Andrew Potter:
But the hard work of thinking is literally what makes us human. 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 defines what you care about, and ultimately, who you are. Which is why whatever else AI does, whatever benefits it brings the world, there is a sense in which it literally steals your soul.
To me, this is an updated version of the Nicholas Carr-type argument that the internet/social media/AI is “rewiring our brains.” From my perspective, if your attention span and “love of reading” didn’t survive the encounter with Angry Birds, it wasn’t very strong to begin with. If your love of “writing and thinking” doesn’t survive the encounter with AI, perhaps you have some painful introspection to do. If anything, I love writing and thinking more now because I have an interlocutor who never gets tired or bored and helps me to build out thoughts that would have remained unfinished otherwise. AI feels to me like I’ve been handed a Zippo lighter. Formerly, I had to bang rocks together, or rub sticks together, hoping for a spark of insight to catch. Now, the fire is blazing much quicker, but the heat is the same as the “organic” version. The “artificial” heat still allows me to do all the things one normally does when one’s teeth aren’t chattering and one’s fingers aren’t numb. AI has not “replaced” the fire. It has replaced some of the fumbling around in the cold, the freezing in the dark while proving your authenticity to the indifferent woods.
In some sense, every moment is a stage for the same eternally-recurring choice: do I make the effort to do this difficult thing, or do I make an excuse to not do it? It could be reading abstruse books, it could be lifting weights, it could be anything that forces the challenge upon us: this will take time and energy; do I want to do it? This choice is never resolved for once and for all. There is no “progress” here. If I give in to laziness and stop lifting weights, I don’t get credit for all the time I put in; I only succumb to entropy. Likewise, if I let an LLM do my writing and thinking for me, my mind turns feeble. If I recognize it happening and fail to change my habits, well, then I fail, period.
I hear complaints like Potter’s and I just shrug. This is the human condition; what do you want me to tell you? You want me to remove all temptation for you? You can’t handle the truth of the choices you make, so you want all the nearby mirrors covered so you don’t have to look yourself in the eye? Yes, it’s always pleasant to have a burden lifted off of you, at least for a moment. But you simply must philosophically commit yourself to some kind of burden. It’s up to you to choose which one.
Technology is very good at neatly separating your aspirations from your actual priorities, however embarrassing that may be for our self-image. The person who “loved reading” until the internet came along may have loved the identity of being a reader more than the activity. The person who “loved writing” until AI could do it faster may have loved the status of having written more than the labor of trying to say something. If the self disappears the moment an excuse becomes available, perhaps the self was already negotiating its surrender.
Premodern life made many choices for you. You rose when there was work to do because hunger, weather, animals, crops, family, illness, and scarcity were not interested in your self-actualization. Burdens were not “meaningful” because people had nobly chosen them; they were meaningful only in the blunt sense that failing to bear them had consequences. Civilization softens that coercion. It gives us heating, refrigeration, medicine, literacy, machines, contraception, abundance, leisure, mobility, digital tools. Good. Thank the gods for it. But then the question becomes more exposed: now that you don’t have to live under necessity, what necessity will you impose on yourself?
Many people start looking wistfully back toward constraint. Not because they truly want cholera, crop failure, childbirth mortality, untreated tooth abscesses, and hauling water. They want the moral simplification of necessity without the misery that came with it. AI intensifies this because it reaches into the last prestige zone of voluntary effort: thought. We already made physical survival easier. We made social communication easier. We made entertainment effortless. Now even the internal friction of composing, reasoning, summarizing, and articulating can be smoothed. That doesn’t mean the soul has been stolen. It means the soul is being asked, rather rudely: “Were you serious?” The old world forced people to endure hardship. The modern world forces people to decide whether they believe in anything strongly enough to endure hardship voluntarily.
There’s a respectable version of “I need external limits,” and then there’s the evasive version, where a person has enough clarity to describe the trap in exquisite detail but somehow not enough clarity to move the cheese. “I notice that this thing makes me miserable, consumes my attention, weakens my patience, and leaves me ashamed afterward.” Good. But the next step does not require a graduate seminar in attention ethics. Delete the app. Put the phone in another room. Use the browser only on desktop. Set a timer. Make a rule. Break the rule, notice it, revise the rule. Go outside. Read ten pages before opening the laptop. Write badly for twenty minutes before consulting the LLM oracle. None of this is mysterious. It is boring, which is precisely why people would rather inflate it into metaphysics.
This is partly why I talk about the gym so much. Lifting weights is applied philosophy for me. It is a controlled environment where you can practice, in a small way, making the eternally-recurring choice to work harder when you’re tempted to rationalize excuses for weakness and laziness. The barbell sits there, mute, impassively ignoring your pleading and bargaining. You either lift it or you don’t. No serious gym-goer would earnestly claim, the way clever bookish types do, “The existence of couches has stolen my quadriceps.” Couches are always there. Laziness is always there. The body is always trying to economize. The question is whether you willingly resist. If my triceps are weak and making it hard to bench press, what’s the answer? Work on the triceps in particular with supplementary exercises. Not, “society should redesign the gym to suit the needs of people suffering from Tricep-Deficiency Disorder, and we need an awareness-raising campaign…”
The same goes for the mind. The blank page, the abstruse book, the hard sentence, the boring revision, the honest objection — these are forms of resistance. AI becomes dangerous when it is used to remove resistance entirely. But it can also be used like a spotter, a sparring partner, a stack of plates, a second mind pressing back.
I don’t deny the structural question. Someone has to think about defaults, incentives, laws, schools, platforms, and public health. But that is not my interest. I am not a would-be social planner, nudger, or shepherd of populations. I say, drolly, that I’m more of a personal trainer for the soul. I am interested in the person who still has enough agency to be addressed as a responsible individual. I do not want to arrange his cage more humanely. I want him to notice the open door. I want him to become self-sufficient, to take an active hand in shaping his own life, to acquire a taste for chosen difficulty rather than cheap release. The question that interests me is not how society can be engineered around weakness, but how weakness, once recognized, can be trained. The social planner asks: how do we organize society so that weak people do less damage to themselves? The trainer asks: how do we make weakness ashamed of staying weak?
The structural frame becomes dangerous when it gives people a respectable language for waiting. “The incentives are bad.” True. “The platforms are engineered for addiction.” True. “The social environment is hostile to attention.” True. But then what? Are you released from action until the Ministry of Humane Defaults issues its findings? “What do you want me to do? They made this stuff addictive. Until society fixes the problem, I have no choice but to continue doing the thing I have just identified as ruining me.” This person knows enough to diagnose the machinery, but somehow not enough to put the phone in the drawer. He can explain dopamine loops, variable rewards, dark patterns, algorithmic capture, and late-capitalist attention extraction, but cannot survive ten minutes without refreshing the slot machine.
We are never free in some abstract, frictionless sense. We are born into bodies, temperaments, habits, economies, families, technologies, desires, and eras we did not choose. But inside those constraints, there is almost always some movable piece. A small refusal. A rule. A delay. A substitution. A chosen inconvenience. A door left unopened. A tool used only after the first draft. Ten pages before the screen. The phone outside the bedroom. Dumb little measures, maybe, but dumb little measures are what self-command is made of. The structural analysis may be true and still be useless to the person using it as a sedative. Yes, the platform is engineered to keep you there. Yes, the incentives are bad. Yes, your appetites have been studied, modeled, and monetized. But unless you are literally chained to the machine, the next question remains yours: what are you going to do with the portion of freedom left to you?