What If the Goal Is the Means?
On human purpose in a post-AGI world
In his 2020 book Games: Agency as Art, philosopher C. Thi Nguyen poses a provocative thesis: Games, he argues, involve an inversion of the way we normally think about goals.
Take Scrabble or basketball as an example. Deep down—gun to our head—we don’t really care about spelling the best words or putting a ball through a metal hoop. Instead, we temporarily and provisionally take up these objectives because they facilitate a pleasurable sense of flow or striving. We think the purpose of playing basketball is to make baskets when, really, the purpose of making baskets is that it lets us play basketball.
In other words: We don’t play games because winning is intrinsically valuable; rather, the goal of winning is a means to a certain mental state.1
I believe this logic extends beyond what we traditionally think of as games. If we were honest with ourselves, we’d probably have to admit that the world wouldn’t end if we took a week off work, or stopped posting on Instagram, or gave up reading the news. Viewed in the right light, these are all just games called “career” or “popularity” or “being informed”; while they may have some more intrinsic value than Scrabble, we surely also play them because they give our lives meaning. Again, the goals here are (at least in part) a means to a certain internal states.
But what happens if and when AI makes our efforts truly and unequivocally superfluous—when robots can perform every conceivable human activity at a superhuman level? As programmer James Somers describes:
A few months ago, I came home from the office and told my wife about what a great day I’d had wrestling a particularly fun problem. I was working on a program that generated a table, and someone had wanted to add a header that spanned more than one column—something that the custom layout engine we’d written didn’t support…I knew it was a good day because I had to pull out pen and pad—I was drawing out possible scenarios, checking and double-checking my logic.
But taking a bird’s-eye view of what happened that day? A table got a new header. It’s hard to imagine anything more mundane. For me, the pleasure was entirely in the process, not the product.
Or to give a closer-to-home example: I enjoy writing. I enjoy not only the creative highs but also the frequently effortful process of smoothing transitions and cutting excess words, of trying to solve the text. But does this enjoyment perhaps depend on the belief that the product will, or at least might, matter in the world? I’d like to think not—that I would continue to write, if only for my own gratification, even if no one would ever read the words. But who knows! Perhaps by the time GPT-10 comes out, human-generated writing will feel like a self-indulgent anachronism.
And in all kinds of diverse fields—art, science, fixing cars—people enjoy otherwise tedious forms of problem-solving because said problems’ importance makes the search itself pleasurable, and because only they can find the answer. Yet day by day, this basic fact of existence becomes less true. Even if we’re aware of the psychic rewards of doing things the hard way, clicking the easy button may prove just too tempting.2
People will make a game out of anything. This is why, most of the time, I don’t worry too much about AI making our passions obsolete; we’ll still enjoy playing something. Indeed, Deep Blue did not spell the end of chess, the forklift did not spell the end of squats and bicep curls, and ChatGPT seems not to have spelled the end of human coding and writing. So if I had to bet, I’d say that even in a techno-utopia of infinite leisure, we’ll still imbue arbitrary human activities with status and meaning.
But on my pessimistic days, I worry about a drift toward nihilism—that when we become the pampered pets or finger-snapping masters of a superintelligence (these being the best-case scenarios), the myth of human purpose will be unsustainable; that all human endeavors will feel like a silly game. In the words of poet Marge Piercy: “The pitcher cries for water to carry / and a person for work that is real.”
If this account seems too clever, it can be illuminating to consider the ways that gameplay often fails: either because competitors take the end goal too seriously, or because they do not take it seriously enough. The Scrabble player who invents nonsense words and fake definitions is annoying because they haven’t cleared the minimal threshold of buy-in to make the activity fun. Conversely, the pickup player throwing elbows and arguing every call is annoying because they’ve forgotten that basketball is “just a game”; that the rules and points are ultimately arbitrary. There’s a sweet spot of play—with some variation between individuals and contexts—that is engaging but not fanatical.
In some cases, it may even become irresponsible to do things the hard way. For example, if we get to the point where driverless cars are many orders of magnitude safer than their human counterparts, I’d argue that it doesn’t matter how much joy someone feels behind the wheel; the risks of “playing driver” are no longer justified.