Sailboat Diaries

I’m in the process of transferring this site to a new web host. I started out with GoDaddy a year ago but decided to switch due to their support for SOPA/PIPA and willingness to completely betray their customers’ privacy. I’m on the new host now but working out all the kinks, hang tight and check back later if there are any problems.

The most powerful lessons are often unspoken. We learn a lot more from people’s actions than their words. Which makes perfect sense- language only developed in the past million years, possibly even as short as 100,000 years ago. But our ancestors have been learning by watching each other’s behavior since long before we developed a common tongue, and long before we achieved the level of consciousness we have today.

The big problem with formal schooling is that students learn more from its structure than its content. We learn much more from the implicit information embedded in the format of our schools than what we are verbally taught. The scary part is, we absorb and accept those implied instructions much easier than something we’re told, because it’s harder to consciously identify and reject them.

Whether intentional or not (and it’s almost certainly intentional), children are being taught some disturbing things in school. They’re not said out loud, they’re not in the textbooks, or on the homework assignments or tested on exams. It’s in the structure of the system.

So I’ll tell you the structural lessons I think I’ve identified, but I’ll admit it’s hard to say exactly how they affect students because they’re not openly discussed. The key point is that structure matters more than content; I’ll give you some examples but you should look at the structure of the system and try to reach your own conclusions. People can argue all day about whether Intelligent Design should be taught, but no one even considers whether teachers should be choosing topics to begin with, or whether students should be graded, or if they should be divided into age groups, or whether they should be required to attend, or whether they should adhere to a dress code, or any of the thousands of assumptions built into the structure of education schooling.

The biggest structural problem I see is who decides what is taught. Representatives and administrators choose the topics, teachers create their lesson plans, and students are expected to learn what the teacher decides. Students have very, very little say in their education, especially at the young end, but really all the way through college- you make a few choices at the beginning of the semester (if you’re lucky enough to have flexibility in your degree plan) and then you’re forced to follow the professor’s bidding. So people learn that their interests are irrelevant, that someone else knows what they need to know better than they do, that they should shut up and do what they’re told.

That’s an assumption I’m skeptical of. I don’t think anyone knows what the next generation needs to know. I don’t think anyone ever has or ever will, and without a doubt, no group of elected representatives has a clue. We just aren’t smart enough to make predictions about our economy 20 years into the future. So why should adults choose what children learn about?

The generic response is: “If adults didn’t force children to study math or history, they would just play video games instead.” So what? If that’s what the next generation wants to do, that’s what they should do. Why do people think knowledge of math or history is more important than knowledge of the maps in Call of Duty? If the next generation is happy playing video games, then we should restructure society so more video games can be played. I highly doubt that’s all kids would choose to do, but the whole point of education should be to help people make themselves happy. We have to understand what makes them happy before we can even begin figure out what knowledge could help them on their quest.

Next, formal education turns school into a game, complete with grades and rankings. Guess what- when you structure something like a game, people play it like a game. At least for me, school was never about learning. That just wasn’t even part of the equation, it was completely irrelevant to the task at hand. School was only about getting a high score, and you don’t have to learn to do that. In fact, it wasn’t even about getting a high score, it was about getting a good grade while expending minimal effort. I remember making a mathematical formula for it in high school (with help from Andrew Byrd). Performance = Grade/(Time-Invested)². Obviously this isn’t an appropriate attitude for learning, but that’s what happens when you structure education like a game. So it shouldn’t be at all surprising that people cheat and cram and procrastinate and bullshit, because that’s how you win.

We’re taught that knowledge is obtained from authority. Supposedly the teachers know all the solutions; every problem can be answered by referring to the back of the book or the teacher’s manual. The structure of the system cultivates a God Complex in each of us.

Even worse are the fears instilled by the current paradigm. Schools require the direct submission of students to the adults around them, using all the silly threats that work on children (uh oh, Timmy has to sign the book, or move his name into the yellow- ooooOOOOoooo). We’re taught to be passive, docile, easily manipulated- to respect fear authority. And if the punishments aren’t enough, they’ll drug you until you can follow orders.

We’re judged constantly, trained to look to others for approval, to seek praise from authority figures. Any deviation from the norm is immediately stamped out- no mustaches, no short skirts, no tank tops, no colored hair, and don’t you dare untuck your shirt in the holy sanctuary of an elementary school. We’re made to conform, and to fear standing out.

The list goes on and on. Whether or not you agree with the problems I pointed out, hopefully you can see how the structure of the system influences the minds of children, and we need to be careful what kids are learning from the situations we put them in.

Amazing TED talk highlighting a lot of these issues:

And if you’ve got a strong stomach, I highly recommend this lecture even though the sound quality isn’t the best:

I’ve made this post into a series of videos. I’ve embedded them within the post, you can scroll through and click the links to reveal each video, or click here to watch the entire playlist.

Watch the introductory video

The universe is not intuitive. Our senses are fallible and our assumptions can be faulty. What seems true usually isn’t. Luckily, we can build a cognitive toolbox to help us reason our way through this strange wilderness we call Reality.

Watch video 2 – Pruning our Belief-trees

But before we can start building, we have to decide on a purpose. A tool is a means to an end; before we pick the tools we have to know what job we’re working on. A cognitive toolbox will allow us to form and reshape our beliefs, “pruning our belief-tree”, but what do you want out of your beliefs? How would you like your belief-tree to grow? What fruits would you have it bear?

We can hold beliefs for various reasons. Beliefs can bring emotional comfort, or more commonly, avoid discomfort. They can influence and be influenced by our social connections. But we have to recognize that beliefs drive behavior, and if we want to behave “appropriately”, we have to hold beliefs that are consistent with Reality. As I wrote in Sanity Check, “holding ‘feel-good’ beliefs is nice, but it won’t help you interact with the universe.”

We take action in the physical world in order to accomplish some goal, consciously or unconsciously. What we believe determines how that goal is translated into behavior. We decide to take particular actions because we believe they will produce particular results. If our beliefs are not consistent with Reality, then the results of our actions will be inconsistent with our goals, and we will have acted “inappropriately”.

“You can ignore reality, but you can’t ignore the consequences of ignoring reality.” –Ayn Rand

The purpose we choose for our belief-system is a critical decision that will affect everything about our lives. The choice to use beliefs for emotional support is dangerous not only because we might act “inappropriately”, but because it could become a trap. Once you enter the maze of self-delusion, it’s incredibly difficult to find the exit. The safest choice is to plant (or re-plant) your belief-tree firmly in the bedrock of Reality. It might feel uncomfortable at first, but we can adapt, and soon grow to appreciate sanity.

Watch video 3 – Cogito Ergo Sum

But where to start? Where should we plant the seed of our belief-tree? We have to go to the very bottom, doubt everything that can be doubted and start from scratch. Let’s begin with the only thing we can know for certain: cogito ergo sum.

So I exist. Now what? Well unfortunately, that’s as far as we can get with absolute certainty. To believe anything else, we have to start making assumptions. But all assumptions are not created equal. We can reason about potential assumptions to determine whether they might help or hinder us. And that brings us to the first tool in our toolbox: Lex Parsimoniae, the Law of Parsimony, more commonly known as Occam’s Razor.

Lex Parsimoniae

Watch video 4 – The Law of Parsimony

Occam’s Razor is generally misinterpreted as “The simplest explanation is usually right”. This tool actually has nothing to do with what is true and what is false. Occam’s Razor only tells us what we should believe, not what actually is. Occam could care less about Reality, he wanted to help us build better models of it in our brains. This is an important distinction, because Reality is very, very complicated. The simplest explanation isn’t usually right. But the simplest explanation is the best explanation to believe.

Occam’s Razor tells us which story we should believe when we can’t tell any difference between their outcomes. But, if you can’t tell the difference between their outcomes, why does it matter which story you believe? We think of a story as a single unit, but they’re really made up of many smaller beliefs and assumptions. When a child finds presents beneath the tree on Christmas morning, there are at least two possible stories that fit the observations. One is that Santa Claus rode in a sleigh pulled by flying reindeer and delivered presents all across the world in a single night. The other is that the child’s parents bought the presents and put them under the tree. To believe the Santa story, you not only have to believe in Santa, but flying reindeer, elves, an endless toy sack, a fat man squeezing down the chimney, and possibly even faster-than-light travel. That’s a lot of assumptions to add to our belief system, and that’s where we can get into trouble: we could start basing new beliefs off of those false assumptions without ever testing them, and end up with a whole chain of incorrect beliefs (which could then make us act “inappropriately” in Reality). The most parsimonious explanation is the story that explains all the observations while introducing the fewest new assumptions, and that’s the safest story to believe. If we’re going to make an error, we should err on the side of disbelief, because it turns out our brains react very differently to different types of errors.

Types of Errors

All assumptions are not made equal because all mistakes are not made equal, especially not in the human mind. Our brains come with built-in cognitive biases that can blur our image of Reality. These are known, proven defects in the way the human mind works. All of us will predictably make these mistakes, unless we know enough about them to counter their effects. It turns out that many of our cognitive biases encourage us to form beliefs before we really should.

We’re pattern seekers. That’s what we do. Our brains are always looking for the signal in the noise, and our amplifiers are turned all the way up. Our environment selectively bred us for pattern-matching overdrive by killing off our less-observant (or more-skeptical), would-be ancestors. Believing things that are false wasn’t as dangerous to our ancestors as not-believing things that are true. Making more type I errors (false positive) than type II errors (false negative) turns out to be a useful trait for survival. The standard example is an early human that hears a rustling in the bushes. It’s extremely beneficial to assume it’s a tiger if it really is a tiger, and the penalty if it’s just the wind is low (which would be a type I error, believing something that’s actually false). If, however, the person doesn’t notice the pattern, or isn’t convinced it’s a tiger, and it is in fact a tiger (which would be a type II error, disbelieving something that’s actually true), then that DNA isn’t passed on.

Watch video 5 – Correcting Errors

This ingrained cognitive bias wreaks havoc on our modern minds. Now that we’re trying to understand the world instead of just pass on our genes, we have to work to overcome our instinctual irrationality. Not only do we tend to believe things before we should, our brains actively conform our perceptions to what we already believe.

This dancer can be seen spinning either clockwise or counter-clockwise, but once you see her going one way, it’s nearly impossible to see the other perspective. But unfortunately, this cognitive bias isn’t limited to our visual perceptions.

This is why it’s much more dangerous for modern humans to believe something that isn’t true (type I error, false positive) than to not-believe something that is true (type II error, false negative). With the former, your cognitive biases start working to conform your perception of reality to the false belief. Not only do we make more type I errors than we should, but they can become cascading errors that form the basis for more mistaken beliefs, and they all appear to be true.

Not-believing something that turns out true is much easier to fix, because once we find compelling evidence that it is in fact true we will be quick to adopt that doctrine. Our brains are constantly on pattern-seeking overdrive, and always working to conform our perceptions to what we already believe; we have to compensate by being skeptical.

Our First Assumption

Watch video 6 – Objective Reality Exists

Now that we’ve laid some groundwork concerning which types of errors we’re willing to make, it’s time to make our first assumption: objective reality exists.

We can’t know if there is anything outside our own minds. All we are aware of is our subjective experience. There is no way to know that our senses are conveying accurate information about the outside world to us. We could be brains in a vat connected to a reality-simulator, or plugged into the Matrix. For anyone who’s ever had a dream, we also know it’s possible to be fully convinced that our subjective experience is real, even when it isn’t. So why do we assume objective reality exists, and that our experience while awake reveals information about this outside world?

It’s a safe, useful assumption to make. We know we’re experiencing something, so it’s useful to try to understand that environment. Even if it’s not the “root” reality, even if we are plugged into the Matrix, this is the reality in which we act, so it would be helpful to have a model of it so we can behave appropriately in pursuit of our goals. And assuming we live in objective reality also fulfills Occam’s Razor: it’s the belief that best describes the evidence while introducing the fewest new assumptions. To believe we are in the Matrix would require new assumptions, and we wouldn’t be able to tell any difference from “the root Reality” anyway, so the Law of Parsimony suggests we assume the external world exists, and that we live in the most “basic” level of it.

Abstractions

Watch video 7 – Abstractions

Next, we can introduce a whole class of tools called ‘abstractions’ to help us understand this objective reality. An abstraction is a concept or idea that simplifies the underlying piece of Reality it represents. A cloud, for instance, is an obvious abstraction. In the physical world of atoms, there is no ‘cloud’. There is just a region of space with a higher concentration of water vapor molecules than the surrounding regions. It has no boundaries to distinguish it from the rest of the universe, and there’s no objective reason to draw a circle around it and call it a single entity. But we call it a cloud because it’s simpler.

The physical universe that (we assume) exists outside our minds is a world without ideas, words, concepts, groups, abstractions or systems. Reality is simply the physical substrate, the particles our universe is composed of: neutrinos, quarks, muons, etc. Each individual particle interacts with each other individual particle based solely on the Laws of Nature, of which we have discovered four fundamental forces. There aren’t even atoms in Reality, that’s an idea invented by humans to simplify the underlying physics of subatomic particles. Our feeble minds can’t begin to comprehend the universe it in its totality, so we simplify it. We organize our perceptions, infer causal relationships, and group things together into a consistent model of the universe. But our model is no where near the real thing.

It’s easy to get caught up in the human perspective, to think that our concepts accurately reflect reality. And sometimes they do. But far more often, they’re only useful tools. Without abstraction, we couldn’t understand anything. We couldn’t plan, think, organize or cooperate. But we have to recognize that abstractions aren’t the same as the things they simplify, and sometimes the devil’s in the details.

Bayesian Inference

Watch video 8 – Bayesian Inference

So we know our beliefs, concepts, and ideas are limited and fallible. How can we tell which ones are better? If every model we conceive of is wrong, which one should we use?

For once, a minister can help us decide what to believe. Reverend Thomas Bayes came up with a procedure for updating the confidence we have in our beliefs when new evidence becomes available. He recognized that the degree to which we believe something can be represented as a probability, where 1 is absolutely certain knowledge (like the belief that I exist, cogito ergo sum, or a tautology) and 0 is absolute certainty the belief is false (like a contradiction, 2+2=5). Every other belief is somewhere in between. We have some degree of belief, some level of confidence in our belief being true.

Bayesian inference is justified by the philosophy of Bayesian probability, which asserts that degrees of belief may be represented by probabilities, and that Bayes’ theorem provides the rational update given the evidence.

Here’s how Bayes’ Theorem works: let’s say we want to determine the probability that it’s raining, given the evidence that the sidewalk is wet. We can write that as “P(raining | sidewalk-wet)”, read “The probability it’s raining given the sidewalk is wet”, the ‘|’ symbol means “given”. What Bayes showed is that this probability is equal to “the probability that the sidewalk is wet, given that it’s raining, multiplied by the prior probability that it was raining, all divided by the prior probability that the sidewalk is wet.”

If we throw in some values, it becomes a little more obvious why this works. The probability that the sidewalk is wet if we know it’s raining is just about 1, maybe .99. Unless there’s some kind of cover, if it’s raining then the sidewalk will be wet. The prior probability that it’s raining depends on your location, but let’s say .1 (meaning at any given moment, there’s a 10% chance it’s raining). The prior probability that the sidewalk is wet is going to be higher than the probability of rain, because there are other ways for a sidewalk to get wet. Maybe sprinklers, or a leaking pipe, or perhaps it rained earlier but the water hasn’t evaporated. So let’s say the sidewalk has a .2 probability of being wet at any moment (meaning the sidewalk is wet 20% of the time). Now we can put all those pieces back together to get our answer, the probability that it’s raining given that we know the sidewalk is wet.

= .495
So knowing the sidewalk is wet lets us increase our belief that it’s raining from the standard .1 (since at any moment there’s a 10% chance of rain) to .495, even though there are other potential explanations for how the sidewalk got wet. It allowed us to improve the confidence we have in our belief that it’s raining from 1 in 10 to just about 50:50. We still don’t know it’s raining just by looking at the ground, but we can update our belief-system to account for the new evidence, and Bayes’ Theorem is the optimal way to do it. The concept of “prior probability” is crucial to understanding statistics. Let’s look at lung cancer for another example. What is the probability that someone develops lung cancer, given that they smoke cigarettes? We’d write that as P(cancer | cigarettes), which we now know is equivalent to

Based on the sources I could find, “nonsmokers account for 15% of lung cancer cases”, meaning the probability that a person smokes given that they have lung cancer is .85. “Based on rates from 2002-2004, 6.98% of men and women born today will be diagnosed with cancer of the lung and bronchus at some time during their lifetime.” So the prior probability of lung cancer is .07. “The CDC says in its Morbidity and Mortality Weekly Report that the prevalence of smoking fell in 2007 to 19.8%”, so the prior probability that someone smokes cigarettes is .2. Plugging that into our equation, we find that

= .2975
So the probability that someone develops lung cancer at some point in their lives given that they smoke cigarettes is just under 30%, which might seem lower than your intuition. The key is that the prior probability of lung cancer is so low, even though 85% of people with lung cancer are smokers, the number of smokers who actually develop lung cancer isn’t nearly as high. Bayes’ Theorem isn’t intuitive, but neither is the universe. The actual mechanics of Bayes’ Theorem isn’t as useful as the high-level insight it provides as to what it means to hold a belief, and how we can utilize evidence to improve our beliefs over time.  Our instincts about statistics and probabilities are so incredibly fallacious that we must disregard them completely if we are to act rationally. Just look at the gambling industry if you need any proof. Billions upon billions of dollars are handed away each year because people don’t understand the odds (or don’t understand how to act “appropriately” given the odds). If your goal is to maximize profits in a casino, the only rational choice is to not play (or own the casino).

Evidence

Watch video 9 – Evidence

Bayes’ Theorem helps us update our beliefs when we find new evidence, but what constitutes good evidence? There are a few key factors we can consider when evaluating new information, the most important being replication. Good evidence will reveal itself over and over again, by different people using different methods. Ancient people began to suspect the Earth was round as they watched ships disappear over the horizon hull-first, every single time. When a ship returned, the mast and sails were always the first to be sighted. This theory was tested with a variety of methods, like measuring the length of a shadow cast by a stick at two different places at the same time. On a flat Earth, two identical rods will cast the same shadow no matter where they are. If the surface is curved, then the shadows will fall at different lengths, because the sun will strike at different angles. Even more evidence came from the circular shadow the Earth casts on the moon during a lunar eclipse. When something’s true, it’s always true, and the evidence will mount on only one side of the scales.

Counter-Example

A single counter-example is enough to disprove any proposition. If someone claims that all odd numbers are prime, all you have to do is show them a 9. If someone proposes that old dogs can’t learn new tricks, teach one old dog a new trick and you’ve disproved their claim. This is a really handy tool for quickly evaluating a new theory. When you’re introduced to a claim, instead of considering whether it’s reasonable, or fits with what you know, just scan for a single counter-example. If you find one, you can throw that proposition away. If you don’t, then you can start to consider adopting that belief, provided there is ample evidence.

Anecdotal “Evidence”

Anecdotal evidence hardly deserves its second word. Witness testimony is just about the most unreliable piece of information you can obtain. People just can’t be trusted to relay good evidence, not because they are consciously trying to mislead you (although that happens too), but because “our senses are fallible and our assumptions can be faulty”. And worse, thanks to the cognitive biases I mentioned before, they may have made a type I error, and now their brains have already started conforming their perceptions to what they believe. People often truly believe things that turn out to be false. Again, we have to compensate by being skeptical. A story is just that, a story.

Systems

Watch video 10 – Systems

Our next class of tools helps us understand how parts interact to form a whole. A system is a type of abstraction that not only simplifies the underlying reality, but also captures the way in which the pieces interact with one another to form a whole that is more than the sum of its parts. Every part is a whole, and every whole is a part. The hard part to thinking about systems is recognizing all the different levels we can look at. We can see a neuron as a system, composed of many different molecules, all interacting as a coherent unit to achieve some function. Then we could zoom out a level to see a whole brain, which is clearly a system of its own, composed of a hundred billion neurons interacting to give the illusion of a single, conscious actor. We can even go a step higher, and look at an economic system comprised of millions or billions of these brains interacting to form a whole that is much more than the sum of its parts. Once we see all these different systems, we can start to ask how are they similar, and how are they different? We can ignore scale, and wonder about the structure between the parts of each system. There are a lot of distinctions we can make, but I’ll just highlight some that I’ve found most helpful.

Linear vs Nonlinear

Without getting too bogged down in the mathematical details, a linear system is one where a given series of inputs will always produce the same output, no matter what order they arrive in. Imagine receiving a series of instructions telling you which way to move. “North 3 steps, East 4, South 1, East 2, North 2, West 7″. It doesn’t matter which order you receive the instructions in, you will always end up 1 step West and 4 steps North of where you started. These systems usually manifest themselves in applications where you assume each part is interchangeable. If every screw is the same, it doesn’t matter which one comes next.

A nonlinear system is any system that doesn’t have this property, where the order of inputs matters. Most natural systems are nonlinear, take a human body for instance. Let’s say we filled our pantry with all the food and water we would need to survive a month. The order in which we consume those inputs matters a lot. If we drank all our water before we ate any of the food, we would either die sooner of water poisoning or later due to dehydration.

This is an important distinction because of its relation to predictability. Linear systems are very predictable. We humans understand linear systems. We can construct long, intricate assembly lines to build very complex things, because each individual part is interchangeable. We can control water pressure and build internal combustion engines and transfer information through cables, because it doesn’t matter which water droplet, or which gasoline molecule, or which electron comes first. Nonlinear systems are unpredictable by their very nature. You can’t “engineer” a solution to a nonlinear problem, the way you can with a linear system. A nonlinear system cannot be reduced into a set of quantifiable inputs, processes and outputs. These systems behave in an entirely different manner, and the consequences of actions can’t be accurately predicted. Systems like an economy, or an ecosystem, or the weather, have to be understood in light of this fundamental property of nonlinearity.

Top-Down vs Bottom-Up

Watch video 11 – Top-Down vs Bottom-Up

Many systems we’re familiar with are hierarchical, meaning orders flow from the top-down. The nodes in these systems are arranged in a pyramidal structure such that one is “above” its subordinates, and “below” its superiors. A modern corporation, or a government, or a military unit can be seen as hierarchical system. A general gives commands to a colonel who gives orders to a major, who is superior to his captains, each of whom leads lieutenants, and so on down the chain of command. Orders flow from the top-down so that relatively few actors at the top control the behavior of the entire system.

In contrast to top-down hierarchy is the idea of emergent order from the bottom-up. In these systems, nodes are connected to one another in a mutually beneficial, voluntary arrangement. Networks like this can achieve “spontaneous order”, or “self-organize” in such a way that they can be more efficient, sustainable, and scalable, than top-down hierarchies. We can see bottom-up order emerge in ant colonies, flocks of birds, schools of fish, the internet, our languages, and a free market economy (if only one existed). Information and decision-making is distributed throughout the system, and order emerges spontaneously from the interaction of individual nodes.

Isomorphisms

Watch video 12 – Isomorphisms

I wrote a long post called Isomorphisms, Formal Systems and Gödel’s Incompleteness Theorem where I wrote in depth about this tool, and how it can help us understand the fundamental limitations of knowledge expressed in Gödel’s Incompleteness Theorem. I’m copying the Isomorphism section here.

I think the best way to try and understand really complex things is to use isomorphisms. An isomorphism is a relationship between two systems that preserves the structure among the parts of each system. As Douglas Hofstadter put it,

“The word ‘isomorphism’ applies when two complex structures can be mapped onto each other, in such a way that to each part of one structure there is a corresponding part in the other structure, where ‘corresponding’ means that the two parts play similar roles in their respective structures.”

A school of fish and a flock of birds are isomorphic: they are both nonlinear, bottom-up systems made up of individual animals, from which emerges a “super-entity” when they work together in close groups.

Brains and Societies

The most interesting and helpful isomorphism I’ve discovered is the “structure-preserving relationship” between our brains and our societies. We can learn a lot about society by studying brains, and vice versa. These two nonlinear, bottom-up systems are structured in an incredibly similar way, and we can exploit this to better understand each of them. Both systems are networks composed of individual ‘nodes’ (people/neurons) which transfer symbols between themselves. The connections in both systems are dynamic, meaning the connections between nodes change as time progresses. You meet and communicate with different people on a daily basis, and neurons continually adjust their synapses. There are a variety of symbols transferred between nodes in both systems. In human societies, the symbols transferred between nodes are embedded in the way we position our bodies (body language), patterns of air pressure (speech), marks made using contrasting colors (writing), pieces of cloth with some of those marks on them (currency), certain elements (gold and silver), voltage carried on a wire (computer bits), electromagnetic radiation (wireless computer bits), the arrangement of colors displayed either on physical media (a painting) or electronically (TV/computer image), etc. We’ve learned to exchange symbols in a many different patterns, and through this exchange emerge our societies. Neurons also use many different forms of communication: dopamine, serotonin, oxytocin, acetylcholine, epinephrine, norepinephrine, histamine and many others. And out of all of this neural symbol exchange, we emerge. We are our neural society.

While the physical substrate that encodes symbols in brains and societies are very different, the overall structure of the systems are amazingly analogous. We can imagine both our brains and societies as giant graphs where each node is an individual (human/neuron) and each edge is a connection between individuals along which communication takes place.

The connectome (graph of neural connections) of C. elegans, a microscopic worm with 302 neurons and a little over 7000 connections

A small sub-network of our social connectome

By drawing analogies between complex structures, we can learn more about one system, say flocks of birds, by learning about its counterpart, schools of fish. We could learn how the individual actions taken by fish, and the feedback between them, results in the emergence of collective action, and we can then apply this knowledge to flock-dynamics. We’re finding a general pattern of interaction that can be applied to many different concepts. Isomorphisms grant us the ability to learn about structure rather than content. They let us learn about patterns instead of particulars. By discovering and investigating isomorphisms, we can leverage our understanding of different concepts by applying knowledge of one system to other analogous structures.

Perspectives

Watch video 13 – Perspectives

These are the lenses through which we see the world, shaping our beliefs and actions. The same underlying reality can look very different through a new perspective. The ability to see from multiple perspectives allows us to find our mistakes, and reshape our beliefs to be more consistent with Reality. We can practice shifting our perspective in all sorts of different directions.

One of my favorites is zooming-in and out, trying to imagine the universe at every scale. We can also try to see from different spatial perspectives, like trying to visualize a city from above, or thinking about where the moon is in relation to the Earth and the Sun just by looking at its current phase. Or we can practice seeing slices through time, like imagining the history of a tree from its modern form, or picturing the evolution of Homo Sapiens from Life’s common ancestor.

Of course, the most useful tool is being able to see through the lens of another self, from another human’s perspective. We can ask ourselves: What would I believe if I was a Roman citizen? If I was born in an Islamic country, would I be Muslim? How would I feel about Communism if I was raised in the Soviet Union? Is this really normal, or does it just seem normal to me? Some of the hardest perspectives to take are the ones that have no observer, like trying to walk a mile in someone else’s shoes when there are no shoes to walk in.

The Anthropic Principle

Watch video 14 – The Anthropic Principle

This idea generally relates to the notion that the Universe was “fine-tuned” for life. The fundamental forces in the Universe seem to be set arbitrarily to certain values, and if any of these had been changed, the Universe would have developed very differently. In many of these hypothetical universes, galaxies might never have formed, or even atoms. How lucky then are we, to be born into a universe with just the right conditions for life? But to fully evaluate that claim, we have to imagine a universe where life didn’t evolve. In a universe with different fundamental laws, where galaxies never formed, there would be no stars, and no planets orbiting them, and no beings to look up and marvel at how the Universe was fine-tuned for life. In the hypothetical universes where there is no life, there is no one there to report it. Only in a universe where intelligent life develops can a being question its own creation. So it’s not surprising at all that we find ourselves in a universe conducive to life.

The Discontinuity of Self

Watch video 15 – The Discontinuity of Self

I also wrote a (shorter) post solely addressing this point, The Discontinuity of Self, and I again plagiarized some of my younger-self’s writing (I’m sure he won’t mind).

An even harder perspective to take is one in which we are an illusion. Our own perspective is so pervasive and convincing, so mind-numbingly obvious that it can be difficult to doubt its veracity. But appearances can be deceiving, and the only rational explanation is that the self isn’t continuous, that our identity emerges anew every instant of every day. We’re so accustomed to our personal point-of-view that we grow to believe it is constant. But the self is a lens through which we see the universe, and it is perpetually shifting. The image only seems constant because the change is generally so gradual. Once we let go of the illusion of continuity and focus on the changes, we can start to see how our selves emerge from underlying processes.

Our identities are constructed by a vast collection of neural networks all working simultaneously. The way we act, the choices we make, the beliefs we form, are the high-level, external projections of a nonlinear system of neural interaction. We’re like the operating system of a highly parallel computer. We are formed by the physical operations of our internal hardware. An operating system is an abstraction that emerges from operations on single bits, and the self is an abstraction that emerges from the operations of individual neurons. We can think of our outward behavior and actions as the image that appears on the computer monitor, a way for the system to interact with the outside world. What shows up on the screen is the result of a coordinated effort by the CPU, RAM, motherboard, OS and hard drive. The image on the monitor can change gradually, giving the illusion of continuity within the system. But at least 60 times a second, each individual pixel is refreshed based on the calculations performed within the machine. It may appear as one congruent form, but that is simply an illusion.

Our behaviors, beliefs, and being emerge anew based on the current pattern of neural communication in our heads. Any semblance of continuity is merely an illusion. This isn’t just some semantic debate about the word ‘self’. The implications for the way we live are enormous.

Understanding that we aren’t the same person as before relieves us of our guilt. It wasn’t you that made those mistakes. It was your younger self, someone without all the knowledge and experience you have today. (Most importantly, the younger version is missing the knowledge of how that mistake played out.) Who’s to say how you (the you in your body right now) would have handled that situation if you were there? The only thing we can do is learn from those memories. We can’t change the past, and there’s no point feeling bad about things we can’t change. When you realize that’s the very best you can do at this point, it actually feels good to recognize the error and move on. You’re now a better person because of it, and the self that made the mistake is long gone. Granted, this perspective isn’t natural and it can be hard to achieve. But it is rational, and with practice we can learn to let go. Like overcoming fear, we have to simulate our desired reactions before we can integrate them into our selves.

Recognizing that we only exist right now can also be extremely liberating. Focusing exclusively on experiencing the moment can bring about enjoyable and enlightening mental states. Many forms of meditation involve the practitioner trying to become aware solely of their immediate perceptions, allowing thoughts, memories, feelings, and sometimes the entire sense of self, to wash past them. Experiences like that can give us great insight into what we really are, a brief peak behind the curtain, beyond the illusion of self. We can learn a lot from Buddha on the subject, but of course,

Believe nothing, no matter where you read it, or who said it, no matter if I have said it, unless it agrees with your own reason and your own common sense. -Buddha

Most helpful of all, recognizing our discontinuous selves can help us shed our fear of death. You will never die. The person in your body right now has no reason to fear death. Some other poor sap wearing your shoes in the future will get to experience that. And since you can’t do anything about it now, there’s no point worrying about it. Worrying about death before it’s happening is the only way it can have a negative affect on your life. We can leave our body’s later inhabitants to deal with the bitter end.

The same goes for anything that will happen in the future that we can’t do anything about right now. After a certain amount of preparation, you can’t change how you’re going to respond in a job interview. You might as well leave your future self to deal with it. After enough studying, you can’t do much to affect your performance on a test. Leave the worrying for the person who takes it. Anything we’re dreading is really a chore for a later version of you, so there’s no point feeling bad about it now. Your future self is going to have to deal with it regardless, don’t let it burden your current self as well. There are, of course, plenty of situations where we can put our future self in a better position. Practice today will make us better tomorrow. But if the self is discontinuous, and it’s not me who’s better off tomorrow, why should I bother?

Pay it Forward

Watch video 16 – Pay it Forward

The attitude we take with regards to our future selves affects our experience in the present. We can only live in this moment, but much of our happiness depends on the actions of our past self. I might be the one who takes a vacation, but it was only because my younger version saved up the money. So then, the real question isn’t “what should I do to make myself happy now?”, it’s “what attitude should I adopt so that each of my selves can be happiest?”

In a lot of cases, the answer to that question is to defer gratification. In general, the longer you wait, the better off the eventual reward. And if we always take that attitude, then eventually we’ll get to be the self that benefits from waiting. If throughout our lives we constantly defer gratification, then throughout our lives we will receive greater rewards than had we always pursued happiness in the moment. This principle is exemplified in the famous Marshmellow Test:
The incredible thing about this experiment was its ability to predict future success.

The first follow-up study, in 1988, showed that “preschool children who delayed gratification longer in the self-imposed delay paradigm, were described more than 10 years later by their parents as adolescents who were significantly more competent”. A second follow-up study, in 1990, showed that the ability to delay gratification also correlated with higher SAT scores.^[1] A 2011 study of the same participants indicates that the characteristic remains with the person for life. Additionally, brain imaging showed key differences between the two groups in two areas: the prefrontal cortex (more active in high delayers) and the ventral striatum (an area linked to addictions).^[6][7] -Wikipedia

Counter-intuitively, deferred gratification is a useful tool for being happier in the present.

The Selfish Meme

Watch video 17 – The Selfish Meme

There’s one more perspective I’d like to talk about, and that’s through the eyes of an idea.

A ‘meme’ is “an idea, behavior or style that spreads from person to person within a culture.” A meme acts as a unit for carrying cultural ideas, symbols or practices, which can be transmitted from one mind to another through writing, speech, gestures, rituals or other imitable phenomena. Supporters of the concept regard memes as cultural analogues to genes in that they self-replicate, mutate and respond to selective pressures.^[3] -Wikipedia

Like a gene, it’s hard to define the borders of a meme. Ideas aren’t exactly discrete units, although we can reason about them as if they were. And like genes, we can imagine ideas to be selfish-replicators, concerned solely with passing themselves on. Of course, memes don’t really think that way. They don’t think at all. But if you look at the behavior of memes, they appear to behave selfishly; they appear to want to spread themselves. Much like a gene or a virus appears to want to spread itself. Obviously these entities have no emotions. These are abstractions we use to simplify the physical world. A virus doesn’t want to replicate any more than a toilet wants to refill itself after flushing. But the viruses that are best at replicating become the most common viruses.

You never hear about the viruses that don’t replicate very well, because they didn’t replicate enough to get recognized. In the same way, you never hear about the ideas that don’t replicate, because ideas have to replicate for you to hear them. Metaphysical solipism, the assertion that no reality exists outside one’s own mind, isn’t a household word because no one tries to spread it. It’s a self-defeating meme. If you believe it, you don’t bother telling anyone. Evangelicalism, on the other hand, is a great replicator-meme. Belief in the idea leads people to want to spread belief in the idea. Memes can also spread via biological reproduction, the passing of an idea to one’s children. All else being equal, a belief that encourages large families is more likely to spread than one that encourages small families.

A good replicator-meme can quickly go viral, and spread across large regions, forming an evolutionarily stable strategy- a strategy that, once adopted by a large population, cannot be invaded by any alternative strategy that is initially rare. Memes have another interesting property, and that’s the potential for immortality. Like genes, memes can be passed on from generation to generation with little or no change in content. An idea can persist for millenia, even if it’s not true (again evidenced by religion). Once we’ve disregarded the illusion of a continuous self, maybe we can replace it with something better: an ecosystem of memes.

Watch video 18 – An Ecosystem of Memes

I was always open to the possibility that the meme might one day be developed into a proper hypothesis of the human mind. I did not know, before I read Consciousness Explained and Darwin’s Dangerous Idea by Daniel Dennett and then Susan Blackmore’s new book, The Meme Machine, how ambitious such a thesis might turn out to be. Dennett vividly evokes the image of the mind as a seething hotbed of memes. He even goes so far as to defend the hypothesis that “human consciousness is itself a huge complex of memes…”

When the meme began, in The Selfish Gene in 1976, the message was a negative one: genes aren’t the only pebbles on the Darwinian beach. In 1998, in Unweaving the Rainbow, I could be more positive: “There is an ecology of memes, a tropical rainforest of memes, a termite mound of memes. Memes don’t only leap from mind to mind by imitation, in culture. That is just the easily visible tip of the iceberg. They also thrive, multiply and compete within our minds. When we announce to the world a good idea, who knows what subconscious quasi-Darwinian selection has gone on behind the scenes inside our heads? Our minds are invaded by memes, as ancient bacteria invaded our ancestors’ cells and became mitochondria. Cheshire Cat-like, memes merge into our minds, even become our minds.” -Richard Dawkins, The Selfish Meme

Recognizing our memetic selves gives us a tantalizing opportunity: partial immortality. If we are a collection of memes, and memes can live as long as they replicate from person to person, then pieces of ourselves have the opportunity to live forever. As long as we transmit ourselves into the future, hopping from body to body in the form of ideas, we will never die.

This TED talk was a huge inspiration for this post: Dr. Derek Cabrera – How Thinking Works

I’ve made this post into a series of videos. I’ve embedded them within the post, you can scroll through and click the links to reveal each video, or click here to watch the entire playlist.

Watch the introductory video

The universe is not intuitive. Our senses are fallible and our assumptions can be faulty. What seems true usually isn’t. Luckily, we can build a cognitive toolbox to help us reason our way through this strange wilderness we call Reality.

Watch video 2 – Pruning our Belief-trees

But before we can start building, we have to decide on a purpose. A tool is a means to an end; before we pick the tools we have to know what job we’re working on. A cognitive toolbox will allow us to form and reshape our beliefs, “pruning our belief-tree”, but what do you want out of your beliefs? How would you like your belief-tree to grow? What fruits would you have it bear?

We can hold beliefs for various reasons. Beliefs can bring emotional comfort, or more commonly, avoid discomfort. They can influence and be influenced by our social connections. But we have to recognize that beliefs drive behavior, and if we want to behave “appropriately”, we have to hold beliefs that are consistent with Reality. As I wrote in Sanity Check, “holding ‘feel-good’ beliefs is nice, but it won’t help you interact with the universe.”

We take action in the physical world in order to accomplish some goal, consciously or unconsciously. What we believe determines how that goal is translated into behavior. We decide to take particular actions because we believe they will produce particular results. If our beliefs are not consistent with Reality, then the results of our actions will be inconsistent with our goals, and we will have acted “inappropriately”.

“You can ignore reality, but you can’t ignore the consequences of ignoring reality.” –Ayn Rand

The purpose we choose for our belief-system is a critical decision that will affect everything about our lives. The choice to use beliefs for emotional support is dangerous not only because we might act “inappropriately”, but because it could become a trap. Once you enter the maze of self-delusion, it’s incredibly difficult to find the exit. The safest choice is to plant (or re-plant) your belief-tree firmly in the bedrock of Reality. It might feel uncomfortable at first, but we can adapt, and soon grow to appreciate sanity.

Watch video 3 – Cogito Ergo Sum

But where to start? Where should we plant the seed of our belief-tree? We have to go to the very bottom, doubt everything that can be doubted and start from scratch. Let’s begin with the only thing we can know for certain: cogito ergo sum.

So I exist. Now what? Well unfortunately, that’s as far as we can get with absolute certainty. To believe anything else, we have to start making assumptions. But all assumptions are not created equal. We can reason about potential assumptions to determine whether they might help or hinder us. And that brings us to the first tool in our toolbox: Lex Parsimoniae, the Law of Parsimony, more commonly known as Occam’s Razor.

Lex Parsimoniae

Watch video 4 – The Law of Parsimony

Occam’s Razor is generally misinterpreted as “The simplest explanation is usually right”. This tool actually has nothing to do with what is true and what is false. Occam’s Razor only tells us what we should believe, not what actually is. Occam could care less about Reality, he wanted to help us build better models of it in our brains. This is an important distinction, because Reality is very, very complicated. The simplest explanation isn’t usually right. But the simplest explanation is the best explanation to believe.

Occam’s Razor tells us which story we should believe when we can’t tell any difference between their outcomes. But, if you can’t tell the difference between their outcomes, why does it matter which story you believe? We think of a story as a single unit, but they’re really made up of many smaller beliefs and assumptions. When a child finds presents beneath the tree on Christmas morning, there are at least two possible stories that fit the observations. One is that Santa Claus rode in a sleigh pulled by flying reindeer and delivered presents all across the world in a single night. The other is that the child’s parents bought the presents and put them under the tree. To believe the Santa story, you not only have to believe in Santa, but flying reindeer, elves, an endless toy sack, a fat man squeezing down the chimney, and possibly even faster-than-light travel. That’s a lot of assumptions to add to our belief system, and that’s where we can get into trouble: we could start basing new beliefs off of those false assumptions without ever testing them, and end up with a whole chain of incorrect beliefs (which could then make us act “inappropriately” in Reality). The most parsimonious explanation is the story that explains all the observations while introducing the fewest new assumptions, and that’s the safest story to believe. If we’re going to make an error, we should err on the side of disbelief, because it turns out our brains react very differently to different types of errors.

Types of Errors

All assumptions are not made equal because all mistakes are not made equal, especially not in the human mind. Our brains come with built-in cognitive biases that can blur our image of Reality. These are known, proven defects in the way the human mind works. All of us will predictably make these mistakes, unless we know enough about them to counter their effects. It turns out that many of our cognitive biases encourage us to form beliefs before we really should.

We’re pattern seekers. That’s what we do. Our brains are always looking for the signal in the noise, and our amplifiers are turned all the way up. Our environment selectively bred us for pattern-matching overdrive by killing off our less-observant (or more-skeptical), would-be ancestors. Believing things that are false wasn’t as dangerous to our ancestors as not-believing things that are true. Making more type I errors (false positive) than type II errors (false negative) turns out to be a useful trait for survival. The standard example is an early human that hears a rustling in the bushes. It’s extremely beneficial to assume it’s a tiger if it really is a tiger, and the penalty if it’s just the wind is low (which would be a type I error, believing something that’s actually false). If, however, the person doesn’t notice the pattern, or isn’t convinced it’s a tiger, and it is in fact a tiger (which would be a type II error, disbelieving something that’s actually true), then that DNA isn’t passed on.

Watch video 5 – Correcting Errors

This ingrained cognitive bias wreaks havoc on our modern minds. Now that we’re trying to understand the world instead of just pass on our genes, we have to work to overcome our instinctual irrationality. Not only do we tend to believe things before we should, our brains actively conform our perceptions to what we already believe.

This dancer can be seen spinning either clockwise or counter-clockwise, but once you see her going one way, it’s nearly impossible to see the other perspective. But unfortunately, this cognitive bias isn’t limited to our visual perceptions.

This is why it’s much more dangerous for modern humans to believe something that isn’t true (type I error, false positive) than to not-believe something that is true (type II error, false negative). With the former, your cognitive biases start working to conform your perception of reality to the false belief. Not only do we make more type I errors than we should, but they can become cascading errors that form the basis for more mistaken beliefs, and they all appear to be true.

Not-believing something that turns out true is much easier to fix, because once we find compelling evidence that it is in fact true we will be quick to adopt that doctrine. Our brains are constantly on pattern-seeking overdrive, and always working to conform our perceptions to what we already believe; we have to compensate by being skeptical.

Our First Assumption

Watch video 6 – Objective Reality Exists

Now that we’ve laid some groundwork concerning which types of errors we’re willing to make, it’s time to make our first assumption: objective reality exists.

We can’t know if there is anything outside our own minds. All we are aware of is our subjective experience. There is no way to know that our senses are conveying accurate information about the outside world to us. We could be brains in a vat connected to a reality-simulator, or plugged into the Matrix. For anyone who’s ever had a dream, we also know it’s possible to be fully convinced that our subjective experience is real, even when it isn’t. So why do we assume objective reality exists, and that our experience while awake reveals information about this outside world?

It’s a safe, useful assumption to make. We know we’re experiencing something, so it’s useful to try to understand that environment. Even if it’s not the “root” reality, even if we are plugged into the Matrix, this is the reality in which we act, so it would be helpful to have a model of it so we can behave appropriately in pursuit of our goals. And assuming we live in objective reality also fulfills Occam’s Razor: it’s the belief that best describes the evidence while introducing the fewest new assumptions. To believe we are in the Matrix would require new assumptions, and we wouldn’t be able to tell any difference from “the root Reality” anyway, so the Law of Parsimony suggests we assume the external world exists, and that we live in the most “basic” level of it.

Abstractions

Watch video 7 – Abstractions

Next, we can introduce a whole class of tools called ‘abstractions’ to help us understand this objective reality. An abstraction is a concept or idea that simplifies the underlying piece of Reality it represents. A cloud, for instance, is an obvious abstraction. In the physical world of atoms, there is no ‘cloud’. There is just a region of space with a higher concentration of water vapor molecules than the surrounding regions. It has no boundaries to distinguish it from the rest of the universe, and there’s no objective reason to draw a circle around it and call it a single entity. But we call it a cloud because it’s simpler.

The physical universe that (we assume) exists outside our minds is a world without ideas, words, concepts, groups, abstractions or systems. Reality is simply the physical substrate, the particles our universe is composed of: neutrinos, quarks, muons, etc. Each individual particle interacts with each other individual particle based solely on the Laws of Nature, of which we have discovered four fundamental forces. There aren’t even atoms in Reality, that’s an idea invented by humans to simplify the underlying physics of subatomic particles. Our feeble minds can’t begin to comprehend the universe it in its totality, so we simplify it. We organize our perceptions, infer causal relationships, and group things together into a consistent model of the universe. But our model is no where near the real thing.

It’s easy to get caught up in the human perspective, to think that our concepts accurately reflect reality. And sometimes they do. But far more often, they’re only useful tools. Without abstraction, we couldn’t understand anything. We couldn’t plan, think, organize or cooperate. But we have to recognize that abstractions aren’t the same as the things they simplify, and sometimes the devil’s in the details.

Bayesian Inference

Watch video 8 – Bayesian Inference

So we know our beliefs, concepts, and ideas are limited and fallible. How can we tell which ones are better? If every model we conceive of is wrong, which one should we use?

For once, a minister can help us decide what to believe. Reverend Thomas Bayes came up with a procedure for updating the confidence we have in our beliefs when new evidence becomes available. He recognized that the degree to which we believe something can be represented as a probability, where 1 is absolutely certain knowledge (like the belief that I exist, cogito ergo sum, or a tautology) and 0 is absolute certainty the belief is false (like a contradiction, 2+2=5). Every other belief is somewhere in between. We have some degree of belief, some level of confidence in our belief being true.

Bayesian inference is justified by the philosophy of Bayesian probability, which asserts that degrees of belief may be represented by probabilities, and that Bayes’ theorem provides the rational update given the evidence.

Here’s how Bayes’ Theorem works: let’s say we want to determine the probability that it’s raining, given the evidence that the sidewalk is wet. We can write that as “P(raining | sidewalk-wet)”, read “The probability it’s raining given the sidewalk is wet”, the ‘|’ symbol means “given”. What Bayes showed is that this probability is equal to “the probability that the sidewalk is wet, given that it’s raining, multiplied by the prior probability that it was raining, all divided by the prior probability that the sidewalk is wet.”

If we throw in some values, it becomes a little more obvious why this works. The probability that the sidewalk is wet if we know it’s raining is just about 1, maybe .99. Unless there’s some kind of cover, if it’s raining then the sidewalk will be wet. The prior probability that it’s raining depends on your location, but let’s say .1 (meaning at any given moment, there’s a 10% chance it’s raining). The prior probability that the sidewalk is wet is going to be higher than the probability of rain, because there are other ways for a sidewalk to get wet. Maybe sprinklers, or a leaking pipe, or perhaps it rained earlier but the water hasn’t evaporated. So let’s say the sidewalk has a .2 probability of being wet at any moment (meaning the sidewalk is wet 20% of the time). Now we can put all those pieces back together to get our answer, the probability that it’s raining given that we know the sidewalk is wet.

= .495
So knowing the sidewalk is wet lets us increase our belief that it’s raining from the standard .1 (since at any moment there’s a 10% chance of rain) to .495, even though there are other potential explanations for how the sidewalk got wet. It allowed us to improve the confidence we have in our belief that it’s raining from 1 in 10 to just about 50:50. We still don’t know it’s raining just by looking at the ground, but we can update our belief-system to account for the new evidence, and Bayes’ Theorem is the optimal way to do it. The concept of “prior probability” is crucial to understanding statistics. Let’s look at lung cancer for another example. What is the probability that someone develops lung cancer, given that they smoke cigarettes? We’d write that as P(cancer | cigarettes), which we now know is equivalent to

Based on the sources I could find, “nonsmokers account for 15% of lung cancer cases”, meaning the probability that a person smokes given that they have lung cancer is .85. “Based on rates from 2002-2004, 6.98% of men and women born tod

I updated the “Tree of Statism” because it stopped short of the root. Unless we help people overcome the fear that allowed the formation of a monopoly of courts and police, we’ll never free humanity.

The Matrix is a pretty great movie even right on the surface, but it’s even better when you can recognize the deep, rich metaphor it presents. The machines are the State. They represent government. The Matrix is “the world that has been pulled over your eyes to blind you from the truth… That you are a slave. Like everyone else you were born into bondage. Born into a prison that you cannot smell or taste or touch. A prison for your mind.”

I highly recommend re-watching it with this metaphor in mind. I get something new out of it every time I watch it.

Relevant: The trick

If a 12 year old can understand all this, so can you. Just think a little. It’s not that hard. Here are some of my relevant posts:
The proper relationship between the Federal Reserve and government (according to Alan Greenspan)
Purchasing Power
The Trick