An apprenticeship model for human and AI collaborative essay grading

Across a range of domains, humans and complex algorithms embedded in systems are collaborating and coordinating action. Tasks are shared because the combined system can be more capable than either agent acting alone. Such systems with shared autonomy raise important research questions in how to design these joint systems to best utilize each of the actors capabilities for optimal performance while addressing important safety, legal and ethical issues. Our work investigates students developing their writing ability throughout their educational career, since writing proficiency is a critical life and career competency. Writing is a skill that develops through practice and feedback. However, the massive effort required of instructors in providing feedback on drafts and scoring final versions is a limiting factor in assigning essays and short answer problems in their classes.

在一系列领域中，人类和嵌入系统的复杂算法正在协作并协调行动。任务之所以被分担，是因为组合系统可能比单独行动的任何一方更有能力。这种具有共享自主性的系统引发了一些重要的研究问题，即如何设计这些联合系统，以便在解决重要的安全、法律和伦理问题的同时，最佳地利用每个参与者的能力以实现最佳性能。我们的工作研究学生在其整个教育生涯中写作能力的发展，因为写作熟练程度是一项关键的生活和职业能力。写作是一种通过实践和反馈得以发展的技能。然而，教师在为草稿提供反馈以及为最终版本评分时所需的大量精力，是在课堂上布置论文和简答题的一个限制因素。