I told my dissertation advisees to use ChatGPT. Here's what happened.

I told my dissertation advisees to use ChatGPT.

Higher education, as we know it, did not end. In fact, it might be better for it.

Using ChatGPT as a tool in the dissertation process is not cheating but rather an example of how AI can enhance the assessment of ideas and extend the possibilities for improving conditions in organizations. AI functions as a means of scaffolding student thinking and enhancing the conversations I have with my students.

As a dissertation advisor, I have utilized ChatGPT in "conceptual blockbusting" conversations with my students conducting organizational improvement studies using mixed methods action research (MMAR).

Here are two examples.

One of the key components of our students’ mixed method action research dissertations is the diagnosis phase. In this phase, students uncover potential challenges which could be addressed with action research. Students do this by a number of means, not the least of which is through stakeholder conversations. To assist my students in this process, we have used ChatGPT to generate conversation prompts.

Providing ChatGPT with a specific organizational challenge and a prompt to generate conversation starters can create a wide range of prompts that can be used to discuss the issue with co-workers.

In this example, ChatGPT generated conversation prompts which spurred my student’s thinking on where they could take the conversation with stakeholders. The student is still responsible for conducting the conversations, analyzing the data, and drawing conclusions based on their findings. ChatGPT is not making decisions or determining the outcome of the research. It provides a tool for students to generate conversation prompts and explore different perspectives.

Moreover, using AI in this way allows for a more thorough and nuanced assessment of organizational challenges. ChatGPT enables the student to consider different perspectives and think more critically about the challenge by providing a range of conversation prompts. This leads to a more effective diagnosis of the problem and, ultimately, better solutions for improving conditions in the organization.

The second example of using ChatGPT with a doctoral student was with a student at a more advanced stage in their study. This student had already conducted a thorough reconnaissance phase and collected mixed methods data on a problem at their college. This section of the dissertation concludes with a nuanced paragraph of "meta-inferences," or new and reasoned insights on the issue at hand.

The student and I used a draft paragraph of their meta-inferences as a prompt, asking ChatGPT to brainstorm potential actions the college could take to address the problem. The language model produced eight ideas, each with a sentence describing the idea. All the ideas were reasonable and feasible and provided a starting point for further discussions.

Using this list, my student began thinking of new ways to converse with the study stakeholders. They could explore which ideas might have traction and what new ideas stakeholders might have beyond the list. This helped the student to think more critically about the problem and consider different perspectives, ultimately leading to a more effective diagnosis of the problem and better solutions for improving conditions in the college.

These examples highlight how ChatGPT can be used to brainstorm and generate feasible ideas around a contextually bound problem, which human stakeholders can then discuss and refine. It shows how AI can complement and enhance human thought and creativity rather than replace it.