The Mind’s I — Fantasies and Reflections on Self and Soul

Recommended by Julian De Freitas, Assistant Professor of Business Administration, Harvard Business School

I’m revisiting this visionary work, which anticipates “centering” — the act of locating oneself as a particular agent in space, time, and possibility — in human cognition. This notion lies at the heart of a recent paper I have submitted, Reverse-Engineering the Centered Self, where we develop a computational framework for how humans solve problems by first inferring who and where they are in the world. Crucially, this capacity for self-orienting appears to set us apart from current AI systems: While humans rapidly identify themselves in unfamiliar environments, allowing them to continuously adapt to new problems, even the top reinforcement learning algorithms struggle with this task and fail to generalize when the setting changes. This discovery has profound implications for where organizations should employ automation, as well as for ongoing efforts to engineer more humanlike AI.