INTELLIGENT FAILURES: CLIPPY MEMES AND THE LIMITS OF DIGITAL ASSISTANTS
Keywords:memes, virtual assistants, humor, digital assistants, artificial intelligence
Often hated during its lifespan in product (1996-2006), Clippy – Microsoft’s Office Assistant, became a pop-culture icon in its afterlife. Delving into the plethora of memes featuring Clippy, we ask: why should a questionable character from a software program that has been out of use for well over a decade have so vibrant an afterlife? If Clippy has become a rhetorical resource, what is it being used to do? We propose that Clippy’s dual status as the original natural-language digital assistant, one that fell critically short in its ability to actually assist, makes it an ideal vehicle for critique of today’s ubiquitous assistants. An analysis of 1,148 meme instances collected from five sites led to a twofold argument: First, Clippy humor relies on the contrast between types of intelligence; Clippy is often too good at one kind, while lacking in another. In particular, Clippy lacks interpersonal intelligence: it serves as a disruptive mediator between its user and the world, as well as other human beings.
Yet this failure in “knowing its limits” and adapting to its environment is also what gives Clippy character. This suggests that digital assistants must attend to multiple kinds of intelligences; attending to any one over others may create an endearing character but not an effective digital assistant. Furthermore, the fact that the unbending yet personality-filled character of Clippy remains ungendered or male gives us insight into the pliant and empty characters of the female gendered Alexa, Siri, and Cortana.