From The Chronicle of Higher Education to The Atlantic, The Washington Post, The New York times and everywhere I turn it seems, everyone seems to be talking–and worrying–about generative AI, such as ChatGPT, and its impacts on higher education. While academics and journalists debate about whether generative AI is good or bad, threatening to higher education’s mission, or an overly hyped non-issue, virtually everyone agrees that it is here to stay, and that we need to deal with it.

In this post, I focus on the four groups, who together constitute the “we” that need “to deal with it”: faculty, academic leaders, students, and industry leaders.

Faculty

For faculty, my biggest piece of advice is to get out in front of this and use your agency. What that means more specifically is something like this:

• Do not be reactionary. Accept that generative AI is already relevant for, and likely to be used throughout, your field, related professions, and your classroom. Prevent cheating to be sure, but the generative AI conversation should not primarily be about cheating, but rather the place of these emerging technologies in the future of your discipline.
• Develop an appropriate level of AI literacy. For most faculty, this does not mean that you have to learn about the role of k-means clustering in machine learning. Instead, it means knowing the differences among some different types of AI (e.g., recommender systems, image recognition, large-language models, classification systems) and a high-level conceptual (rather than technical) understanding of how they work. Additionally, I strongly recommend to spend a few hours knocking about with these systems in some kind of sandbox doing something you’re good at to get a sense for what they’re like to work with, what their capabilities and limitations are, etc.
• Approach the issue with an appropriate level of granularity. Instead of thinking about “what is the role of AI in higher education?” break the question down so that you are thinking about a specific AI technology and a specific disciplinary practice: how will large-language models influence writing essays about history? How might image generators contribute to digital advertising (or packaging design, or poster design, etc.)? How might image recognition of satellite imagery influence discoveries in minerals science?
• Talk to students and alumni. Find out from students how they are using these technologies, what they think about them, what excites them, and how they see them unfolding in their futures. Talk to alumni about what their organizations are looking for in terms of AI knowledge in new hires, whether and how they are using generative AI today, and what programs they have to train their workers about generative AI.
• Design your discipline and your whole curriculum, not just your class. It is up to your field, not your college administrators, to decide how and in which ways generative AI is best used in your discipline, in its related professions, and in the curriculum. Think big, and working together, take your agency as faculty to determine what ought to be done for the well-being of the field. Remember that scholarship of teaching and learning counts as research and can have a significant influence on your discipline.
• Be thoughtful and explicit about generative AI in your class. Talk to students. Make the boundaries between what is and is not acceptable super explicit. Be thoughtful and clear about what you put in your syllabus. Teach students how to attribute content to generative AI and why they should do it. Empower students to take ownership of work they did while using generative AI.

Academic Leaders (Chairs, Deans, etc.)

For academic leaders, the role is to set a vision and to empower the faculty to do what they do in light of it.

• Coordinate with other entities. Leaders may have peers at other institutions or other units within their own institution: What are they doing? What are the contents of the conversations (are they talking about what we’re talking about)? What is working/not working for them? What are the national higher education trends? Who is out in front on this–perhaps invite them to come in and present their work.
• Resource the process of redesigning/reimagining research, teaching, and service in this era. Leaders have access to many resources, from the bully pulpit to creating task forces, providing small grants, assigning professional staff, collecting and distributing knowledge resources and best practices, offering workshops and panels, holding town halls, and so on.
• Offer policy guidance. Offer clear guidance on what is/is not allowed (e.g., how generative AI plays with FERPA, courseware, etc.). Alert faculty to potentially hidden issues, e.g., the ways that paid accounts might give students who can afford them better tools with which to complete assignments, introducing an actionable equity complaint from students who can’t afford them. How does generative AI intersect with university IP policies? Can they require their students to use generative AI? What accommodations are available for these technologies?
• Encourage faculty creativity and leadership. Communicate to faculty that you are looking for and support creativity and leadership, using your bully pulpit to move the conversation beyond reactionary thinking. Those leaders who are closer to specific disciplines (e.g., chairs, more likely than provosts), encourage and even (as appropriate) participate in discipline-appropriate conversations and curricular changes. Incentivize creative ideas and innovations, never punishing (reasonable) risk-taking if things don’t quite pan out as hoped on the first try. For those who take up the challenge, reward them somehow, even if their efforts don’t yield traditionally measured outcomes (e.g., publications and grants).
• Set goals, aspirations, and expectations. Make them public, invite conversation and input, settle on one or more achievable goals, resource those working on it, and measure the outcomes.

Students

Students, you have more agency in this than you likely realize: show us what you can, and want, to do.

• Talk to your professors. Tell them what you think, wonder about, want to do, or think you should know about generative AI.
• While obeying course policies, throw yourself at generative AI. Try stuff. Be creative. Push the limits (of the technology, not of ethics!). Create clubs, run hackathons, and above all show your results. Help faculty see generative AI through your eyes.
• Attribute. If you use AI to help you generate any sort of content, disclose what you did (e.g., share your prompts), which technology you used, what its outputs were, and how you used those outputs in your work.
• Don’t cheat. I wish I didn’t have to write this, but I do. When you cheat, you cheat yourself, not your professor, or “the system,” or anything else. You cheat yourself and anyone in the future who works with you or depends on you.

Alumni, Recruiters, and Industry Leaders

Your primary job is to do your part to help higher education stay aligned with professions during this time of change.

• Communicate with universities about how generative AI is changing your work practices. Whether you are starting to hire different sorts of workers, are changing your work procedures, are changing your tools, and so on, communicate back to your higher education contacts. While job training is only one part of what higher education does, we all benefit when there is an appropriate alignment between higher education and professions.
• Help higher education understand future AI-related everyday competencies. If you are developing future tools, processes, products, or services for future consumers or end users who will need to have certain AI-related competencies, communicate those expected competencies to universities.
• Communicate with appropriate granularity. As noted earlier, “generative AI and higher education” can be unhelpfully abstract. Which kinds of AI are of increasing importance in what specific work practices or processes in your industry? What specifically do you wish graduates already knew or were at least positioned to learn?

If all four of these groups each do the sorts of things listed here, professions, academic fields, and individual careers will be advanced. Doing so will require practical partnerships, though, and none of us afford to wait for some other group to deal with this for us.

Acknowledgement: I participated in a panel on “The Impact of Generative AI in Our Classrooms” hosted by Josh Wede, Brad Kozlek, and Bill Goffe, with fellow panelists Jessica Driver, Larkin Hood, Wendy Mahan, and Tiffany Petricini. Their insights contributed to my thinking on this topic, which I gratefully acknowledge.