Universities are in trouble. Tuition is sky-high but salaries for adjuncts—one of the largest teaching classes—are abysmal. Amid the discontent adjuncts are forming unions to bargain for higher salaries, but this might not solve the greater problems plaguing the higher education system. But what few recognize is that these problems are just a mirror for our larger society as it wrestles with the future of technology, particularly the rapid rise of AI. How do the problems within universities warn us about the greater dangers in our AI future as well as suggest any potential solutions?
First, a quick note about the problem at universities. Progress in society, technology, and economy is founded on knowledge, much of it developed in universities. But as universities, particularly in the United States, have slowly morphed from public institutions dedicated to fostering such knowledge to increasingly mimic private enterprises competing for students, changes have ensued with negative implications for faculty and society. Perhaps the most problematic is the dramatic increase in adjunct faculty—professors who teach the same classes, but do not have the same research obligations or pay as tenured faculty.
For the university, these “gig workers” are attractive because they are dramatically lower cost and teach several times more classes than a tenured faculty member . While this efficiency game is attractive for universities in the short run, it comes at a horrible cost to adjuncts who must exhaust themselves teaching vastly more classes but leaving no time for the research required to win a coveted tenure-track position. What’s more, their contracts are almost always uncertain, creating anxiety about whether they will be employed the next semester, and sadly they are often treated as second or third-class citizens with no voice in decisions and little respect or prestige.
No wonder then that they want to fight back by unionizing and demanding the modicum of respect they deserve. Meanwhile, the university feels it has little room to move as bloat in student benefits and administration, as well as decreased public funding support, have made balancing budgets challenging. Although the adjunct “gig” solution might work in the short run, in the long run, it hurts the adjuncts, the university, and the larger society as a class of smart and talented people are undervalued, and thus underproduce, the knowledge on which our future prosperity hinges.
Although it feels like a catch-22, we forget that this outcome is a choice about how we create knowledge, encourage progress, and ultimately share prosperity. Having spent time in universities in the U.S. and Europe, it is clear there are other models. Many universities in Europe are either free or a fraction of the cost of their U.S. counterparts, and while many of my compatriots like to believe that U.S. universities are an order of magnitude better, this is not necessarily true. The mathematical training in Dutch universities, for example, is world-class and costs pennies by comparison. It is true that the system does not have as many student perks, supports smaller administration, and has not entirely solved the adjunct problem, but it does suggest an alternative model.
What then does this have to do with AI and the future of society? The problem at universities is the creation of a two-class system with lower-class, lower-paid “gig” workers with few rights who are controlled by a smaller, well-paid group with the power to make the decisions. As the inequity becomes greater, history teaches us with exacting clarity that suffering, conflict, and eventual revolution ensue.
Outside the university, the U.S. has some of the highest income inequality among all the developed nations and at the same time some of the lowest income mobility. In other words, like the university, if you are at the bottom you are screwed and there isn’t much you can do about it. These differences, as well as our social media technologies which are constantly creating envy for a life beyond reach, are at the root of much of the discord in the U.S. today. Interestingly, much of this income disparity has been driven by productivity gains of technology which have agglomerated to the most wealthy, further exacerbating the problem. Although we may appear to be better off than fifty years ago “on average,” the differences between the haves and have-nots have increased, creating a cauldron of future trouble.
AI unfortunately may exacerbate these differences even more, creating a class of companies and individuals who are the “shepherds” of AI, as global chess champion Garry Kasparov, described it to me versus the rest of us. These shepherds will have ever greater power, control, and wealth magnifying the problem further, and like in the university, unrest that hurts individuals and society will ensue. This is an existential challenge.
The good news is that this dystopia can be avoided. What we learn from universities, and from the history of technology, is that we make choices about how progress benefits a few or many. And ironically, in the long run, those at the top, like the university in general, are actually better off when everyone has the chance to do great work. Just like universities in Europe demonstrate a more equitable, even if imperfect model, for letting students and professors do great work, so too can we make choices about how we apply AI to complement people to do their best work vs turning them into disempowered gig workers. And just as we could design a more fair and equitable university system, so too can we shape our economic systems.
Obviously, some people chafe against the idea of shaping and designing, arguing for the benefits of the free hand of the market. But the “freehand” is a bit of an illusion if we are honest—everything happens in the context of rules and culture, advantages and disadvantages. Adam Smith—the economist who originally wrote about the invisible hand of the free market, wrote his iconic work while he lived at home, coming down to a dinner prepared each night by his mother who did so not for payment, but out of affection. She created the context in which he could do his best work. In a similar manner, why not ask, how do we create the context for everyone to do their best work? The benefits of such an approach are simple math: although there might be some initial investment, the gains to progress, productivity, and prosperity would be immense.
コメント