Moral AI

一本平衡且發人深省的指南，探討人工智慧與倫理的重大議題

電腦能理解道德嗎？它們能尊重隱私嗎？我們該如何確保 AI 安全且公平？

人工智慧革命已經來臨。如今，街道上有自駕車，軍隊中有自主武器，醫院裡有機器人外科醫生——AI 在我們生活中的存在只會越來越多。有人將此視為創新與便利的新紀元，也有人對其潛在破壞力感到警惕。但有一件事是清楚的：這是一項前所未有的技術，它引發了對人類智慧與道德定義的深刻思考。

在《Moral AI》中，世界知名的道德心理學、哲學與人工智慧研究者——Jana Schaich Borg、Walter Sinnott-Armstrong 和 Vincent Conitzer——直接面對這些棘手問題。他們以清晰、冷靜的筆調，呈現這一尚處於萌芽階段領域的最新進展，剝除誇大與誤導性的論述。書中深入探討 AI 核心的道德議題，涵蓋種族公平、個人隱私、假新聞到自主武器等範疇。最終，他們主張人工智慧可以被安全且合乎倫理地建構與使用，但其潛力的實現必須建立在對我們希望賦予 AI 的價值進行審慎反思的基礎上。

這是任何有思考能力的人都不可或缺的入門讀物。

Can computers understand morality? Can they respect privacy? What can we do to make AI safe and fair?

The artificial intelligence revolution has begun. Today, there are self-driving cars on our streets, autonomous weapons in our armies, robot surgeons in our hospitals – and AI's presence in our lives will only increase. Some see this as the dawn of a new era in innovation and ease; others are alarmed by its destructive potential. But one thing is clear: this is a technology like no other, one that raises profound questions about the very definitions of human intelligence and morality.

In Moral AI, world-renowned researchers in moral psychology, philosophy, and artificial intelligence – Jana Schaich Borg, Walter Sinnott-Armstrong and Vincent Conitzer – tackle these thorny issues head-on. Writing lucidly and calmly, they lay out the recent advances in this still nascent field, peeling away the exaggeration and misleading arguments. Instead, they offer clear examinations of the moral concerns at the heart of AI programs, from racial equity to personal privacy, fake news to autonomous weaponry. Ultimately, they argue that artificial intelligence can be built and used safely and ethically, but that its potential cannot be achieved without careful reflection on the values we wish to imbue it with. This is an essential primer for any thinking person.