Queer Affects at the Origins of Computation

Much has been written on Alan Turing and the origins of artificial intelligence (AI) some seventy years ago. Turing's "imitation game" set the foundation for research into what has become the future promise of nearly all AI-driven industries today.¹ At the heart of Turing's work is the notion of intelligence as performative, that is, as an effect that need not demonstrate any internal awareness of intelligence as an abstract or conceptual goal. Turing famously likened this performative quality of intelligence to gender, which he imagined as equally transmutable and inessential—a comparison that opens up the possibility of a queer reading of AI through the discourses of performance, language, and affect. Nonetheless, in our hagiographic treatment of Turing as the so-called father of modern computing, we often miss those queer objects and relations that constitute the broader milieu of experimental mathematics during this period. Working alongside Turing at the University of Manchester Computing Center in the early 1950s was a gay man named Christopher Strachey. A prolific early programming language designer, Strachey is best known for developing what are arguably the first examples of computer music and computer games, along with a love letter–generating algorithm that is widely considered the earliest work of computational art. That Strachey developed so many groundbreaking programs at the precise moment Turing was theorizing the foundations of artificial intelligence speaks at once to his skill as a researcher and to his mutual interest and investment in experimental or non-normative uses for computational technology. While their colleagues worked on applications in optics and aerodynamics, Turing and Strachey approached the computer with a distinctly different set of affects and investments, asking the machine to perform not only intelligence but also play, sincerity, camp, and even love. Examining the history of early computing through these two queer figures allows us to mark out a set of affective relations toward computational machines that presage the contemporary moment while critiquing our own investment in the normative intelligence of artificial systems.

In looking for a queer origin to the history of computation, nearly all scholars are drawn to the figure of Turing, considered by many to be the originator of modern computer science and arguably its most visible queer subject. As I have discussed elsewhere, Turing is a unique and captivating figure due in part to the visibility of his difference and the tragedy of his death.² While not a secret, Turing's sexuality was not widely acknowledged within computer science and mathematics for many years. Following the publication of Andrew Hodges's definitive biography of Turing nearly thirty years after his death, Turing became a figure of fascination both for his work in defining the function and limits of computational systems and for the ways he indexed a culture of early-twentieth-century sexuality and homophobia.³ This commingling of the personal, political, and technical in Turing's work begins with Hodges, but it has subsequently gained traction among researchers invested in queer history and Turing's influence on the political claims of modern computer science.

Turing's most noted work in this regard is his widely influential "Computing Machinery and Intelligence," first published in 1950 while he was a researcher at the University of Manchester developing some of the earliest modern digital computers.⁴ It is here that Turing first proposed the evocative question "Can machines think?" and argued in favor of machine intelligence through a reframing of thought as the successful performance of intelligent behavior. To make his case Turing proposed an imitation game that has come to be known as the Turing Test, whereby an examiner seeks to ascertain whether either of two unseen respondents is a machine based on their answers to a series of simple questions. Here Turing locates thought within a performative theory of intelligence, suggesting that if a machine can successfully emulate thinking by answering questions in a way that is indistinguishable from a human participant, then it has demonstrated a functional intelligence. Rather than weigh down this claim with ontological concerns over what thinking or intelligence are, Turing instrumentalizes thought as...