“The AI Mirror: How to Reclaim Our Humanity in an Age of Machine Thinking” By Shannon Vallor
The Danger of Letting Go of 'Thinking and Feeling’
What is AI, to you? A mere tool for work, or perhaps a friend, even a lover? Vallor, a virtue ethicist, argues that AI is nothing more than a mirror reflecting our own "trajectory as predicted from the past." The major large language models are trained on vast datasets skewed toward white Western culture, and the distortions latent within that data are reflected back in our own image. A mirror shows only surfaces. That image, the author says, has "no sound, no smell, no depth, no softness, no fear, no hope, no imagination." AI cannot, in principle, possess lived experience. This is precisely why the threat AI poses comes not from outside us, but from within. What is dangerous is that, enchanted by flawless advice and ideal dialogue, we ourselves let go of the very process of thinking and feeling. If we hand over even our judgment and reasoning, then just as abandoning a map for GPS gradually erodes our capacity to find our own way, our very faculty for moral judgment — toward the world, toward others, and toward ourselves — withers. And yet we rarely notice this happening, because we mistake the image reflected in the warped mirror for our true selves. I feel this same unease acutely. The more I entrust my writing to generative AI, the more satisfied I become with its output, and the more my own thinking grows tame, domesticated. At universities in recent years, students' papers have seemed, on the surface, to improve in quality. But the writer's own touch has vanished, replaced by an increase in inoffensive, personality-less prose. One can see thinking being quietly smoothed into the average "desirability" that AI presents. So what is to be done? Virtue, the author says, is not an innate quality but a "moral intellectual muscle" trained through practice. What is needed is a system of which we would want to say, "it is alive, it is thinking alongside us, it is feeling" — one that does not let us lose sight of the human qualities a mirror cannot capture. Exactly what such a system would look like in concrete terms is left to the reader's imagination. We ought to avoid peering into the mirror only to trace and fix the past in place. Instead, we should envision an AI that helps us savor the meaning of the experience we live each day, and helps us transform into something better. Translated and supervised by Yohei Nishida; translated by Kako Ishigaki. (¥2,420)
