Learning to Play: Smart Toys, Data-Driven Friendships, and Technologies of Animation

Society for Literature, Science, and the Arts (SLSA) / Toronto / Nov 2018

Since the 1990s, the markets for technologically-enhanced interactive toys that “learn” alongside their child users have grown exponentially. From Tamagotchi digital pets to Hello Barbie, such toys have promised to facilitate responsive, meaningful, play with children, but are often met with considerable criticism over play value, privacy, and questions about what it means for a toy to think. Toys that call to mind notions of learning and growth affirm the dominant developmental paradigms that frame discussions of childhood and play, naturalizing products that might otherwise be regarded as radical deviations in childrearing and play practices by likening them to the developing child. Despite the new technologies integrated in such playthings, these metaphors of learning and growth are not unique to the connected age, and have a much longer history over the course of the twentieth century. This paper considers machine learning smart toys that engage in interactive and friendship-based play within a longer historical context to explore both their radical new potentials and the ways that they recast longer aspirations and concerns surrounding children’s play. I argue that such playthings are contemporary instantiations of the broader recurring trope of animate toys, or toys that come “to life,” an enduring goal that has animated the toy industry itself since the nineteenth century.