How Games Are Shaping Our World

This interview is part of The Rules We Live By, a series devoted to asking what it means to be a human living by an ever-evolving set of rules. The series is made up of conversations with those who dictate, think deeply about, and seek to bend or break the rules we live by.

¤

“GAMES ARE A KIND of untrue truth,” writes neuroscientist and physicist Kelly Clancy in her new book Playing with Reality: How Games Have Shaped Our World. In a game, we suspend reality, allowing ourselves to inhabit new roles and identities and explore the realm of “what could be.” For this reason, games have a storied history as simulations and tools for testing our intelligence and creativity.

Games also helped usher in probability theory and game theory, transforming chance from “the mouthpiece of God” into a series of laws that are susceptible to analysis and which enable prediction. But Clancy anticipates a complicated future as we more intentionally use games to model and shape reality. “I think games are more likely to temper a dystopia than usher in utopia,” she tells me. Relying on powerful historical examples, Clancy considers the attributes that make games enjoyable and rewarding, the empty promises of gamification, and the reasons games are so fundamental to understanding who we are.

¤

JULIEN CROCKETT: You dedicate Playing with Reality to the next generation, writing, “If play is the engine of creation, I cannot wait to see what new worlds you build.” A recurring theme in your book is the primacy of play and games to what it means to be human—that playing games is universal, an instinct. What makes games and play so important to understanding ourselves?

KELLY CLANCY: Play is something that predates humans. It’s fundamental to how animals engage in and understand the world. It’s how we test and figure out the rules of our environment, how social relationships work. Part of the idea to focus on play in the book was generated from David Graeber and David Wengrow’s book The Dawn of Everything: A New History of Humanity (2021), where they go through different ancient societies and show not only how incredibly diverse they were—contrary to the idea that there is a monolithic trajectory from hunter-gatherer to agriculturalist to industrial society—but also how incredibly playful all of these societies were. More recently, at events like Carnival in Renaissance Europe, people would put on costumes and participate in taboo things like gambling, having affairs, and drinking. There were all of these opportunities to release the strictures of society and try something new.

This is something that I think we’re missing in our current society but that maybe the internet is giving us—an opportunity to put on a new mask, a new personality, and to act in a way that we wouldn’t necessarily in our everyday interactions. I think of play as a way we try on different identities and work our way toward a society that, hopefully, eventually works for everyone.

Looked at from another perspective, play also helps reveal that there is no baseline, fixed human nature. We are extremely malleable, changing depending on the game we’re playing, the roles we inhabit in society. You write that one of the purposes of the book is refuting B. F. Skinner’s claim that “no theory changes what it is a theory about; man remains what he has always been.” What about Skinner’s perspective is so incorrect, in your view?

There is a lot to this. I didn’t anticipate writing a book addressing B. F. Skinner, but his ideas kept coming up, particularly in psychology research and the type of behavioral-economics research relying on game theory to understand human nature. Game theory is a description of an optimizing function. The people who created it initially thought it could be a good model of human behavior because humans are selfish. Game theory is a beautiful field of mathematics, which has been really useful for operations research, for example, but which has been squarely refuted by psychologists. Humans are not selfish. We change our behaviors depending on the context we’re in.

For example, you can have the exact same game, but by simply changing its name—say, from “Community” to “Wall Street”—people behave remarkably differently. These findings led me to believe that we’re very sensitive to our environment, to the cues we’re picking up, to social contexts, to values, and we’re not necessarily optimizing any one thing the way that a game-theoretic agent would be. That’s where Skinner’s idea of people being immutable, monolithic, unchangeable entities falls apart.

You present the history of game theory through the history of games themselves. Starting pre-Enlightenment, you explain that our fascination with games came from the belief that God operates through uncertainty. Even though we had a long history of gambling and mathematics, as statistician Florence Nightingale David argued, one reason people did not develop probability earlier was because their “habit of mind […] made impossible the construction of theoretical hypotheses from empirical data.” What about our early belief systems delayed discovering probability?

David believed there were a number of reasons we didn’t strike on the mathematics of chance until fairly recently in history despite people playing dice for thousands of years. Part of it is just mechanical—we were using Roman numerals, which were very clunky for representing certain kinds of numbers. But also, chance just seemed so remote from logic. It couldn’t possibly be understandable. In the same way that it took us a long time to realize that the same gravity that works on an apple on Earth is the same gravity that works on celestial bodies in the sky. That was Newton’s big insight, which in many ways came thanks to astronomers looking at data and proposing and rejecting hypotheses. The ancients just did not have much of an empirical tradition.

But even though we have probability theory today, we still retain superstitions around chance—like blowing on dice to swing chance in our favor.

We do retain a lot of primitive intuitions. I think of it as our brain’s way of trying to retain a certain amount of control over reality.

What role did games play in the discovery of probability theory?

The birth of probability theory came from Pierre de Fermat and Blaise Pascal. Pascal was a sickly young child and math prodigy, but when his father converted their family to Jansenism, in which math was considered akin to masturbation, Pascal was prohibited from doing math. After his dad died, however, he went into a period of degeneracy, gambling, and doing math. A friend posed some questions about, well, if I bet on this roll of dice, I usually get a decent payout, but if I bet on this other one, which seems to be the same, I don’t. What’s up with that? Pascal wrote to Fermat and their exchanges gave birth to the new field of probability.

It blew everyone’s mind. Chance was supposed to be unknowable, the mouthpiece of God. It was crazy that we could find laws in chance, and that suddenly we had turned something intrinsically chaotic into something understandable. It was a huge win for the scientific method. If we could systematize chance, then what else could we do? And the fact that games—gambling!—led to this really powerful insight caused Enlightenment thinkers to take games more seriously. Gottfried Wilhelm Leibniz, for example, became obsessed with games and started a games academy in Berlin where they used games as a model for thought, innovation, and invention. It was also where the idea to use games to model warfare came from.

Which the Germans perfected with Kriegsspiel.

Right. Eventually, a couple hundred years later, German military officers started taking this idea seriously—first with chess, which itself is an abstraction of war. It was meant to model four divisions of the ancient Indian army, and Germans reconcretized it as Kriegsspiel, adding more realism, expanding the board to around 1,600 squares, and eventually doing away with the board and playing on scaled maps of real territories. They would simulate actual battles—both battles they were undertaking and historical battles to try to understand what went right and what went wrong. It became a scientific endeavor to prepare a military offense.

Kriegsspiel was directly responsible for the creation of the German Empire in the 1870s and then for much of Germany’s successes in World War I and World War II. It also led to their downfall in World War II because the Russians were also using Kriegsspiel to prepare their counteroffensive.

How did Germany’s use of Kriegsspiel help give birth to game theory?

It comes down to John von Neumann. Von Neumann was a Jewish mathematician born in Hungary, who was also a child prodigy. He was the most brilliant man of the 20th century. Luckily, he was able to flee Europe in the 1930s, but he was completely devastated by the war. He was proud of his European roots but horrified by the racism and brutality of the Nazis. Many of his friends and family were killed or died by suicide. In an effort to make sense of the war, he set about doing the impossible with mathematics: axiomatizing or trying to mathematically model human nature. He thought that the rise of the Nazis came down to economic reasons. His friends tried to convince him that it was really due to animalistic instincts, but he was sure that we could empirically understand their behavior. So, he used games as his model for human nature and started with the simplest game you can imagine—a zero-sum game where if one person wins, the other person loses. A game like chess or checkers.

In collaboration with economist Oskar Morgenstern, von Neumann published a book in 1944 explaining game theory. It was at first completely ignored by economists because the math was alien and weird and it also didn’t seem like a very good model—markets are not supposed to be zero-sum, they’re supposed to be positive. But as other mathematicians generalized the field and made it more applicable to non-zero-sum games and to games with more than two participants, it eventually took over economics, and it is now the heart of the field.

Game theory has also successfully spread across fields. What has made it so successful?

Game theory is clean; it’s understandable. And there are many reasonable applications, particularly for modeling supply chains, for example, or computer networks. These are things that are supposed to be maximized. But it’s a poor model for human beings. Something that blows my mind is that, although game theory gives us a poor model of humans, there is a huge industry of psychology research that publishes these discrepancies between the model and how humans behave as something surprising. So, for example, the research concluding that humans are biased or not rational derives from the assumption that humans are game-theoretic rational. There isn’t really anything interesting here because the base assumption is wrong. And if humans are constantly thwarting this theory, then let’s find a better theory for human behavior. But of course, it’s actually very productive for academics to have a theory that’s constantly throwing up surprising results, because that is something they can publish.

Going back to von Neumann—at the same time as he was trying to understand and discern the physics of human nature, there was also a pull to move military decision-making away from humans and give it to a synthetic intelligence. Where did this movement come from and how did it ultimately lead to AI?

This is clearly super salient today. The draw to having nonhuman decision-making was inspired by the birth of the first nuclear weapons, where suddenly military officers had the fate of the world in their hands. War used to unfold on the timescale of months as troops would slowly move across continents and overseas. Now war can unfold in minutes, when you have missiles and airplanes and nuclear bombs. So we had a need for a new kind of diplomacy, and people like von Neumann were promising that game theory could provide a mental prosthesis to aid our decision-making, or even replace it. This was the premise of the movie Dr. Strangelove (1964), where the idea was that we would enforce peace between two nuclear powers like the United States and USSR with mutually assured destruction. If one party hit the other, the offended party would retaliate with so many bombs that it would basically end the world.

This is not a rational strategy. If we’re the US and we get attacked by Russia, is it rational or moral to end the world? What person could actually press the button and do that? So, the game theorists concluded, don’t let a person do it. Have a machine do it. In Dr. Strangelove, the USSR has developed a machine that will automatically end the world if the US attacks them, but they haven’t announced it yet. A rogue US general attacks the USSR, and it results in the end of the world. Of course, at the time the film was made, we were nowhere near having automated decision-making systems, and I don’t think we’re there today, but that was the goal: to obviate our moral responsibility in some way through an automated decision-making system.

Moving to the present, French sociologist and philosopher Jean Baudrillard argued in the 1980s that “we increasingly live in our maps of reality rather than in reality itself. The territory has been replaced with towering layers of abstraction.” At the same time, you write that we’re “transitioning from being players to architects of the games we move in.” How did this self-conscious transition come about?

If we think about the way humans saw the world before probability theory, our lives were orchestrated by fate. Chance was a reflection of God’s whims. With the birth of probability theory, we gained a little more agency and realized that if we have the right data, we can make better decisions. We can, to some extent, see into the future and steer things in a better direction. That might be a bit of a fiction we tell ourselves, as you discussed with Robert Sapolsky, but regardless, it’s a kind of a fiction that gives us agency and the ability to obsess over our decisions because we feel that we could have done better.

With the creation of game theory, mechanism design, and just the more conscious designing of games, we are realizing that we do not just play games but can also change games. This brings us full circle to what we talked about earlier with prehistorical societies, where they were trying out new forms of society. I think we’re going back in that direction with online governance, where people are trying to find new ways of being together.

The concept of “gamification” has also taken hold over maybe the last 10 years. I understand the benefit of making difficult or annoying tasks more enjoyable, but there does seem to be something nocuous about manipulating ourselves into not noticing the friction of those tasks.

I remember when everyone started talking about gamification and how it was going to change the world and solve problems. Everyone would be able to exercise and have fun doing it. Boring trainings at work would become fun. I have experienced basically none of that. I think part of the problem is that the people trying to gamify things are not using the fun parts of games. They’re using the grimy parts, the parts that make them addictive. For most people, I think getting points or badges is not actually that fun because we see through it. It’s like Monopoly money.

I think the more successful gamifying apps put you on a leaderboard, pit you against your family members, give you competition, something more engaging and social. But when it’s just getting points, it’s cheap and lazy. A lot of education could be gamified. Indeed, games are inherently pedagogical. But the vast majority of what we’re exposed to as average consumers is just really bad and not enjoyable.

In Playing with Reality, you home in on the tension between games or simulations being helpful for modeling outcomes and the fact that games and simulations don’t necessarily reflect something fundamental about the world. How do we bridge that tension?

We talked about Kriegsspiel, and to this day some form of Kriegsspiel is used for military affairs. These kinds of games can be really useful for thinking through what a particular scenario looks like. But of course, you take what you learn with a grain of salt. When the United States was preparing for World War II, they had gamed out different scenarios and had predicted Pearl Harbor. What they hadn’t prepared for, however, was kamikaze pilots. The US couldn’t conceive of people killing themselves for their country in this way. A game is limited by the imagination of the people playing it.

One of my favorite examples of war games being useful was an exercise in 1983 led by political scientist Thomas Schelling. It was called “Proud Prophet.” Schelling led 200 of the top US politicians and military officials through a scenario where Russia attacked and we had a nuclear war. They spent weeks workshopping, and the politicians realized that the rhetoric they were using regarding tensions with the USSR were totally off base. We didn’t have the right technologies to offset an attack, and all of the scenarios, save for one, ended in the end of the world. The only “good” one ended with half a billion people dying outright and many hundreds of millions more dying of radiation poisoning. This scared the politicians enough that the Reagan administration shifted rhetoric and opened sweeping arms control negotiations with the Russians. There is something important about the experiential aspect of a game. It’s not enough to just talk through things.

I want to end with how games will be used in the future. You write that “[m]odern thinkers have returned to Plato’s hope that games might teach players how to improve themselves and become better citizens. […] Games will not just be the official pastime of Utopia; they’ll also help usher it in.” What utopia are we creating by gamifying society?

This was a sarcastic take on gamification and the idea that work and even mopping floors would become enjoyable through gamification. Of course, we can’t make something that is inherently unfun into something fun just by slapping game dynamics onto it.

Games are opiates. They can be really engrossing. And we can even get a feeling of accomplishment from them. But there is only so much from games that we can bring into the real world. My biggest hope is that they will help revitalize education and be used to build empathy within people. But I’m still wary of the tendency people have to get buried and lost in games. I think games are more likely to temper a dystopia than usher in utopia.

¤

Kelly Clancy is a neuroscientist and physicist who has held research positions at MIT, Berkeley, University College London, and the AI company DeepMind. Her research focuses on uncovering the principles of intelligence, and she has invented novel brain-computer interfaces to investigate the biological underpinnings of agency. Her essays on neuroscience and AI have appeared in Wired, Harper’s, and The New Yorker. She spent her childhood stuck on the first level of the video game Myst, and being repeatedly murdered by her younger sister in GoldenEye 007.