Scholarlink

There is widespread awareness of the potential undesirable consequences of AI. In this paper we explore the grounds of such concerns from the perspective of risk management. We propose a decomposition of AI risk into three sorts of risk (inertial, disruptive, fundamental) that can be approached in different but complementary ways. From this differentiation we advocate for Value Engineering as a pertinent approach to address fundamental AI risk.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

From the Pascal Wager to Value Engineering: A Glance at AI Risks and How to Address Them

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Pablo Noriega,
  • Pompeu Casanovas

摘要

There is widespread awareness of the potential undesirable consequences of AI. In this paper we explore the grounds of such concerns from the perspective of risk management. We propose a decomposition of AI risk into three sorts of risk (inertial, disruptive, fundamental) that can be approached in different but complementary ways. From this differentiation we advocate for Value Engineering as a pertinent approach to address fundamental AI risk.