When I started this six-week-long series by asking whether my daughter would be attending college in 2035, the year she turns eighteen, I was pretty sure I already knew the answer. Nine years is a short amount of time, and something disastrous—or, I suppose, downright liberatory—would need to happen in the culture to make it unlikely for her to shuttle off to some campus after high school.Still, thinking about the future of higher education has convinced me that her path to a bachelor’s degree will be very different from the one I began in 1998. I want to end the series with a set of of predictions about what will happen to higher ed in the next nine years:The enrollment cliff—the result of a drop in births that started during the 2008 financial crash—will have a devastating effect on small private schools and regional public schools. A winner-take-all dynamic will emerge quickly, as hundreds of campuses will either be shut down completely or absorbed into larger institutions.The percentage of high-school graduates heading off to college will continue to decline. In 2009, seventy per cent of high-school graduates—an all-time high—went on to pursue some form of higher education. Today, the number is sixty-one. Much of that is simple regression after fifty or so years of unstoppable growth in the college-attending population, but I cannot see a single reason why that number would go up. Typically, a major recession boosts enrollment in higher education, as a lack of jobs leads young people to stock up on degrees while they wait for the market to improve. But, given the groundswell of outrage about the high price of college and the increased emphasis on student debt, it’s hard to imagine that a whole lot of broke, jobless young people would expect great value from a college degree.Colleges will continue to have a tenuous and frankly embarrassing relationship with A.I. companies. Some will announce “partnerships” to increase the robot presence on campuses, as the University of Chicago did last week; in STEM fields, in particular, this might lead to changes in the ways that classes are taught and that students learn. But the cheating problem will not be resolved any time soon; nor will professors in the humanities find many novel ways to incorporate A.I. into their pedagogy. At some point, those professors will have to accept that, while some students still want to read “The Brothers Karamazov” or contemplate a painting of the Buddha, others will be turning in A.I. assignments that the professors will have to dutifully pass, because the university can’t afford to lose any more students.Some struggling universities will take the partnership idea more seriously, and do more than buy subscriptions for their students. Think of the headlines: “The University of Vermont”—or wherever—“becomes the first state school entirely powered by A.I.” And a handful of new, for-profit, degree-granting institutions will open and simply offer guided A.I. tutorials. But the winners of the next nine years will continue to sell what they’ve sold in the past—whether it’s a fun football team, entry into the cultural élite, religious education, or radical politics for the children of law-firm partners.The concurrent crises in the reading, writing, and math competencies of incoming college freshmen—which, I have begrudgingly accepted, are real, and not merely the familiar griping you hear in faculty lounges—will get worse, and within a few years a course that used to be appropriate for tenth graders will become the standard 200-level course in many universities across the country.The prestige of the truly élite colleges will, sadly, not change all that much. Sometimes I entertain the fantasy that grade inflation and rampant A.I. cheating will prompt employers to seek out other forms of validation than Ivy League degrees. But that fantasy requires one to believe in another: that employers hire Ivy Leaguers because of merit and not because of network effects and cultural affinities.In good news, community colleges should be some of the biggest winners of the next several years. If A.I. and economic ruin eliminate the middle and lower middle classes of universities, more students might decide to get associate’s degrees and then transfer to four-year institutions. This shift will, I hope, happen higher up on the socio-economic spectrum and become normal even within liberal, wealthy, suburban enclaves. Some flagship state universities might also expand, whether by absorbing struggling campuses in their network or by adding more spots for undergraduates. The schools that do this will almost certainly be in the South and the Southwest, the two regions that face the least amount of demographic decline, in large part owing to their growing Latino populations.In other good news, the phenomenon of “administrative bloat,” in which a school hires hundreds of administrators and very few faculty, will abate, mostly owing to financial pressure. In boom times, it’s easy for administrators to add more administrators to the administration borg. But, when it’s cutting time, it’s easier to lay off student-success coördinators than it is to eliminate the handful of professors who teach chemistry to all of your undergrads.As a columnist, I’m almost embarrassed by the dearth of truly hot takes in this list. This is hardly Armageddon for higher education. But the future does kind of suck, and looking over this list makes me wonder if the more appropriate framing for this series would’ve been to ask about my future grandchildren rather than my children. Will they be attending college in the twenty-fifties or the twenty-sixties, when the attrition of the next decade or two has run its course, universities have consolidated, and we’ve come to some livable peace with technology? If my hypothetical grandchild grows up like my actual daughter, in an educated household in an expensive town surrounded by educated friends with educated parents, what will pull them to a campus? For me, it was an imperative: go to college or end up homeless on the street. Or so it seemed. For my nine-year-old daughter, in 2035, I imagine it will feel more like inertia: go to college because you might as well have a degree in an uncertain and changing world, where every day some A.I. company or another announces yet another industry that will soon be taken over by the robots.Inertia, though, is finite. If the university is no longer just the place you go to get a degree that gets you on a job, what is it? As far as I can tell after my inquiries during the past six weeks, almost no one has a particularly good answer to that question. In the coming years, as the pain starts at so many of these universities, their administrators, faculty, and those with a direct stake in the future of higher education need to figure out what a college education is for.My suggestion: It’s easier to come up with a collective vision for an educated population than it is to keep flogging exclusivity, élite credentials, and whatever else colleges sell these days. A.I. will, at minimum, shift the access points to all sorts of information, flatten much of the differences between formerly disparate forms of education—as the robot learns more, it’ll arrive at some bland but acceptable amalgam of all it has read, and everyone who relies on the robot will take that mush as the consensus—and radically change the way students work, especially in STEM fields. Under those conditions, can colleges still sell class mobility through an exclusive education? Or will they do better to look at the existing infrastructure of affordable community and regional public colleges, hear what young people and their families are saying about the cost of tuition and student debt, and then make a broader appeal for education for the sake of the public? ♦