In this episode, I explore the "artificial I" — the sense of self that feels so solid and obviously real, yet works a lot like the AI systems so many of us are worried about right now. Drawing on my work in data infrastructure, my studies in applied AI, and the Noah AI tool I built, I explain in plain terms how AI learns from training data and predicts the next word, and then show how our own minds were trained on family, culture, and early experiences we never chose — like the child who learns love has to be earned and grows into an adult who can't rest. I connect this to the Buddhist teaching of no self through the word "artificial," which originally meant "made by skill" rather than fake, and the Pali word sankhara (formations, conditioned things, literally "put together"): the self isn't fake, it's made, and what's made can be seen, questioned, and even remade. I look at the three things people fear AI will do — deceive us (hallucination, illustrated by the split-brain "chicken shed" experiment), manipulate us (the predictive mind and the teaching of the two arrows), and make decisions without our consent (habit) — and show that the self has been doing all three for as long as we've been alive. Using the image of a megaphone and the three poisons of greed, hatred, and delusion, I suggest that AI amplifies whatever mind is holding it, which makes an unexamined self dangerous to ourselves, to others, and to the planet. Mindfulness, then, is interpretability turned inward, and the practice for this week is simple: when a strong reaction shows up, name it as a prediction ("my mind is telling me...") and ask, "What was this trained on?"
Explore more at eightfoldpath.com: full transcripts, guided meditations, courses, and Noah AI, an AI you can chat with about any episode or teaching.
Hosted on Acast. See acast.com/privacy for more information.