Stepan ParunashviliTwitterBooks

Agent vs System

Does it feel like society is heading off a cliff?

Social media companies create platforms so addicting that children can’t help but stick to their phones. Politicians make increasingly vicious personal attacks, to the point where elections look like dog-fights. Newspapers pump out so much biased click-bait that people have vilified opposing views. Companies produce such horrible food that obesity is now one of the most deadly threats. Universities teach so poorly that all the youth end up with are vague ideas and a mountain of debt.

This makes the future bleak, but it gets worse when you try to find a cause.

It feels like there are simple solutions right in front of us: Facebook could just change their algorithm to be less addicting. Politicians could just uphold a modicum of respect. Newspapers could just increase their journalistic standard. Food companies could just sell healthier food.

And when we consider those solutions, we can’t help but feel indignation: It’s like our leaders have forsaken us. They can just make those choices, but they don’t. They imperil our society, even the livelihood of their own grandchildren, for what seems like avarice and short-sightedness.

With that view, one solution sprouts up: what if we forced our leaders to do the right thing?

We could replace Mark Zuckerberg with someone who would make sure the algorithms were less addicting. We could make the New York Times and Fox News report a balanced viewpoint. We could tell food companies to remove sugar from their food. We could ask Literature professors to teach something more useful — perhaps prose.

Simple right?

Well, it turns out we’ve tried this before. Many, many times, actually. At innumerable points in our history, we’ve tried to solve moneylending, conspicuous consumption, alcohol, prostitution, various child-rearing practices, emigration, immigration, the list goes on. Most of the time, we employed the same methods and reached the same result: utter failure.


Let’s look at Facebook for clues. What would happen if Mark Zuckerberg made Facebook’s algorithm less addictive? If our conviction is that Facebook is the reason people cling to their phones, we would expect a sustained decrease in social media usage. But would that really happen?

If Facebook’s algorithm became less addictive, the only change we’d notice is a decrease in Facebook’s market share, not social media usage. Twitter, Tik Tok, Snapchat, and many a startup would gladly eat Facebook’s lunch.

Here we enter our first logical fallacy. We’ve ascribed too much power to Facebook. People don’t exist, so Facebook can show them ads. Facebook exists because people want to see what their friends are up too, and are okay with seeing ads. The underlying force that decides the demand for Facebook is people and their wants. Facebook doesn’t dictate what people should want.

Abstract “Facebook” and “Social Media” away, and you get to a general principle. The “system” (in this case, “Social Media”) is what drives behavior, not the “agent” (in this case, “Facebook”). The agent is merely a player, powerful to the extent that they can satisfy the system.

You can see this by looking at Facebook’s history: had Facebook failed to transition to mobile or to purchase Instagram, it would be a lot less relevant today. Social media, however, would very much remain relevant.

The other ills of society fit this abstraction. Presidential Candidates lower their standards because the people who select them respond positively to dog fights. The New York Times writes biased articles because people prefer to read them. Harvard teaches English Literature, because people still study it.

So we come to an uncomfortable truth: what you see on your feed isn’t quite as up to Mark Zuckerberg as we think.

If we really want to solve society’s ills, we need to think one level higher. How does the system work? Why is the current incentives in social media gluing us to phones? Why is it most profitable to share divisive news? Why do dog-fights win elections?

The answers to these questions are the clues we need to change how the system works. But, finding the answers is so hard that we’ve stopped trying. Instead, we look for solutions that come to mind immediately.

What solution comes to mind immediately? Regulation.

Instead of replacing Mark Zuckerberg, what if we regulated the algorithms that social media companies could use? Similarly, what if we made rules about what is “balanced” news? What if we specified what politicians were allowed to say? What if we made laws about college curriculums?

These ideas flower up over and over again across our civilizations.

In the middle ages for example, we tried to solve consumption with sumptuary laws: regulations on the amount of money classes could spend on clothes. If you’re a knight, your wife can spend at most this much for a coat, if you’re a doctor, this much, and so on.

More closer to home, there was prohibition. The majority agreed that alcohol was bad for society. The moral choice seemed simple: let’s just ban it. If alcohol sellers aren’t allowed to sell alcohol, then people would stop drinking, and society would be better off.

Of course, both of these ideas failed, and they look ridiculous to us now.

Why did they fail?

The consumption that sumptuary laws were trying to curb was driven by an underlying human need: the need to distinguish oneself. This is so core to our collective existence that no amount of regulation could curb it. As Montaigne astutely pointed out, the laws even made consumption more attractive: why not get the goods that “only princes” could wear?

The same was true for prohibition. Make alcohol illegal, and watch the bootleggers flourish. The way a river finds the shortest path down a mountain, a seller finds the shortest path to a customer. The need was too strong.

Worse still, these laws share a common blindness: they don’t consider to second order effects. Sumptuary laws failed to react to changes in fashions, and didn’t consider the counterintuitive increase in desire for distinguishing goods. Prohibition didn’t consider the bootleggers.

This is because regulation works in large sweeps and requires concerted effort, while systems are versatile, decentralized, and can change faster than the time it takes you to finish this sentence. We used a slow, one dimension process to change a fast, multi-dimensional system. The flaw is inherent in our method.

So, our simple solutions are no longer so simple. The picture may seem bleaker, but at least we’ve now pruned some incorrect methods. So what paths are promising?

As far as I know, two methods have worked.

The first, is Natural Selection. In the same way that the environment forces a gazelle to be fast, reality forces our society to be effective. Imagine if the United States, for example, descended into socialism. This wouldn’t mean that the whole world would be doomed to it. If the consequences of socialism bear fruit, our society would lose productivity. As long as some society was able to maintain freedom from this kind of coercion, given enough time that society would become the world power. What we did to the Soviet Union, this more productive society would do to us.

Natural Selection works, but it is a bleak option to rely on: it takes a while and has no mercy. We’d sacrifice lives, increase suffering, and may have to wait through a new dark age. And the world isn’t so theoretical: while we wait, there are existential threats that could end society completely.

Thankfully, this is not the only option. Humanity has one trump card up its sleeve, and it’s saved us over and over again: Innovation.

Consider electric cars. If we tried to force people to buy electric cars, we’d end up with the same problems as prohibition. As long as electric cars suck compared to gas-guzzlers, people will want gas-guzzlers. Tesla solved that problem. How? they out-innovated gas-guzzlers, and better served the human want for transportation.

For news, we don’t need to force the New York Times to report fairly. Substack may do more to fix reporting quality than any regulation could have. Now, journalists can build an audience and support content that isn’t dependent on clicks. To the joy of reporters and readers alike, they changed the incentives that governed journalism.

What’s the analog of this for Facebook? Could we build a social network that made offline interactions as fun and intuitive as online ones? Could we build platforms that made constructive debate more engaging than Twitter? Could we integrate political transparency in our government, so that presidential candidates would filter to high-integrity? Could we make a University that encouraged free thinking and creation, at a hundredth the cost?

It’s hard, but once we choose to search along viable paths, the outlook is optimistic.

There have been many points in our past that imperiled our society, and innovation was up to the task. It is just as powerful, decentralized, and versatile as the most pernicious of society’s ills.

It’s a wonder that so many people today can live a life that lords could only have dreamed of a few hundred years ago. Slavery is eradicated in most of the world, and feudal societies are the rarest. Our production supports billions of people — an idea that would have seemed ridiculous to Malthus.

It has been done before, and we know this path has viable solutions, while our simple solution is known to fail. If you’re a person who wants to make a dent in the world, now you know two things: what not to do, and a potential path to success. Regulation or moralizing won’t help. Innovation will.

You may balk at the task ahead of you: climate change is a problem so daunting that it’s hard to think about. Luckily, innovation is opaque. You don’t need to attack problems directly.

Penicillin just showed up on Fleming’s desk. Rocket engines evolved from car engines. Ideas tend to compound together, forming ever-more complex ideas. This means you don’t have to go big to achieve big results. You can start by looking at problems around you, and try to solve them.

If everyone just did this and went no further, we’d likely solve most of our ills in society, and enter an age of unprecedented prosperity. But it gets better. Not only will smaller innovations compound with each-other, but each innovation makes the innovator more audacious. The more problems you’ll solve, the more daring you’ll become. I can’t begin to predict where that would lead us.

We can wait it out, or we can innovate. Let’s innovate. We can start right now.

Look around you, what’s a small problem you might be able to help solve?

Thanks to Daniel Woelfel, Cody Breene, Joe Averbukh, Mark Shlick, Sebastian Marshall, Jacky Wang, for reviewing drafts of this essay

Thoughts? Reach out to me via twitter or email : )