Students are Right to Jeer the Promised AI Future

https://youtu.be/5MYggR_PPRg?si=KLC-6XD8svqYzdEf

Generative AI instantly produces lots of things: computer code, recipes for cookies, images of ninja dogs, and Muppet-versions of Pulp Fiction. But one underappreciated thing it is generating right now is resentment. Multiple commencement speakers touting the technology, notably ex-Google CEO Eric Schmidt, have found hostile audiences to their pro-AI message.

Schmidt’s speech suggests the anger stems from fears of job loss, and thus recommends students not mourn old careers but lean into AI ones. But this underestimates the harm, failing to recognize that the threat of AI extends beyond our jobs; the deeper threat is to our agency and, with it, our capacity to determine how we want to live. And students recognize this most acutely because their schools are aggressively flattening and homogenizing their education in the name of AI, depriving them of the challenges they signed up for. Their jeers are a call to arms against the grim future AI promises.

We should be clear: many technologies in our history have been transformative, and job destroying, in the way tech boosters claim AI will be: the printing press ended the scriptorium, the car shuttered many horse stables, and the computing machine replaced number-crunchers. But they were also empowering and connecting, opening new avenues for realizing ourselves at work and play. Mass media, from books to radio to TV, has made art common and accessible rather than rare, informing people of different worldviews and novel ways of living from across time and space, allowing them to share news and organize politically, and making an educated citizenry not just possible but fundamental. The freedom afforded by a car permitted people to travel and move around the country, and it quickly became a center piece of our national identity, captured in songs like “Life is a Highway,” books like On the Road, and still grounding movie franchises like Fast and Furious. And for me, multiplayer video games and internet chat boards gave me space to find myself through interactions with people distant, different, and often disagreeable people who bond over a shared passion for fighting robots.

There is no similar narrative at work in current AI. What they promote is dystopian, a disempowering and isolating existence that deprives us of the space for self-creation. Current AI’s default involves depriving agency, not just by taking our jobs, but more mundanely by summarizing our emails and drafting responses before we’ve even glanced at them. Google’s search function increasingly limits our access to real websites in favor of AI-generated “answers” cribbed from other sources, burying the niche blogs with Reddit threads filled with debate about the best coffee grinder or worst stout. PowerPoint and Adobe often generate clip art or images based on our work without asking, aiming to pre-empt a search for the perfect Simpsons gif with overly literal slop. While it is possible to bypass and ignore these functions, companies are betting on resignation—accepting the slop because it is available, regardless of whether it is what we want.

Current AI is also deeply isolating, often pushing a solipsistic experience where we increasingly encounter only distorted versions of ourselves. The stories of deluded, self-destructive users hardly surprise us anymore, since apps promise us “friends” that are tailored to our own desires and to meet our whims. Social media sites are attempting to mass produce these sycophants, populating our timeline with artificial friends liking our posts and photos. Image programs promising boundless creativity are instead becoming a samey stream of hyperrealist slop, where every woman is 14 with bedroom eyes and every man fit enough for a Marvel movie. For AI, the world—with its challenging relationships, social ambiguity, and mundane beauty—is unpleasant, complicated, and ugly. The replacement is a smooth, effortless existence of affirmation and sexiness.

It is the youth who grasp the danger because schools are leaning heavily into AI. Classes and departments in the humanities and social sciences are cut or dumbed down as students can simply summarize texts. Writing assignments are increasingly rare because of the threat of AI cheating, with the result that everyone loses out on critical feedback and advice. Research papers are often treated as pro forma, as students are encouraged by university administrators to have AI populate an annotated bibliography rather than read articles themselves. Fundamentally, students recognize they are receiving an inferior education than the one offered only a decade ago, and that their university boards and presidents are cheering this on rather than lamenting it.

The gutting of their education stems from an underappreciation of what makes agency possible. Everyone can play chess badly or stumble through a book without understanding it. But doing either well takes practice. And it is only through extensive grinding that we can achieve the competency required for a skill the be empowering—before we can feel confident playing a game of chess or grabbing the new Pulitzer novel, rather than encounter it with fear at our own failure.

Schools are designed to accomplish this skill-building, where mentors can guide students in gaining the basic skills required in reading, writing, and coding that they can hone over a lifetime. Universities extend this, teaching students the fundamental skill of self-improvement—not just recognizing how to be good at something novel, but also an appreciation of the grind, accepting feedback, and pushing ourselves through the embarrassment of mediocrity. And this demands students suffer through difficult, challenging texts, like Plato’s Phaedrus or Morrison’s Beloved. There is no other path for gaining the skills than doing it oneself; as Confucius warned, the teacher can only lift one corner of the rug, but it is up to the student to lift the other three.

AI-centered university’s promise is the end-product without the process. They contend the future of employment doesn’t require careful reading or skills at writing. But, as Ted Chiang notes, this is encouraging students to bring a forklift to the gym to do the lifting; it fundamentally misses the point. It is a profound waste of universities for students to be surrounded by people who’ve devoted their lives to building skills around their passions—published authors, distinguished speakers, artists and actors—and yet spend their time asking questions of their computer.

There may be a kind of person who wants AI to write books in their name or tell them the plot of Vineland. But they are not the students I see who are teaching themselves Japanese and practicing kenjutsu to better appreciate anime. They certainly aren’t my students staying late to debate criminal justice or ethical relativism or asking for comments on an article they’re writing for the school newspaper. Students want more out of school, and are resentful schools keep offering less in the name of AI.

We should share their resentment at the promised AI future. Asimov’s robots and Star Trek’s Data opened our eyes to how AI might challenge our humanity and force us to appreciate what makes our own lives unique and worth living. But current AI is in the hands of people without a vision of humanity’s future beyond passive consumers. Their vision is incurious and boring—and it behooves us to ensure it does not come to pass.