In the Mirror

Artificial Intelligence (AI) is now part of the daily news, with prominent public figures issuing dire statements about existential risk and governments scrambling to devise regulations. Lost in this conversation — which sometimes seems designed to induce fear and panic — is any discussion of what AI systems do and how they do it. The goal of this course is to step back from the headlines, reflect on AI as a human construct, and ask what it can teach us about ourselves. Chris suggests that by viewing AI as both a technological experiment and a cultural phenomenon, we can indeed learn a lot.

The first session, AI, language, and meaning, examines Large Language Models (LLMs) such as ChatGPT and their analogs in other domains. These systems naturally raise questions about the use and meaning of language, and of other human productions such as art and music. Such questions have been asked by philosophers for millenia, and by scientists for over a century. What does AI teach us about these questions?

The second session, Myths of human-like AI considers the history of “artificial humans” in mythic literature, and the relation of these myths to the ideas of “human-like” AI systems in academic research and in popular culture. Is human-like AI desirable? Is it feasible? What does the power of this myth — both attractive and scary — tell us about being human?

The final session, AI as a cultural artifact is a conversation with Eric Dietrich. We ask what it would be like to treat AI not as a threat or anything alien, but rather as yet another human cultural artifact that reveals something about who and what we are.

Recommended Reading

• Does Chat GPT-4 Pass the Turing Test?
• AI Wars, 1950–2000 and their consequences