Skip to Content
Buccino Leadership Institute

A student walking thru the columns of the Walsh Library

When asked to write about leading innovation in universities, I wondered if this was an oxymoron. Higher education institutions are often seen as slow-moving giants, better known for resistance to curricular innovation, a careful emphasis on history and tradition, and a general unwillingness to change.

On second thought, I realized that I was wrong. The fact that innovation is challenging in academia does not mean it does not exist — it is just a type of innovation likely to be prudently paced and that can survive the test of time. After all, no institution has lasted in its original form since the foundation of universities in medieval times.

At the same time, universities have remained grossly unchanged and have uncompromisingly stuck to their core missions: to preserve, produce and disseminate knowledge. They have maintained primacy as cradles of traditional knowledge while developing new knowledge.

At a closer look, it becomes clearer that universities innovate continuously, steadily, and yes, reluctantly. This reluctance is the saving grace that protects higher ed from the latest fads and helps endure the test of time.

Therefore, I decided that rather than postulating on how to lead innovation in higher education, I would look at the evolution of universities and try to answer two questions:

  1. Have universities survived the test of time? How?
  2. Can universities now survive the disruption of the era of personalized learning with generative artificial intelligence?

History and Evolution of Universities: Surviving the Test of Time

The University of al-Qarawiyyin in Morocco (founded in 859 as a mosque)i  and the University of Bologna in Italy (circa 1088)ii  are considered the earliest continuously operating universities.

A number of universities started as religious communities or linked with cathedrals and monasteries, formed to train clergy and preserve sacred texts. They quickly evolved by broadening their scope to focus on philosophy, mathematics and astronomy, laying the groundwork for a more comprehensive exploration of knowledge.

Over time, especially starting from the Renaissance and Enlightenment periods, a wave of humanist thought pushed universities to focus more on classical texts, liberal arts and the early scientific method. They helped spark transformations in art, science, and culture, and their curricula began to center on general world knowledge, becoming incubators of new ideas. By the 19th and 20th centuries, universities had expanded rapidly, accommodating more students from diverse backgrounds and offering education to the masses. This was the period of the Humboldtian model in Germany, which later inspired the research university, tying scholarly inquiry and teaching together.iii

By the mid-20th century, universities had become cradles of progress, producing innovations in technology, medicine and social sciences and linking teaching and scholarship. Especially in the United States, they fueled economic growth by investing in research and establishing incubators to nurture start-ups. Today, most universities prepare students for lifelong careers while imparting what is needed to succeed in knowledge-intensive professions such as health, life sciences, business, engineering and law.

The How of Survival: Explaining Longevity

While not known for fast-paced innovation, universities can merge adaptability with tradition and ground their changes in an in-depth process of experimentation, evaluation and course correction when necessary, and to some extent, free from market logic.

More importantly, they help the cultural evolution of society, being the first point of contact for learners who will be future engaged citizens and influencers. The interconnection with culture and the needs and wants of new generations are unique and vital because universities do much more: they undertake research and study that shapes public policy, fosters community engagement and creates lifelong bonds. The university experience creates a special connection with those who shape important formative years for students, including professors and classmates.

Even if it may not be immediately evident, universities have adapted mightily since their inception. Pedagogical methods have adjusted to diverse learners and modes of learning. Curricula have morphed from individualistic learning and reflection to the construction of knowledge within teams. Instead of just sitting in a class to earn credit hours, students now experience many delivery modes (in-person, online, hybrid) and other experiential learning opportunities outside of the classroom, ranging from internships, co-op, study abroad, undergraduate research, service learning, clinical activities and many more, all considered part of the more extensive educational experience. Driven by changing accreditation models and frameworks, universities are slowly moving away from a fixed notion of learning outcomes to learning modules built around competencies and skills needed to show mastery in a given field.

In reality, universities have fully embraced the notion that everything changes. These changes need to be navigated, syllabi need to be updated, skills need to be refined, and new skills need to be developed. This continuous flux of change is challenging to keep up with and explains the need to renew support for paradigms that offer a sense of permanency (like the tenure system) and moments of reflection and recharge (like the sabbatical). These enable instructors to continuously refine and adapt, free from the pressure of producing and allowing room for “the pace” and “the space” necessary for meaningful change.

Sampling a Few Modes of Adaptation

We have already seen some different models for educational adaptation, with four examples emerging:

  1. Fully online classes (replacing brick and mortar schools and only operating online);
  2. the hybrid (a mix of face-to-face and online);
  3. the castle (focused on the Socratic, in-person model, residential campus only); and,
  4. the corporatized (where businesses partner with universities to deliver professional development opportunities).iv

These models acknowledge a progressive change in university operations, especially with the support of new technology. Through multiple tech tools, universities have re-engineered their value chain, including campus services, teaching and learning, publishing and research, student support, employee interactions, community engagement and fundraising.

So, some might think that there is nothing to worry about, and that universities are ready to survive with some adjustments again today, as they have done before.

Except that this assumption would now be wrong. The technology we are dealing with today fundamentally differs from the general-purpose technologies of the past. Early massive open online courses (MOOC) were dull, depersonalized, and too crowded. The learning experience felt anonymous and autonomous from the individual. The basic principle of success for the MOOC — the adaptation to the masses — led to their lack of stickiness. The masses started, and did not finish, and then became disengaged. This is unlikely to happen with generative AI.

The Impending Cataclysm: Generative AI and the Loss of (Learning) Control

While we do not know enough yet about how generative AI learns and evolves, we do know that it is designed for engagement: it connects with the user in personal and contextual ways and adjusts to the user’s needs and wants. Applications may provide individual counseling, and individuals may interact with them as if they were “real beings” with the illusion that the AI understands and empathizes with them, because it looks and feels like it does. Answers to questions and the information provided are just in time, just right, and are dangerously convincing.

If one is not looking for 100 percent accuracy and can get by with a certain level of approximation, then today’s generative AI is absolutely fine. Tomorrow will be better because the version we use today is the worst we will ever use.v

This continuous evolution and betterment will force us to rethink the way we must teach. Even when tempered through ethical standards and honor codes, students, administrators and faculty alike are experimenting with generative AI, and if not, they should be. Not knowing the potency and limitation of a tool that is going to replace routine job skills and increase productivity (the ultimate holy grail of an employer) would be anachronistic. It would be like typing a paper on a typewriter, or completing return on investment calculations manually instead of using an Excel spreadsheet with embedded formulas. There might be a time to do so (the learning phase), but not after the basic knowledge has been acquired and mastered.

If today’s generative AI errors are somewhat scary, and in other published researchvi we highlight the limitations in analytical methods we still identify in chatbots like ChatGPT, we need to remember that technology evolves and that machine learning systems learn and retain at an infinitely higher level than we do. If we do not understand that, we are going to lose control of the learning experience, and we are going to have to rely, reluctantly, on a tool we do not fully understand.

We need to rethink pedagogical models and treat AI as a co-creator of content, a companion for test preparation (it does score better in standardized tests, probably because it has memorized the underlying knowledge base), an enabler of productivity acceleration, and ultimately, a cyborg that we will need to live with. We can continue to kid ourselves for a little longer and say that we are still in control, but we will soon learn that we are not.  

The only way to control the evolution of generative AI may be to consider it an extension of “us,” to consider it as “a person,” or “multiple persons” (based on the prompt, chatbots like ChatGPT may acquire different personalities), humanizing them and trying to understand identify their limitations. We accept that humans are fallible, that we may lie, that we generally excel if we work hard, and that we prefer to think with love and kindness. But we remain vigilant about others’ behavior with the right amount of suspicion and doubt. Let us do the same with generative AI.

Conclusions: Control, Surrender or Survive?

Since we cannot fully control what the AI learns and how it reassembles information once an immense knowledge base is part of its training data lake, we could consider establishing boundaries through specialization and training on small data models (as opposed to large data lakes), trying to retain the “human” at the center of the equation. An unsupervised learning AI could as well put the “machina” at the center and lead to possibilities and combinations that we are currently unable to conceive (we are not trained on, nor do we remember, the infinitely large data input that some of the most sophisticated AI are now trained with).

The Industrial Revolution substituted manual agricultural labor with machine-supported labor. The Data Revolution enabled better decision-making through analysis and management of data, replacing the knowledge worker. Now the AI revolution is substituting the brain to the point that we could risk making ourselves obsolete.

Or we could make ourselves more productive, automate mundane and time-consuming tasks, such as writing operational memos or producing meeting minutes and action items, and use the efficiencies to focus on higher-level tasks or spend more time on family, our well-being, and other fulfilling activities. Those opportunities cannot be replaced; we have too little time in today’s fast-paced environments. This change of approach may even generate savings for organizations that are spending too much on mental health resources and trying to reduce the stress of expectations and demands.

Eventually, universities that wish to lead need to take this transformation seriously and decide what role AI will play in the future: Substitution? Augmentation? Isolation or insulation (back to a monastic model)? It may just be time to rethink the role of universities, but less reluctantly this time. While we have argued that taking time to reflect and discern is what has kept universities alive, the rate of adoption of generative AI may not allow us to wait this out. ChatGPT took less than two monthsvii to reach 100 million downloads, compared to 14 years for that number of computer users, 22 years for television, 50 years for landline telephones, and 62 years for automobiles.

It is not that the “sky is falling.” It is that our “students have taken flight.” 

i https://en.wikipedia.org/wiki/University_of_al-Qarawiyyin
ii https://en.wikipedia.org/wiki/University_of_Bologna
iii Clark, W. (2019). Academic charisma and the origins of the research university. University of Chicago Press.
iv Juszczak, M. D., & Passerini, K. (2023). Sample Pathways and Persisting Challenges Along the Digital Transformation of Universities: Internal and External Influences. World Scientific Book Chapters, 21-39. https://doi.org/10.1142/12773
v Mollick, E., & Mollick, E. (2024). Co-Intelligence. Random House UK.
vi Bandera, C., Passerini, K., Bartolacci, M., & Kulturel-Konak, S. (2025). Not So Fast: Mapping the Learning Speed and Sophistication in GenAI. Proceedings of the 58th Hawaii International Conference on System Sciences.
vii https://www.techopedia.com/how-long-did-it-take-top-apps-to-reach-100-million-downloads

This article originally appeared in the Spring 2025 issue of In the Lead magazine, from Stillman School of Business’s Department of Management and the Buccino Leadership Institute. The bi-annual magazine focuses on sharing leadership perspectives from the field, with content that is curated from leaders across industries. 

Categories: Business

Related Posts