BriefGPT.xyz
Nov, 2023
AI对准的基础道德价值
Foundational Moral Values for AI Alignment
HTML
PDF
Betty Li Hou, Brian Patrick Green
TL;DR
解决人工智能对齐问题需要明确且可靠的价值取向；本文提出了从道德哲学中提取的五个核心、基础价值观，即生存、可持续的代际存在、社会、教育和真理，并表明这些价值观不仅为技术对齐工作提供了更清晰的方向，还作为一个框架来突出人工智能系统对获取和维持这些价值观的威胁和机遇。
Abstract
Solving the
ai alignment problem
requires having
clear, defensible values
towards which AI systems can align. Currently, targets for alignment remain underspecified and do not seem to be built from a philosophica
→