Philosophizing with AI: happiness

A few years ago, I wrote a short text illustrating a problematic situation regarding the judgment of whether a particular fictional person, Bento, led a happy life or not. I frequently use this text in my introductory classes as a didactic resource to explain the nature of philosophy, its role in our understanding of the world, and to demonstrate its main challenge: the aporetic nature of philosophical questions. These questions do not yield unanimous or incontrovertible solutions; they always demand choice and engagement. The merit of this text (which will be reproduced in the dialogue below) lies in the fact that almost everyone, upon reading or hearing it, has an immediate and irresistible inclination to consider Bento's life either happy or unhappy. Opinions tend to differ on this matter. In all the countless times I presented it to an audience, the responses were never unanimous. Disagreements always arose, illustrating the aporetic nature of philosophical questions. As soon as I learned about ChatGPT and its extraordinary capabilities as an artificial intelligence (AI) system, I decided to test it with my text. Its answer to the question of Bento's happiness was remarkably clear; perhaps the most precise response I had ever received. I was genuinely impressed yet somewhat frustrated. Despite providing a rather sophisticated philosophical analysis of the situation, ChatGPT refused to make a choice, to decide whether or not the character led a happy life. It informed me that, as an AI system, it has no values and does not make choices. It neither likes nor dislikes; it does not prefer. This revelation shocked me because choosing, preferring, and deciding, even in the absence of complete information, are, in my view, not only the essence of philosophy but also the essence of human intelligence. Eager to see ChatGPT make choices, I adopted the strategy of asking it to assume a human persona, to pretend to be a person, and I began chatting with this individual. I reasoned that if it played the role of a person convincingly, it must make choices. And so, our philosophical conversation commenced.