Generative AI Goes Back to School
Where Things Stand in Fall 2023
The commercial release of ChatGPT and other forms of generative AI created a firestorm this spring in and well beyond higher education. Other types of AI were already being used in university settings, particularly by researchers, but the sudden availability of consumer-friendly tools capable of generating responses to virtually any query transformed AI from a specialized tool into a mass-market product. The rapid spread of these easy to use tools raised important questions about teaching, learning, and research practices in higher education. Colleges and universities have spent most of the year playing catch up.
While the pace of adoption of flagship products like ChatGPT slowed over the summer, the need to develop policies and resources to manage the incorporation of generative AI is a significant strategic issue for universities. Last week, Ithaka S+R hosted the kick-off meeting for our Making AI Generative for Higher Education project, a collaboration with 19 institutions focused on developing evidence-based, proactive, institutional decision-making about generative AI in teaching, learning, and research contexts.
At the kick-off meeting, 88 individuals from 19 institutions came together to share details of their institutional responses so far and insights into the challenges and opportunities that will drive conversations about generative AI during the 2023-24 academic year. Several key takeaways emerged:
- Institutions are moving away from framing AI primarily as an academic integrity problem in favor of assessing its appropriate pedagogical uses.
- Universities will need to focus increasing attention on the place of generative AI in the research enterprise: AI methods and ethics are too important to the mission and reputation of institutions to be left to publishers.
- Responding to AI requires coordination across the university and across universities.
The importance of the last of these can’t be overstated. Generative AI is impacting all areas of the university. This is evident in the diversity of the individuals who participated in our kick off meeting, who represented units from across campus: libraries, centers for teaching and learning, student and academic affairs, graduate schools, research offices, IT, legal and tech transfer offices, and faculty from disciplines ranging from music to engineering. The institutions that successfully integrate generative AI will be those that put together the human networks required to think about the university and its strategic priorities as a whole while attending to the distinct missions and needs of specific units. Cross-campus working groups, committees, and collaboration are essential, and members of our cohort have created teams that reflect that.
The institutions that successfully integrate generative AI will be those that put together the human networks required to think about the university and its strategic priorities as a whole while attending to the distinct missions and needs of specific units.
However, institutions do not exist in vacuums. Their norms, practices, and policies exist in relation to those of individual disciplines, other universities, and large scale cultural, technological, and economic trends. Perhaps the most exciting part of our kick off meeting was seeing how much value came from the opportunity to share knowledge with peer institutions and from institutions that might often see themselves as having little in common. Throughout the kick off meeting teams busily took notes about and inspiration from what their peers were doing.
As university decision making around generative AI continues shifting from reactionary to strategic, the value of cross-unit and cross-institutional communication will only increase. The Making AI Generative for Higher Education project is one of several nascent projects facilitating this communication. The recent announcement of the Future of Law Libraries initiative, a series of regional roundtables bringing together law library stakeholders, advocates, and community partners to discuss both the risks and opportunities presented by AI technologies, is another.
Ithaka S+R will continue monitoring cross-institutional initiatives and services developed by individual universities throughout the 2023-24 academic year and will publish periodic updates of notable developments. This fall, we will release two important resources for institutions committed to strategic planning around generative AI. The first is a tracking tool focused on AI products marketed to universities, faculty, and students, and a catalog of policies, best practices, and guidelines around Generative AI. The second is an analysis of existing instruments and metrics for assessing the AI-readiness of faculty and students and entire universities, with recommendations for future refinement.
We welcome conversation with stakeholders about this topic. Please contact Dylan Ruediger (firstname.lastname@example.org) if you or your institution would like to learn more.