Will universities exist when today’s infants are 18?
August 19, 2025
By Matthew Pietz
This article was written and edited without the use of AI.
If my family were expecting a baby right now, I would not open a tax-deferred college savings account (commonly called a 529 plan).
One of the consequences of AIs exceeding human intelligence in 5-10 years, around which there is growing consensus (though no certainty) will be to shift the locus of higher education away from universities. Even if they radically transform, the current cost of a US college degree may no longer be worth it by the time today’s tykes are grown.
A university education has long been a cornerstone of the transition to adulthood, for those who could afford it, so the idea of that era ending may sadden some. It saddens me. But I think some universities will continue in a new form, and other institutions may fill the gap, as discussed below.
The three main benefits we seek from college are:
The degree, a powerful token in the job market
Knowledge imparted by professors
Life experience, in an environment specialized to that purpose
The all-in cost to get these benefits, for an average bachelor’s degree in the US, is about $150,000. Already, 40% of adults feel a college degree is not important for getting a job. That number will go up as a result of AI, while the willingness to pay such an already-painful cost goes down. Here’s why.
AI super-professors. If and when AI agents become more intelligent than any human, they’ll also be supremely talented educators, with fun, useful and interactive lessons fully tailored to each student’s career goals and learning needs. I doubt it will be kids passively sitting in front of screens, either. There will be hands-on assignments and deep conversations, and virtual reality can provide a “classroom” environment for those who want it. I feel most parents will still want human teachers for younger children, but for 18+-year-olds, you can’t beat both free and better than a professor. (“Cognitive offloading”, humans becoming less capable as they use AIs, is a serious concern, including for employers. But an AI education does not mean using AI to complete traditional schoolwork. Quite the opposite: AIs will be able to elicit responses directly from each student, whether during class or homework, so students must use their own brains to answer.) Equity of access may be much like access to AI is today, available to anyone with internet and a device, both of which will continue to drop in cost.
Devalued degrees. As the super-professors arrive in every home, university degrees may matter less. A certificate showing AI course completion may not hold much value either, as they’ll be easy to fake. Employers could instead use bespoke tests to verify a candidate’s potential (along with the resume and interview). This was always a good idea in theory, but costly or complex to customize to each job, which AIs will help do. Subtle and well-structured simulations could also test for leadership ability, critical thinking and soft skills. It’s true that today a degree also signifies responsibility, a work ethic, a seal of approval from a respected institution, and in a sense, membership in an elite club. But at the end of the day, the question is who can do the job, and economic pressures will probably force most employers, for most jobs, to hire the person who demonstrates the best performance. That may be the person who had the best customized training. Even if there are exceptions to this, the overall trend will likely be less emphasis by employers on degrees.
A different kind of life experience from universities. Institutions, like living things, seek to persist. They want to grow, to propagate, and they can fight tooth and claw to stay alive. Universities will be keen to find new sources of income. Meanwhile, families may still want to give young adults a college-like experience, a formative adventure away from home. It’s possible some colleges may reinvent themselves as “adulthood incubators”, where young people could live together and carry out hands-on leadership, emotional wellness, or physical achievement tasks, all designed around human interaction. Or campuses could just be a place to live during a first internship or job–a communal space with low rent and cheap food, with engaging events continually on offer. Incubators need not be on campuses, but could pop up anywhere. There are many ways young people could cohabitate in a nurturing environment during these formative years.
So, will universities exist?
Not in any form like they do today, in my opinion, as their income from current sources will precipitously drop.
Even if incubators or something like them arises, attendance will doubtless be below current college enrollment, and without human professors, today’s tuition simply doesn’t make sense.
Moreover, human PhDs may no longer be the key generators of new knowledge, making it more difficult for universities to secure research grants. A portion will thrive, working on frontiers like AI development or synthetic biology, but many of the 6,000 colleges in the US, especially the smaller ones, will struggle to pivot. And even remaining researchers may eventually be replaced by AIs (with a whole other set of consequences).
Institutes of higher learning are already facing headwinds in the United States, with investigations and threats to cut federal funding, while immigration restrictions send the world’s brightest minds to other countries. Even if a new administration were to reverse course in 2029, the process I describe above may be well underway.
But universities, and especially the larger, wealthier, and more elite schools, are not institutions to go gently into the good night. They are deeply entrenched in the upper echelons of society, gatekeepers of access and nodes of powerful personal networks. They will fight the arrival of AI professors and researchers, though I believe they’ll eventually lose those fights, and try to find ways to raise the value of the degrees they confer, or whatever replaces degrees. However they find ways to survive, I believe they will not emerge looking as they do today.
About that 529
In families well-off enough to save, shortly after a couple announces their pregnancy, expecting grandparents start to ask “Have you opened the 529 yet?” That may not happen much longer.
What if I’m wrong, and it takes 20 years for AI to surpass human intelligence, or it never does? Doesn’t it still make sense to put aside something for college? Certainly, but I wouldn’t use a 529 if I were a new parent, because of the 10% penalty (on top of tax owed) for using it for non-education expenses. Even though 529 funds can be spent on various types of learning, not just university degrees, the chance that these will cost a significant amount of money in 18 years–when AIs and YouTube can already teach you so much, virtually for free, now in 2025–seems so small that I, personally, wouldn’t take the risk.
If universities do fundamentally change, and lose their current place in society, the disruption will be stressful for many of us. But every major revolution humanity has undergone entailed such disruption. We will be disoriented, but new opportunities and types of community may arise to fill the societal gaps left by colleges, and our young people might be better-trained than ever.
What kinds of jobs will exist for them to train for? That’s the subject of a future post.
If you would like to speak to Keranaut about planning for education as the world changes, please contact us here.
