LECTURE 3: Deciding How to Decide Part 2: Human and Artificial Judgment in a Second Best World
The development of artificial intelligence and related technologies will likely keep advancing, bringing important new possibilities and raising intricate legal and policy questions. But we imperil the value of much that is positive in civilization, including language, respectful deliberation, and mental acuity, if we embrace the idea that the risks of artificial intelligence can be easily managed, that human judgment is so imperfect that it’s worth displacing as soon as reasonably possible, or that it’s a forgone conclusion that the benefits of the technology in any given context outweigh the costs. The burden of proof is on anyone who thinks it would be different with AI. Robust human judgment and deliberation is worth defending in a messy, second-best world of manufactured intelligence. Precisely because it’s imperfect, it forces us to specify what we value and why, and to fight for it. Too much convenience could make us far too docile even as the risks we create grow in the background.