Andrew Potter: You're going to miss your pre-AI self

By turning normal mental effort into something that feels like a defect, AI is a challenge to the very concept of who you are.

Humans at work.

By: Andrew Potter

It’s hard to screw up a university commencement address. Take a captive audience of optimistic young people and their proud parents, have a person of accomplishment feed them some platitudes of how the world is their oyster and they should follow their dreams and go change the planet, and everyone goes home content in the belief that it was four years and a few hundred grand well spent.

But in the graduation season just ended, at least three commencement speakers in the United States had their words of wisdom greeted with a chorus of boos and catcalls from the assembled graduates.

At the University of Central Florida, the audience erupted in booing at a ceremony in early May after real estate development executive Gloria Caulfield called AI “the next industrial revolution.” After she implored the audience “can I finish?”, someone yelled out, “AI sucks!”

At Middle Tennessee State University, the boos came when music executive Scott Borchetta told the grads that AI was a tool they needed to learn how to use.

And at the University of Arizona, the jeering went on for minutes after Google CEO Eric Schmidt told the students to get on board with AI. “When someone offers you a seat on the rocket ship, you do not ask which seat,” Schmidt said. “You just get on.” The loud response from one member of the crowd? “Fuck this guy.”

It is possible this was all lingering resentment at the knowledge that somewhere around two-thirds of their classmates had coasted or even cheated their way through the last couple of years of their degree. Or perhaps it was more forward-looking: A bleak job market becoming more unappealing by the minute, thanks in part to the very technology being pushed on them by the speakers who had been invited to send them on their way.

But is it possible the hostility might have been coming from a place even more profound? Could the students have been, if only subconsciously, mourning the imminent demise of their pre-AI selves?

For all of human existence until about a year or so ago, the most significant feature of intellectual labour, writing in particular, has been how hard it is. Thanks to the success of the behavioural economist Daniel Kahneman’s 2011 book Thinking Fast and Slow, we have a good grasp of why that is the case. The distinction between “System 1” and “System 2” to describe two distinct ways the human brain processes information and makes decisions has become part of our received wisdom. It has been overused in places, and many of Kahneman’s claims have ended up hard to replicate, but the basic scheme is still a useful way of understanding how our minds work.

System 1 is what is often referred to as our gut instinct or intuition. Think of driving a car on a flat open highway on a clear day, knowing that two plus two equals four, or catching a ball thrown at you from a few feet away. It is fast, almost effortless, and in a lot of ways completely automatic and involuntary; we often don’t even really know how we know or do these things.

On the other hand, whenever you try to memorize a new phone number or email address, do long division or multiply two three-digit numbers, play a musical instrument or learn a dance routine for the first time, you are using System 2. It is slow, voluntary, and explicit, forcing you to hold a number of thoughts and ideas in working memory as you work through the steps. It also requires an enormous amount of effort.

These are not actual modules or parts of the brain, they are modes of thinking, or ways of making judgments, and each involves a key tradeoff. What you gain in speed with System 1 you lose in accuracy, as the cognitive shortcuts it relies on often lead to errors in judgment. For System 2 tasks, the high cost of mental effort is justified by the need to get things right and to make things explicit. When someone insists you “show your work,” they are asking for the evidence of System 2 labour.

The act of writing is pure System 2 work. It is sequential, explicit, and transparent. You have to work through an argument, or a narrative, line by line, clause by clause. Writing a simple newspaper column can take an entire day, a proper book can take years of research, planning, and execution.

AI vaporizes that. Take a few seconds to prompt a chatbot and the sentences come pouring out, sophisticated arguments, inferences, and evidence, leads and kickers and turns of phrase. Some of it is great, some of it isn’t. But any writer who has spent any time with a decent LLM has experienced that moment where it says something better, edits it smarter, makes it more creative, in seconds.

There are AI apologists out there arguing that AI is to writing what the washing machine was to laundry, or what store-bought butter was to the butter churn: it takes away the hard drudgery of the task and leaves the fun, creative part to the humans. Treated as a new tool in the cognitive arsenal, AI is just another step on the way to freeing humans from menial labour, both physical and mental.

But the analogy doesn’t hold. The washing machine and the butter churn allowed humans to trade stupid work for something smarter, while AI does the exact opposite. A new study from researchers at Carnegie Mellon and MIT found that when people used AI for even 10 to 15 minutes, they performed significantly worse when the AI was removed, even on tasks they should have been able to answer easily. They argue that repeated AI use appears to prevent you from building the mental model you need to solve a problem yourself. And when instant answers from AI chatbots become routine, using your bare brain for System 2 tasks starts to feel disproportionately hard, “a kind of recalibration that turns normal effort into something that feels like a defect.”

This isn’t just AI-induced laziness. System 1 is your primate brain, the legacy of our evolutionary heritage from the ancient savannah. The hard, focused attention that characterizes System 2 thinking is a currency, and what you choose to spend it on literally defines who you are, what you care about, and ultimately what makes you human.

The rise of AI isn’t just a further stage in the liberation of humans from physical drudgery. It is a direct existential challenge to who you are. The students who booed and catcalled the speakers lecturing them on the exciting new digital tools that were transforming the world under their feet might have been merely irritated by the cheats in their midst, or anxious about the job market out into which they were being sent.

But more than likely, the jeers reflected a deeper unhappiness, stemming from the certain knowledge that they had spent their last days with their pre-AI selves. And they were going to miss them.

---

The Line is entirely reader and advertiser funded. No federal subsidies, no bailouts. If you value our work, please consider supporting us by subscribing or making a donation. Donations are not subscriptions and do not unlock paywalled content, but they help keep The Line independent

---

To contact The Line with a general inquiry or comment, please email info@readtheline.ca. For other ways to connect with us or to follow us on social media, please see our LinkTree.