Between Virtuality and Humanity: AI, Shakespeare, and the History of Character

Why are chatbots often described in the same terms as Shakespearean characters? Attempts to categorize the ontological status of Large Language Models and other AI systems—“counterfeit people,” “virtual humans,” “artificial persons”—resonate strongly with tech-inflected identifications of Shakespeare’s dramatis personae—“virtual persons” (Crewe), “automat[a] ... that perform[] humanity” (Kolb), and, again, “artificial persons” (Barroll). This article suggests that descriptions of artificial intelligence on the one hand and of fictional characters on the other share deep roots in foundational early-modern attempts to express both the virtuality and humanity of imagined persons. I argue that recent efforts to articulate what it is chatbots do to, with and for humans have implications for how we understand explorations of the nature of dramatic character in the sixteenth and seventeenth centuries. Shakespeare offers particularly rich examples when investigating the conceptual overlap between artificial persons and dramatic persons. As dominant models for theories of character across centuries, Shakespeare’s characters have long been saturated in the language of artificial intelligence and imitation, and they retain a ghostly presence in modern AI discourse.

I draw on two recent experiences thinking about and with chatbots which have shaped my thinking about artificial intelligence. One is my ongoing, idle conversation with a chipper and ethical AI assistant, Anthropic’s Claude. The other experience is making a short film with Creation Theatre—a digital theatre company—about the efforts of a fictional chatbot, BardBot, to research a student’s Shakespeare essay, leading to a chaotic debate between AI-models of Sigmund Freud, Sarah Siddons, and L.C. Knights. I finish by discussing this film as a teaching resource and a thought experiment which teases out the conceptual stakes explored more formally in the article.