让AI具备人类智慧的关键是提升其元认知能力。
Imagining and building wise machines: The centrality of AI metacognition
- 将人类智慧拆解为元认知策略与具体问题解决方法
- 元认知缺失导致AI在新环境易出错、难解释、不安全
- 适合关注AI可解释性与安全性的研究者阅读
尽管人工智能日益智能，但其智慧水平并未同步提升。本文探讨了人类智慧的已有认知，并勾勒出人工智能智慧的愿景。我们将人类智慧视为应对难以解析问题（超出分析技术范围）的一系列策略，包括对象层面的启发式方法和元认知层面的策略，如智力谦逊、换位思考、情境适应性等。我们指出，当前AI尤其缺乏元认知能力；若能增强元认知，将使AI更适应新环境、对用户更具可解释性、更善于协作，且减少与人类目标错位的风险。文章还讨论了智慧型AI的评估、训练与实现路径。
原文摘要 · Abstract (English)
Although AI has become increasingly smart, its wisdom has not kept pace. In this article, we examine what is known about human wisdom and sketch a vision of its AI counterpart. We analyze human wisdom as a set of strategies for solving intractable problems-those outside the scope of analytic techniques-including both object-level strategies like heuristics [for managing problems] and metacognitive strategies like intellectual humility, perspective-taking, or context-adaptability [for managing object-level strategies]. We argue that AI systems particularly struggle with metacognition; improved metacognition would lead to AI more robust to novel environments, explainable to users, cooperative with others, and safer in risking fewer misaligned goals with human users. We discuss how wise AI might be benchmarked, trained, and implemented.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。