Why Does Anyone Do Anything?
On finding purpose in the age of AI
When I was a teenager I played a lot of chess. I studied books about chess, I memorized chess openings, I went to chess tournaments where I’d play chess with my friends in between rated chess matches, and eventually got good enough to win some of those tournaments. Over time, I lost interest in the game somewhat and my skill atrophied, though I still play occasional blitz games online.
I put in all those hours even though human chess players were essentially obsolete by the time I learned the game. In 1997, when I was 10 and didn’t know what the Ruy Lopez was, Gary Kasparov lost to Deep Blue, the first time a reigning world champion couldn’t defeat a computer. Around 2005, when I graduated high school, chess computers had become so powerful they could consistently win against any human player. If the point of chess is to make as many optimal moves as possible, humans have become completely superfluous to the game for going on 20 years.
As AI becomes better and more widely adopted, we’re going to see machines dominate more and more intellectual activities that, like chess, used to be the sole purview of humans. AI can write memos and term papers far faster than people can, it can solve puzzles, it can write code, it can create images and videos, it can serve as customer support for businesses, and on and on. As AI permeates society and culture, the challenge will be to assert why AI shouldn’t be used to do something.
In the case of chess, unbeatable machines have not ruined the game, because they haven’t been allowed to take it over — it’s still a game played between two humans. Chess is intellectually stimulating to study and fun to play, but it’s neither of those things if you’re pitting yourself against a god-like robot (or using a god-like robot to cheat). An AI-dominated version of chess would be pointless, like running a race against a car or completing a crossword by googling every clue. When the outcome is predetermined by technology, competition is literally pointless.
This would seem obvious, and yet this lesson is not universally understood. For instance: Some people are using AI chatbots to solve puzzles in escape rooms. That is, they are paying to play a cooperative game that’s supposed to challenge them, then cheating at that game to remove the challenge. “And it’s not just escape rooms,” Michelle Santiago Cortés writes in the Cut. “ChatGPT is being used to enshittify fun in all sorts of places. It’s used at trivia nights… It’s writing flirty messages and ‘Good morning’ texts full of em dashes.” People are treating the ChatGPT app “like a pocket librarian and portable therapist,” Cortés continues, using it to “come up with recipes, write personal messages, and figure out what to stream.”
AI clearly has many legitimate uses. Just to pick one minor example from my own working life, it has erased the hours journalists used to have to spend transcribing interviews. There are a million minor, essentially administrative tasks that AI will inevitably automate. But as Cortés points out, apparently many people are using AI to automate their leisure time, too. You can use AI to (for instance) send messages to people on dating apps, generate letters to Santa, churn out fake stories to farm engagement on Reddit, engage in fantasy football trash talk, or send condolence messages. That last link offers users the option to select different “message styles” ranging from “gentle and grounding” to “light humor or humanity.”
It’s unclear how many people are using these services; hopefully, as we all become familiar with AI, that number will approach zero. We need to remember that the purpose of human interaction is to interact with humans. Even on relatively impersonal platforms like dating apps and Reddit, everyone is there under the assumption that they can reach out and have another person reach back1. You write messages to people because taking the time to write something demonstrates that you care about them. Drain the human element from these activities and they become as pointless as a chess game played between two nigh-omniscent computers. AI can be a labor-saving device, but it can also be a shortcut to nowhere.
A relatively common AI usage is having it write essays for college classes, which is not so different from using it in an escape room. These students are being asked to write essays partially to test their knowledge of whatever subject is being taught, but also to push them to synthesize information from multiple sources and formulate arguments. Importantly, these students are paying (or their parents are) for this experience, because learning how to think in this way is valuable — but then they’re skipping that process, and learning nothing. You can glide through college on the wings of AI, just as generations of clever slackers have glided through with a combination of cheating and buying human-written essays. But when you find yourself graduating without learning anything and saddled with debt, what’s the point?
I’m sure I’m not the only former student who misses college not for the cliched party-adjacent reasons but because you got all that time to read, and think, and write. It’s fun to stretch your brain in an intro to philosophy class, just like it’s fun to study chess openings or learn to draw or debate someone or memorize poetry or a thousand different mental activities that now can be outsourced to AI. Creative and intellectual work is a pleasure because it is a challenge, and when you smooth away the challenge, as AI does, the pleasure vanishes as well.
More simply, thinking is a joy. It is the one true unalienable right. Chop off my limbs, throw me in prison, burn my books, seize my computer and phone (seriously, please, take my phone), put me on the beach that makes you old from The Beach, lock me in a room with my screaming, unreasonable four-year-old who wants to drive in the car without her seatbelt on — do all that and I’ll still be free to think. Bind me within the confines of a nutshell and I’ll be a king of infinite space. Your thoughts are the only thing you’ll always have; thinking is the reward you get simply for being human. Every previous generation of labor-saving technology has allowed humans to spend less time on chores like churning butter, washing dishes, and sewing clothes so that they could spend more time thinking. Even today, when your drugerous workday is done you probably relax by doing something to stimulate thought. You read books, watch TV, you listen to podcasts, all because these things provide your mind with fuel. And yet there are people apparently willing to outsource their thinking to a machine at the first available opportunity.
No one can take your ability to think away, but you can destroy it yourself. Can you regain it? We’d better hope so.
If you enjoyed this essay, please leave me a comment and tell someone about it in real life.
Obviously catfishing and fake posts predates AI, but these have always been regarded as antisocial and ban-worthy behaviors.
Thanks for this, Harry. Really good points. Reminds me of a thought experiment I like that asks, would people work if they didn’t need to? Like, if they didn’t need the money.
Since retiring I seeing many people working — usually volunteer work — who are working for others, their community, charity, social change and so on because it’s important to them.
I can imagine a future scenario in which AI enables most people to not need to work. I would hope that those folks would find meaningful “work” that results in explosions in the arts, in government, education, science, media and community support. Maybe current utopian fantasy.