ABSTRACT
In order to introduce his famous “imitation game,” a test for determining human-level machine intelligence, Alan Turing first appealed to an analogy of gender passing. In this survey of AI fictions, we argue that questions of how to perform humanity have always been intertwined with questions of how to perform gender “correctly,” and, if to be read as transgender is in many ways to be read as inhuman, artificial intelligence narratives written by both cis and trans authors prove a particularly revealing space for exploring trans issues. Science fiction typically constructs and understands artificial intelligence through dichotomies such as natural/unnatural, real/artificial, genuine/simulated, and its plots frequently evoke the “rhetoric of deception” that transgender philosopher Talia Mae Bettcher argues surrounds transgender persons. An insistent “having to know”—to borrow Sedgwick’s phrase—hounds AI characters and interrogates their possession of human cognitive and emotional attributes or, indeed, specific gender identities, speaking to larger cultural anxieties surrounding gender transgression, as we can see in key AI fictions including Philip K. Dick’s 1953 story “Second Variety” and Richard Powers’s 1995 novel Galatea 2.2. As K. Surkan has argued in relation primarily to Spielberg’s film A.I.: Artificial Intelligence (2001), the android thus often becomes a trans passing figure, and nonbinary SF author Annalee Newitz’s 2017 novel Autonomous even evokes transgender identity explicitly in its depictions of AI, reflecting on the human need to gender consciousness and commenting on gender performance and construction. Authors who deliberately “trans” AI narratives can reframe the conventional dichotomies in ways that no longer relegate transgender identity to the abject, just as Susan Stryker has called for a reclamation of the unnatural in a reflection on that prototypical AI narrative Frankenstein.
