用或然因果关系量化责任--以气候Action为例
Quantifying Responsibility with Probabilistic Causation -- The Case of Climate Action
摘要:
许多现实世界的伦理和经济相关性的情况下,如集体(在)行动方面的气候危机,不仅涉及不同的代理人,其决策以复杂的方式相互作用,但也有各种形式的不确定性,包括可量化的风险和不可量化的模糊性。在这种情况下,评估道德上不受欢迎的结果的道德责任或避免这些结果的责任是具有挑战性的,并且容易产生决策不足或决策过度的风险。与现有的方法,采用因果关系的必要性和充分性的组合或某些逻辑的基础上,专注于“负责任”与“不负责”的二元分类的概念,我们提出了一套定量指标,评估的概率单位的责任度。为此,我们适应广泛形式的博弈树作为框架,用于表示决策方案和评估建议的责任函数的基础上正确表示一组分析评估的范例场景。我们在真实决策场景的简化表示上测试最佳性能指标,并能够计算有意义的责任分数。
英文摘要:
Many real-world situations of ethical and economic relevance, such as collective (in)action with respect to the climate crisis, involve not only diverse agents whose decisions interact in complicated ways, but also various forms of uncertainty, including both quantifiable risk and unquantifiable ambiguity. In such cases, an assessment of moral responsibility for ethically undesired outcomes or of the responsibility to avoid these is challenging and prone to the risk of under- or over determination. In contrast to existing approaches that employ notions of causation based on combinations of necessity and sufficiency or certain logics that focus on a binary classification of `responsible' vs `not responsible', we present a set of quantitative metrics that assess responsibility degrees in units of probability. To this end, we adapt extensive-form game trees as the framework for representing decision scenarios and evaluate the proposed responsibility functions based on the correct representation of a set of analytically assessed paradigmatic example scenarios. We test the best performing metrics on a reduced representation of a real-world decision scenario and are able to compute meaningful responsibility scores.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。