In this paper, we examined two cohorts of students in the Digital Media Department at Drexel University who participated in the course Future of Digital Media: Seminar (DIGM 475). The first cohort took the course over the summer in an online asynchronous format, consisting of N=23 students, including N=20 undergraduates and N=3 graduate students. The second cohort took the course in the fall in an in-person format, comprising N=17 undergraduate students.

The course includes two main assignments: the first involves AI prompting using chatbots, where students generate essays with AI and then revise them. The second is the final project, in which students speculate on developments in digital media over the next 5 to 10 years.

The key difference between the two cohorts was the research approach. In the first cohort, due to the online asynchronous format, students conducted their research individually. In contrast, the second cohort, being in person, was required to conduct research in groups and was provided with a pool of topics to explore. The results showed that students in the first cohort primarily based their speculations on class materials from the first five weeks. Meanwhile, students in the second cohort often went beyond the class content, integrating multiple topics from the provided pool. In some cases, it was the integration of multiple topics from the pool of topics. However, there were no significant differences in student performance in the individual AI-prompting assignment between the two cohorts.

Keywords: Speculative Research, Digital Media Education, Generative AI in Education, Online Vs in-Person, Speculative Digital Media.