18 Rapid proliferation of generative artificial intelligence (AI) tools forced many instructors to adjust their syllabus policies as a stop-gap measure, banning the use of generative AI. But detecting how much of a student’s work is generated by AI remains a challenge. Turnitin now includes an AI detector in its reports, but reviews are mixed. These policies are unlikely to be a deterrent to students if they know it is impossible for them to be caught. These policies also ignore the ubiquitousness of AI tools. The APA recommendation that students not only cite AI tools, as well as provide their AI prompt and its output, can only be achieved using external tools like ChatGPT. The integrated AI tools in Bing and Google search engines do not save both the prompt and output, and Microsoft’s Copilot, which will soon be available in many Office 365 applications, is designed to not require a prompt at all. It will offer AI suggestions on student work automatically, similar to the design recommendations in PowerPoint. Instead of focusing on policy changes, instructors must reconsider their course assignments and classroom activities. Educational research has shown that students retain the most from assignments that incorporate personal experience and peer interactions, neither of which can be convincingly replicated by artificial intelligence. This requires spending less class time on lectures and quizzing and more on smallgroup discussions and peer review. Rubrics for out-ofclass assignments that build on in-class discussions should address criteria such as student voice, personalization, critical thinking, and the use of sources. Faculty members should also consider revising their courses to focus on the process of learning. Tradi- tionally, students are assigned a project, and the instructor reviews only the final product. Instead, instructors can offer formative feedback on drafts or smaller portions of a large project submitted through- out the course. Instructors can also encourage students to use online writing tools that maintain a document history, so that students can review the evolution of their work with the instructor or with peers. Analytic thinking requires thorough consideration of a problem and all its possible factors and solutions. To develop this skill, students must think more slowly than they are used to. Additionally, the development of generative AI is a temptation that many will find hard to resist. By structuring assignments and in-class discussions to normalize uncertainty, exploration, and constructive critiques, faculty members will help students develop skills essential to analytic thinking. o CAROLYN FITZPATRICK, M.A. Instructional Designer IV, H. Wayne Huizenga College of Business and Entrepreneurship Fitzpatrick has more than 15 years of experience in designing quality online instruction. She holds a master’s degree in history and is completing a second master’s degree in instructional design. In We Can Remember It for You Wholesale, science fiction author Philip K. Dick imagines a world in which a corporation can artificially create past human experience, promising memories even better than “the real thing” since “actual memory, with all its vagueness, omissions, and ellipses…that’s second best.” Almost 60 years later, as we face the reality of the most sophisticated AI programs yet developed— ones that can seemingly replicate human writing and art—we are pushed to consider the difficult question of what really makes us human and why our own creations matter. It is uncomfortable to consider how human communication, for instance, may be reproduced without the actual human experience of writing and thinking— especially when, for many, it may seem that an AI program can write or draw Shark Point of View
RkJQdWJsaXNoZXIy NDE4MDg=