Lawless William F, Moskowitz Ira S, Doctor Katarina Z
Department of Mathematics and Psychology, Paine College, Augusta, GA 30901, USA.
Naval Research Laboratory, Information Technology Division, Washington, DC 20375, USA.
Entropy (Basel). 2023 Sep 11;25(9):1323. doi: 10.3390/e25091323.
In this review, our goal is to design and test quantum-like algorithms for Artificial Intelligence (AI) in open systems to structure a human-machine team to be able to reach its maximum performance. Unlike the laboratory, in open systems, teams face complexity, uncertainty and conflict. All task domains have complexity levels-some low, and others high. Complexity in this new domain is affected by the environment and the task, which are both affected by uncertainty and conflict. We contrast individual and interdependence approaches to teams. The traditional and individual approach focuses on building teams and systems by aggregating the best available information for individuals, their thoughts, behaviors and skills. Its concepts are characterized chiefly by one-to-one relations between mind and body, a summation of disembodied individual mental and physical attributes, and degrees of freedom corresponding to the number of members in a team; however, this approach is characterized by the many researchers who have invested in it for almost a century with few results that can be generalized to human-machine interactions; by the replication crisis of today (e.g., the invalid scale for self-esteem); and by its many disembodied concepts. In contrast, our approach is based on the quantum-like nature of interdependence. It allows us theorization about the bistability of mind and body, but it poses a measurement problem and a non-factorable nature. Bistability addresses team structure and performance; the measurement problem solves the replication crisis; and the non-factorable aspect of teams reduces the degrees of freedom and the information derivable from teammates to match findings by the National Academies of Science. We review the science of teams and human-machine team research in the laboratory versus in the open field; justifications for rejecting traditional social science while supporting our approach; a fuller understanding of the complexity of teams and tasks; the mathematics involved; a review of results from our quantum-like model in the open field (e.g., tradeoffs between team structure and performance); and the path forward to advance the science of interdependence and autonomy.
在本综述中，我们的目标是设计并测试开放系统中用于人工智能（AI）的类量子算法，以构建一个人机团队，使其能够发挥出最大性能。与实验室环境不同，在开放系统中，团队面临复杂性、不确定性和冲突。所有任务领域都有不同的复杂程度——有些较低，有些较高。这个新领域中的复杂性受到环境和任务的影响，而环境和任务又都受到不确定性和冲突的影响。我们对比了团队的个体方法和相互依存方法。传统的个体方法侧重于通过汇总个体可获取的最佳信息、他们的思想、行为和技能来构建团队和系统。其概念主要特征包括身心之间的一对一关系、脱离实体的个体心理和身体属性的总和，以及与团队成员数量相对应的自由度；然而，这种方法存在诸多问题，比如近一个世纪以来众多研究人员投身其中，但能推广到人机交互的成果寥寥无几；存在当今的复制危机（例如自尊量表无效）；以及有许多脱离实体的概念。相比之下，我们的方法基于相互依存的类量子本质。它使我们能够对身心的双稳态进行理论化，但也带来了测量问题和不可分解的性质。双稳态涉及团队结构和性能；测量问题解决了复制危机；团队的不可分解方面减少了自由度以及可从队友那里获得的信息，从而与美国国家科学院的研究结果相匹配。我们回顾了实验室环境与开放领域中团队科学以及人机团队研究；拒绝传统社会科学同时支持我们方法的理由；对团队和任务复杂性的更全面理解；其中涉及的数学知识；开放领域中我们类量子模型的结果综述（例如团队结构与性能之间的权衡）；以及推进相互依存和自主性科学的前进道路。
