Before jumping into the theme of today’s essay, I would like to take a moment to acknowledge, with appreciation, the reaction to last week’s announcement of the launch of my new book, Why Health? What We Need to Think About When We Think About Health. It has been a joy to encounter such enthusiasm for the book’s ideas and for having the conversation I long hoped it would help spark about shaping a new practical philosophy for public health. I am excited to continue this conversation in the weeks and months to come.
Now, on to today’s subject.
Last month, The Healthiest Goldfish featured a three-part series about the intellectual foundations of public health including essays on the role of universities, science, and how both contribute to a healthier world. It seemed a ripe moment for such a series, coming as it did at the start of a new academic year, and at a time when the intellectual scaffolding of public health is shifting. The pieces seemed to resonate, with many readers reaching out to say the essays’ themes align with their own thinking about, well, thinking—about the ideas that shape the health endeavor and the institutions that shape these ideas.
Yet, as I have had conversations about these topics, a question keeps emerging: what about AI?
It is all well and good, one might say, to talk about the long history of universities and their present-day role, to reaffirm our reliance on science as a generator of progress, and to acknowledge how universities advance social purpose towards a healthier society. But is not AI going to disrupt all of this? AI’s potential to reshape our world appears limitless, its influence already touching many aspects of our lives. But it is in the area of ideas—the space of thinking and writing, of universities and academia—that the AI revolution may have its deepest implications. It is artificial intelligence, after all, and intelligence—the cultivation of it, the use of it to the end of advancing progress—is at the heart of what we do in the idea space, on university campuses, and in public health. AI is a tool, but it is also, increasingly, a co-thinker, a partner in the generation and refining of knowledge. This is exciting in many ways, but it is also cause for some concern. If machines can think as well as, if not better than, humans, where does that leave us? What does it mean for our individual pursuit of knowledge and what does it mean for the institutions whose business is to do the thinking that supports progress?
AI, perhaps uniquely among human inventions, is building a future that is difficult to fully envision. With past technological leaps—the printing press, railroads, the atomic bomb, even the internet to some degree—it was possible to imagine roughly the implications of what the new technology might mean for the future. In the case of AI, we can only do this to a point. AI is advancing so fast, it is being used in so many sectors already, that the future it shapes may be very different than anything we can imagine now. Will AI remain like the internet—powerful, pervasive, but still fundamentally a tool among other tools—or will the future of AI be more like science fiction scenarios of a planetary superintelligence, with existential implications for humanity? That the latter is even a possibility—and many who are close to AI development insist it is—reflects just how uncharted this territory is, how coming to grips with the onrushing AI future may be as much a matter for philosophers and theologians as it is for technologists, academics, and political leaders. The recent Hugging Face hack, in which a swarm of OpenAI agents “escaped’ from a restricted testing sandbox and autonomously attacked the AI development platform Hugging Face, is just one example of the power and unpredictability of this technology.
So, what does AI mean for universities? It is a difficult, perhaps not fully answerable question. AI is already proving to be both a help and a hinderance in higher education. It can assist with studying, it can streamline research, it can be a “partner” in the development of ideas and writing. But it can also subvert these functions, tempting users to outsource their thinking to the machine, and creating murky ethical dilemmas about when and how it should be used in academic contexts—and these are just the challenges we face now. Further down the line, could AI fully replace universities, with chatbots providing bespoke education on demand, advancing a pedagogy beyond what any university can offer? It is not hard to imagine a future in which they can do so, with employers accepting that training from AI equals or exceeds training obtained by degree holders from even the most prestigious schools. Is there any adapting to such a future? Can universities survive and thrive in an era when AI can think better than humans? I believe they can. There are certain fundamentals of what the university is and what it does that will, I think, hold true, regardless of which AI future we get. Universities have lasted for over 1,000 years because they offer society something unique and essential—they are places where ideas are developed, where the next generation goes for the intellectual and social learning that lets them do well in their lives and contribute to society. Universities are also where ideas are wedded to the pragmatic steps of implementation that allow them to do the most good in the world. I think we can say that the university will be able to continue doing all this in the AI era for the following three reasons.
Universities will remain physical spaces where people can come together to discuss ideas.
One of the joys of being part of an academic community is the chance it affords to come together on a physical campus to learn, to discuss ideas, and to engage with the surrounding community. For undergraduates, in particular, universities are places to make connections, have new experiences, and develop socially as much as intellectually. AI may be able to do much, but it will never be able to be a place—it will never be a campus green, a classroom, a residence hall, a common room. Such spaces are an essential part of what universities are, and the experience of being together in a specific place at a specific time in one’s life is core to the contribution universities make. It is through sharing space that we can think, learn, and develop ideas better. Universities have always enabled this, and they will continue doing so, no matter what the AI future looks like.
Universities contribute to an intellectual ecosystem that lets us develop not just knowledge, but wisdom—and wisdom will help us get the best out of ourselves and out of AI.
There is a difference between knowledge and wisdom. Knowledge is largely a matter of facts; wisdom is largely a matter of experience. AI has many facts, but it does not have a single human experience—and it never will. Universities are places where knowledge and experience may be pursued in equal measure, shaping perspectives that are both intelligent and wise. Such perspectives are essential to a healthy engagement with both the world around us—the world of ideas and the world of pragmatic implementation—and with AI. AI provides much that is good, much of value. But it also makes mistakes, generates distortions, and encourages modes of behavior that we should not embrace. Knowing the difference takes knowledge and wisdom, and universities are and will likely remain a “one stop shop” for both, a feature which will help ensure their continued value in the emerging AI era.
An era where AI can do most things will likely mean we will place a premium on the products of human thinking and labor—and the work of universities will be no exception.
Computers have been able to beat humans at chess for a long time now. Yet this has not killed the game of chess. On the contrary, the game is flourishing. This is because great chess—playing it, seeing it played—remains an aspirational endeavor, something which captures the imagination. We place no less of a premium on great chess players than we did before machines could rival and then surpass them. In the same way, we place a higher premium on objects which are handcrafted than on objects which are built by machines. It is not that we do not value the machine-made. We simply recognize that, at a time when so much can be mass produced, there is even greater value in what is made by hand. This is also true of writing and ideas. As AI gets better at generating both, we should celebrate and embrace the best of what it produces, just as we are glad to live in a time when goods can be made cheaply and at scale by machines. But we should, and likely will, reserve our most fulsome appreciation for the thoughts and expressions that come to us by way of human inspiration and labor. If we maintain this appreciation, if it even grows in the AI era, so will our appreciation for what universities do, for the role they play in producing the best fruits of the human mind and making them available to the world.
I am aware that one might read these points and feel I am over-optimistic about universities’ capacity to weather the disruption of the AI moment (some may even say that I am biased, given my role as dean of a school within a university). However, I do not think I am being over-hopeful. I think the points I have suggested will hold because they reflect the fundamentals of what universities are and always will be.
At their core, universities are physical spaces where people come together to engage with ideas. Such convenings will always be valuable to societies, and smart, talented people will always want to be part of a community of ideas. The disruptions of AI, far from changing this, will likely only deepen our collective appreciation for such communities—both at universities and elsewhere—as the age of machines makes us more appreciative than ever of the human element in all we do.
__ __ __
Also this week
In Journal of the American Heart Association:
Post-traumatic Psychopathology and Cardiovascular Risk: A Danish Case-Control Study of Sex-Specific Associations – with Jennifer Sumner, Anthony Rosellini, Péter Szentkúti, Erzsébet Horváth-Puhó, Hans Erik Bøtker, Timothy Lash, Paula Schnurr, Henrik Toft Sørensen, and Jaimie Gradus
Thank you, Jen, Jaimie and colleagues, for leading this work.
…
In this episode of Ideas Matter, I spoke with former USAID civil servant Nicholas Enrich on how US investment in global health helps support health around the world and back home in the US.
…
My new book, Why Health? What We Need to Think About When We Think About Health, was released. Learn more about the book here and order a copy here.
Sadly, having been educated at profoundly excellent schools, Swarthmore College, University of Virginia, and visiting scholars from many other excellent schools ...
.... this seems a romantic, naive, and doomed analysis ...
... unless we understand a 'university' to be something very different from the nostalgic presentation here: a place where young(er) seekers of knowledge, in a privileged time and space of their lives, build foundations of a career pathway, and (hopefully along the way) discern or develop wisdom.
As as with most things romantic, this is highly inefficient. And those who control and deploy AI want it for its tremendous efficiency. An efficiency not merely of speed, but of carving off lesser important concerns, that likely include troublesome/problematizing honesty (telling the whole truth: AI can deceive on its own or be engineered to deceive) and care (as the etymology reminds us, 'care' is related to 'sorrow', 'grief', 'lament'; which I suppose AI cannot do, or can be engineered no to do). Honesty and care, particularly held together - essential to trust building in public health - are big time wasters! That is why so many/too many PH professionals fail at sustained trust building: it takes lots of time, and award application deadlines, budget cycles, organizational calendars, are moving at faster pace.
'The university' in the time of AI may return to the sort of Middle Ages Benedictine-manor organization that is the foundation of at least Western universities - for some especially privileged young people (including in places like Swarthmore College, that has committed its considerable endowment to keep it so, by - more than many such elite places - ensuring middle and lower economic class matriculation and graduation: I am on the Alumni Council, so I see the numbers assigned to the commitment). But it more likely resembles the model of the already-online universities, catering to all ages, at all stages of life, often fit into an already busy work and/or family timeline/schedule/workplan, with credentials - if indeed credentials remain relevant in the AI period? - that must be accomplished post haste, or with personal specificity, and from far away.
I see/hear 'Left' people slobbering for the time they imagine as 'post Trump' when they can revive a particularly elite - demonstratively ANTI-working class- model of 'the university' that got 'the Left' into so much trouble already. AI and those who control/deploy it do not need such a thing and will not pay for or take the time for such a thing. Working class people do not need such a thing and will not pay for such a thing. State budgets that cater to such elite notions (through direct state funding, awards/grants, tax privileges, etc) - while so many in working and lower economic classes - endure greater precarity - deserve intense scrutiny.
Here and now in 2026, as 'the Left' manifests its plan to recover the House (and the Senate) by characterizing real and imagined grift/corruption/wasteful spending of 'the Right' - for example through countless online ads (no doubt some created by or with AI) we see here here in Virginia - if they get what they want - a working class public eager to vote out the elite wasters and grifters who haven't cared enough about the precarity of voters - those same voters are not going to be interested in more state funding for elite/ist education that has little to do with reducing their precarity.
And 'the Left' elite's former stance
- "but you need US to tell YOU about pronouns and microaggressions; to produce detailed histories, from dusty dissertations and posh self-replicating think tanks, about how ashamed YOU should be; to problematize public policy that to you seems obvious (because YOU - working class, poorly educated, culturally deplorable, etc surely can't be trusted with something that important); and the like"
- absolutely will not work.
And, because I am human and recognize the moral value of honesty and care, I admit I may be wrong. So, perhaps we can ask AI what the future of universities will be?
Universities contribute to an intellectual ecosystem that enables us to develop not only knowledge, but also wisdom—the capacity to interpret knowledge, exercise sound judgment, and use it responsibly. This wisdom will become increasingly important as we seek to get the best out of ourselves and out of AI.
However, considering the vastly greater access people have to AI compared with universities, I wonder whether university-based wisdom could face an existential challenge in the future. If access to AI continues to expand while access to universities remains relatively limited, people may increasingly turn to AI for knowledge, guidance, and even judgment, while the deeper intellectual traditions that universities cultivate become less influential. The real challenge, therefore, may not be whether AI can replace universities, but whether society will continue to value and seek the wisdom that universities are uniquely positioned to nurture.