Is the Most Accurate AI the Best Teammate? Optimizing AI for Teamwork
Is the Most Accurate AI the Best Teammate? Optimizing AI for Teamwork
复制标题
最准确的AI就是最好的队友吗？
DOI：
10.1609/aaai.v35i13.17359
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Daniel S. Weld
中科院分区：
文献类型：
--
作者：
Gagan Bansal;Besmira Nushi;Ece Kamar;E. Horvitz;Daniel S. Weld
人工智能从业者通常努力开发最准确的系统，隐含地假设人工智能系统将自主运行。然而，在实践中，人工智能系统经常被用来为从刑事司法、金融到医疗保健等领域的人们提供建议。在这种人工智能建议的决策过程中，人类和机器组成一个团队，由人类负责做出最终决定。但最精准的AI就是最好的队友吗？我们认为“不一定”-可预测的性能可能值得在人工智能的准确性上做出轻微的牺牲。相反，我们认为人工智能系统应该以人为中心的方式进行训练，直接针对团队绩效进行优化。我们研究这一建议是针对一种特定类型的人-人工智能团队，在这种情况下，人类监督者选择接受人工智能建议或自己解决任务。为了在这种设置下优化团队绩效，我们最大化团队的预期效用，以最终决策的质量、验证成本以及人员和机器的个人精度来表示。我们在真实世界、高风险数据集上使用线性和非线性模型进行的实验表明，最准确的人工智能可能不会导致最高的团队性能，并通过考虑人类技能和错误成本等参数，在训练期间通过改善预期的团队效用来展示建模团队工作的好处。我们讨论了现有优化方法的不足之处，超越了已有的损失函数，如原木损失，并鼓励未来在人工智能协作的人工智能优化问题上的工作。
AI practitioners typically strive to develop the most accurate systems, making an implicit assumption that the AI system will function autonomously. However, in practice, AI systems often are used to provide advice to people in domains ranging from criminal justice and finance to healthcare. In such AI-advised decision making, humans and machines form a team, where the human is responsible for making final decisions. But is the most accurate AI the best teammate? We argue "not necessarily" --- predictable performance may be worth a slight sacrifice in AI accuracy. Instead, we argue that AI systems should be trained in a human-centered manner, directly optimized for team performance. We study this proposal for a specific type of human-AI teaming, where the human overseer chooses to either accept the AI recommendation or solve the task themselves. To optimize the team performance for this setting we maximize the team's expected utility, expressed in terms of the quality of the final decision, cost of verifying, and individual accuracies of people and machines. Our experiments with linear and non-linear models on real-world, high-stakes datasets show that the most accuracy AI may not lead to highest team performance and show the benefit of modeling teamwork during training through improvements in expected team utility across datasets, considering parameters such as human skill and the cost of mistakes. We discuss the shortcoming of current optimization approaches beyond well-studied loss functions such as log-loss, and encourage future work on AI optimization problems motivated by human-AI collaboration.
影响因子：
15.2
作者：
Patel, Bhavik N.;Rosenberg, Louis;Lungren, Matthew
通讯作者：
Lungren, Matthew
DOI：
--
发表时间：
2001-06
期刊：
--
影响因子：
--
作者：
B. Zadrozny;C. Elkan
通讯作者：
B. Zadrozny;C. Elkan