial intelligence?

Recognition Pattern #32. Archival Pigment Print © Marcus DeSieno

DeSeino’s work pushes the viewer into a stark confrontation between the realities of visual imaging systems and choice, creating a sense of tension, or cognitive dissonance. The very tools we are asked to work with and live with are still within the realm of choice. But for how much longer?

—Rachel Wolfe