Shannon’s information theory and Turing’s computer set the stage for the next generation of science. Largely in ways they wouldn’t have even predicted. Recently the idea that the world is a video game or simulation has caught on as a silly meme idea. Okay, whatever, some silly philosopher or tech billionaire thinks we live in a simulation, what does that even mean?
A more interesting question is why has this idea surfaced? What is it about our knowledge of the world that has made it start to seem feasible? In the 17th century to not be a Christian was insanity. Religion was by far the most reasonable and pragmatic explanation for human existence. For fun let’s Imagine a counterfactual world: You are born and grow up on a 500×500 foot plot of land, which has all you could need. Your parents and a few other families live there as well, and they can all trace their lineage back four generations, when one day two people appeared on the ground. If you go to the edge of your plot there is a reflective mirror you can’t go past. Your great-great-great grandparents said a nice man named Steven came down one day and told them he was creating a cool new experiment to create people whom he loved.
You would have to be one hell of a person to doubt that story, it’s incredibly reasonable. It is perhaps the simplest explanation. The idea of a small plot of land with reflective mirrors sounds weird to us, so it shocks our thinking into a new frame of reference. But the truth is it’s no weirder than reality, we’re just used to the strangeness.
But perhaps in this reality thousands of plots of land exist next to each other, and the reflective walls end up 3 miles, and spores were crossing the reflective barriers and seeding new plots with new life — super weird right? That seems way less likely than Steve making us. This is why so many brilliant people never asked the question. Eventually some weirdo tries to climb the reflective barrier and measure spore content with new technology. Eventually some weirdo comes up with an idea that species change over time.
Kuhn called this a paradigm shift. In his mind a paradigm shift is some profound change that sets off a new series of scientific and societal revelations. While he thought the paradigm shift was an essential aspect of science, to me it seems more of an empirical observation. Either way, it does seem to have been the case. Before Darwin there was Galileo. He would have been the guy who proved the reflective barriers actually don’t extend upwards infinitely in my counterfactual world.
These ideas began to chip away at the mystical. When something appears so far out of our ability to explain we usually don’t just play it cool and admit we have no idea. We prefer to make up some story. Galileo and Darwin are classic examples. I suspect though little things like the germ theory of medicine played a role as well. In the past they thought the plague was due to divine providence and punishment.
The funny thing is like relics of the past, these broken ideas still infect weaker or susceptible members of our population. We have people who are against vaccination and believe evolution is evil. In fact billions of people believe a bedouin warlord from 600 AD who was violently murderous and rapacious (in his defense they all were) had his heart cleaned by the angel Gabriel in the desert, and now follow his life advice. I’m coming far closer to edgy internet atheism than I would like. Honestly religious people are some of the best I know, Alyosha Karamazov for example. Still, I think it’s a fun way to describe Islam due to the current Western obsession with acceptance.
Now Shannon and Turing come along. What if the universe can be explained by information content? What if using binary information we can create machines that compute information in highly complex and dynamic ways? If you showed Turning and Shannon World of Warcraft I bet they would have never predicted the dynamic interaction of their ideas could result in a basic simulation of reality.
Building on these ideas Kalman created a filtering device, which was based on the concept of filtering information from complex and noisy environments. Now we have adaptive neural networks. A model class that on at least a few dimensions seems a lot similar to our brain. The quantitative nerds of the 20th century had math, but their math was distinct from history and literature. It’s use in the social sciences was at times remarkable, but still distinct from the messy noise of reality.
The mathematical modelling of the 20th and 21st century the goal was for a clever human to conjure up the near inexplicable magic of his brain to solve and produce mathematical ideas that match our economic and political realities. My previous boss was an academic, and when I asked him how he came up with solutions to his stochastic differential equations he said “I look up to the sky at God, and I say ‘God, give me a solution to try.’”
We could model nuclear war using game theory, or run regressions on economic trade outcomes. Often at arbitrarily high levels of mathematical equilibrium refinement, but all the work following Nash wasn’t exactly blowing open the hatches of reality and letting us stare at the fundamental truths of human behavior. Stochastic differential games are cool, and pretty tough to crack, but as smart as those mathematicians are they weren’t shifting the paradigm.
With our new models in the 21st century we are getting closer to blowing that hatch open. Or at least we can see where the tunnel might lead. When we can create a computer that can study, classify, and interpret human social sciences better than a human brain we will begin to put the nails in the coffin of guys like Hegel and Marx. At least we can prove them wrong. We can’t really prove them wrong now, because we have no sufficiently advanced method that can bring precision to these grand overarching solution concepts that as of now only the human brain can tackle.
Philosopher David Stove wrote “For even in a single paragraph of Hegel, say, there is, presumably, not just one thing that has gone wrong, but half a dozen things which have all gone wrong together; while we are not able to identify a single one of them.”
It’s still hard to imagine what a proof of incorrectness would look like, but a computer that could scan all the information of history that operates stronger than a human brain could at least undetermine Hegel’s arguments by dynamically learning and finding counter-examples to his historiography.
Suddenly the mysticism of our brain starts to die. No longer will the brains of some revered theorist, philosopher, or economist be held up as magical black boxes of understanding. Debates on what Democracy, on what justice means, were viewed as meaningful and worth having, rather than simply our computational brains working together to filter out an abstraction that’s meaningful to us from a noisy world.
And what happens to these concepts if we kill the magic of our brains? Our brains deal in algorithms. Lots of the high-dimensional correlation and matching features of our brains far exceed the models we have. Yet machine learning models are slowly starting to overtake and match the human brain. Machines can drive a car now. Gone are the days of the mystical ability of eyes and interaction. Now we know that sight is simply light information processed by our brain. If we take that same information and codify it and have it be processed by a computer is this not, in every meaningful way, the same sight?
Our world is closer to The Matrix than our great thinkers ever understood. Instead of inexplicably sophisticated humans using language and knowledge to define human rights, government, and beauty, we are evolved computers living in a world defined only by information. If art is beautiful it is due to an evolutionary quirk in our programming. If Bach is extraordinary it’s due to his gifted and well-trained computer used to generate melodies we evolved to find beautiful. If justice and human rights are important, it’s because our experiences lead us to identifying a pattern of information that we believe improves outcomes, classifying it as justice or rights, and claiming it is important.
If this is all true, how would we view our current political system and debates? I imagine it all as complex atomistic functions (humans) interacting in a simulation. I’m hardly the first person to imagine it this way, current Political Scientists code and study simulations. Their simulations are too rudimentary to capture the complexity of the world. So we’re left with our brains, which can run bastardized simulations.
We can’t come to precise estimates, we can’t meaningful interact people in our brain in a mathematical way, yet we can consider highly-dimensional data. Our brains won’t tell us if we have enough data, and it’s hard to validate our model. One heuristic I use is for every hour of additional data I add by studying the simulation at hand, how much does my outcome change? If it changes a lot my brain hasn’t converged, if it changes little my brain is converging.
To run a proper simulation in your brain you need to be able to completely detach yourself from who you are, from your own emotions, wants, biases, and interests. This is the most fundamental reason I know of why activists cannot be strong scientists — at least not in the realm they work as an activist. They are so rooted in their conception of the world they struggle immensely to view the world through someone else’s eyes. It’s funny that cartoonists are great at it, maybe because cartoonists need to understand the world from many perspectives.
My favorite example of the activism and science dichotomy is in George Orwell’s relationship with the British far left and communist party. I mean, you had to be a socialist in Britain in the early 20th century. For essentially the first time in history there was a massive consumer surplus at the top of the social distribution, industrialization was not meaningfully improving the lives of the working class, and the idea of socialism had been born yet wasn’t tested. The main debate among the far left was to go full communist, or to try to integrate the best pieces as social policies.
This is part of why progressives drive me absolutely mad. I was talking to a friend recently about how it’s not fun bashing on Christian conservatives in the U.S. because it’s too easy. Scott Alexander wrote
[E]ven though conservatives seem to be wrong about everything, often in horrible or hateful ways, they seem like probably mostly decent people deep down, whereas I have to physically restrain myself from going on Glenn Beck style rants about how much I hate leftists and how much they are ruining everything. Right is the New Left
His explanation though is that a small subgroup of people are interested in being contrarian, and the pendulum swings back and forth. He explains it with an analogy of cellular autonoma, it’s worth reading. It has a valid claim towards being correct, but if it’s wrong I think this is why:
Progressives essentially run our country. Their reach doesn’t run through every town, so the idea is not absolute. If you are a woman who wants reproductive health care you justly want the ideas of progressivism to exist across all towns. I might disagree with a progressive philosophy of science, but to set the stage that doesn’t imply I am somehow radically against their every goal and fight.
Progressive ideas do control all of our top universities, our social policy, our foreign policy, our primary media, and have a strong claim on our richest, smartest, and most fashionable. Control sounds sinister, but it’s not sinister or a conspiracy, it’s simply a natural outcome that I believe is based on an idea–a model– of progressivism thought, that stretches back centuries. Moldbug calls it The Cathedral — it’s impossible to beat his explanation.
It is based on a religious belief that progress is clear and measurable, there is a set of actions to achieve it, and so long as we work towards these goals our world and country will improve year over year.
It forces you to view your opponents in one of two categories: Either they must be evil or they must be stupid. There is no alternative. If you replace the word stupid with uneducated though, it lets you feign a sort of condescending empathy. If you look at British Leave voters you can see half the articles are on racist bigots, and the other half on a campaign of misinformation and lies interacting with the uneducated masses. I used google to get those two links, there are thousands. It’s the hot thing to do to write one of two archetypal posts if you’re a smart progressive who is horribly depressed.
If the progressive classification of history is wrong, how is it wrong? And where did it go wrong? There is no reason it can’t be fixed either. We fix things on the fly all the time. I’m not going to suggest the answer is to burn it to the ground. I’m honestly not sure it can be fixed, but that’s an empirical question.