A Better Way of Understanding Systems

Or, why everyone is telling you a college degree is pointless so you should totally get one

Two people are unjustly sent to prison, a complicated system where surviving is not guaranteed and thriving is reserved for a select few.

Once there, the first one starts to learn the system. He gets a feel for which guards can be bribed to bring contraband, what gangs he can join and what gang he should fear, when and who to fight to get respect, which are the best cells and how to get sent to them, etc. After 20 years of fighting for survival he gets out of prison, gang tattoos covering him head to toe, looking old and scary, with no friends left on the outside except for some of his old prison buddies; he resigns himself to a short life of crime. He doesn’t fear going back to prison anymore, for he knows that system well and can thrive in it.

The second one takes a different approach to learning the system. He learns the shifts of all the guards, looks for where the security cameras are, where the gates are and when they open, how to buy smuggled civilian clothing, etc. After 3 months a small riot breaks out; he sees his lucky break and puts his escape plan into action. After a few days of running, crawling and hiding, he manages to hitch a ride to Mexico. He starts a new life as an English teacher, his American accent goes a long way in giving him credibility. He learns to enjoy his new country, pays a few dubious people to get a new identity as a guy named Gonzales, buys a home, marries and lives happily ever after.

It can be said that both people understood the system they were in very well, both of their achievements would classify them as “experts” on the subject. However, prisoner one had a toxic understanding of the system, whilst prisoner two had a productive understanding of the system.

This is, I think, the case with many systems. They are so complex that you have a lot of ways to approach understanding them. Even though the starting point might be similar, the end result could be completely different.

It’s often hard to figure out what the end results of a certain way of understanding are, but you can sometimes figure out what the end results of a way aren’t, before pursuing to the very end.

Yet, it’s often the easiest way of understanding that has the worst end results (more on this later), but it’s also the most tempting way to approach things.

Useful and Useless System to Understand

We are very good at finding pattern in noise. We are very good at understanding, navigating and building social structures. Yet most people waste this ability by choosing to focus it on the “wrong” thing.

If you are given the choice to analyze n systems, and you can get some surface intuition about how much you can use them once they are within your grasp, the seemingly obvious choice is to try and learn about those which seem most malleable to you.

Let’s look at politics for a second. Internal politics is a system which too many people struggle to understand and almost always end up with a useless (or even harmful) understanding of. Indeed, there’s hardly any way to understand internal politics in a “useful” way unless you have a tremendous amount of influence. After all, your vote only counts for so much, you could have a pitch-perfect understanding of how your country’s democracy operates , and you’d be no better off for having it (unless you work in politics).

Indeed, I’d argue that a rule of thumb such as: “Ask 3 people which you consider to be ethical and smart who to vote for, and if they all agree ask no further questions until the next election” is sufficient for the vast majority of people in terms of understanding internal politics.

Yet, in spite of them influencing this system so little, so many people struggle to understand it. So many people are proud about their choice in voting and would view me as a madman for voting the way I specified above.

On the other hand, a system that I spend some amount of time trying to understand is taxation. Understanding this system can be the difference between a life of debt and a life of wealth, between living close to poverty and living as part of the 1%, between life and death (though few countries still implement literal death sentences for filling taxes the wrong way, you can still get a lot of years in prison).

Happily enough, I have people helping me understand this system as well, but it’s one I take much more interest in, both in my country of residence and outside of it. Whilst I would not recommend myself as an international accountant, I think the vast majority of people would be better off heading my advice than going off on what they know.

Yet, out of all the people that spend hours a day reading about internal politics, few could name a good country to open a holding company, explain to me how the EU VAT laws work or name an advantageous fiscal entity for earning a specific type of income in a given country.

This seems to me like one of those near-paradoxes that prove how irrational we are. Both politics and taxation are highly interwoven. Indeed, an international tax expert probably gains a pretty solid understanding of geopolitics as a side effect of his research, and vice-versa. If you analyze the taxation system of a specific country past the surface level, it is often an indicator of its whole social structure.

Understanding politics very well gives you similar insight, but it gives you these insights from a perspective of helplessness. The insights are not a lever you can use to lead a more fulfilling-life (or at least a life with significantly more money to throw at problems), they are a burden you mostly just carry, forbidden knowledge which slowly gnaws at your sanity.

This is the second issue with how people learn to understand system, they not only choose the wrong way to understand them, but they can choose the wrong systems to understand. You can have two very similar systems, understanding them is tied to similar knowledge and yield similar insights, but learning to understand one will keep you helpless, whilst learning to understand the other will give you useful tools.

Useful Understanding is Uncommon

The problem with “useful” understanding is that it can never be common knowledge.

It’s either hard to realize that it’s useful, thus not many people know it (e.g. figuring out that you should learn how the internet works in the early 90s).

Or it’s difficult to learn if everyone knows it’s useful (e.g. figuring out how to build and repair a car engine).

Note: Here I mean an arbitrary combination of boring, takes a long time and requires certain peculiar mental abilities, resulting in “difficult”. Learning about taxes does not require you to be particularly smart, nor does it take a long time, but it’s about as interesting as watching paint dry. Learning the experiments and theory that lead us to the standard model is interesting and can probably be done without being a genius, but it might take a few years of your life to get there.

Even worse than that, I think most people don't know that they have the former “uncommon” type of understanding. The understanding may seem “obvious” to them, it may seem to be a part of understanding a greater system or a combination of other systems.

The obvious examples here are individuals which are very good at talking to others, charismatic people. Granted, charisma sometimes comes from looks or a pleasant voice, but I’ve seen plenty of mediocre looking guys with less than pleasant voices that are nonetheless amazing with people. Still, I don’t think I’ve ever seen one of these amazing individuals be able to put into words what makes them so amazing, or even realize they have this ability.

That’s most likely because charismatic people are often least likely to have a view of the world that contains a “human social interaction system”. Indeed, if you try to build a systematic model of how to interact with other humans, you will probably fail. You might end up with something like evolutionary psychologists, a bunch of researchers trying to understand human behavior from first-principles, yet paradoxically this supposes understanding is not enough to keep most of them from the list of people you’d least want to date or associate with.

You can however learn simple (albeit hard to follow) rules like: “If you try to always be nice to people, even if you don’t particularly like them, they will probably be nice to you”; OK, this is something I can’t back up with data, but it sure seems to apply to “magically charismatic” people I know.

The other issue with uncommon ways of understanding is that it can be hard to spread, since it looks “wrong” to most people.

This is a big pet peeve of mine, since I consider one of my few skills to be finding uncommon~ish ways of understanding systems; yet whenever I try to share various ways of understanding with people they almost always try to find a counter argument for it.

An obvious example of this that I hold near and dear to my heart is the need to have a degree to work in an unregulated tech career (e.g. web development, machine learning, data science, marketing, sales).I think an amazing example of this is the nonsensical ending analysis stack overflow ran on whether or not developers have college degrees.

They essentially find that:

Most programmers don’t have a degree, and even fewer have a degree in computer science
Most companies don’t require a degree, and even for those that do, it’s sometimes listed as optional (and, in my personal experience, most will make exceptions. I can’t find statistics on this, but I do take essentially all of my job offers as empirical evidence)
Whilst the article doesn’t say it, and I couldn’t find exact numbers on this, playing with SO’s salary calculator, seems to imply that a BA adds less than 10k to your salary and is equivalent to 4 years of experience (keywords used: none, javascript, linux + aws, pytorch + tensorflow). Not to mention this doesn’t account for confounders like age, IQ, social connections, location and wealth, all of which are positively correlated with having a college degree.

According to Nick Larsen, the person who seems to have provided the interpretation of the data, since the article is full of his quotes:

Regardless of the degree level in the requirement, it’s usually okay to not have that degree if you can show a history of success in the area relevant to the problems you’ll be solving at the company. Even for a job looking for the Ph.D level, a bunch of published papers relevant to the work they need done is more likely to get you hired than a Ph.D with fewer or non-relevant public articles.

The text goes on to state other things, such as the fact that “boot camps” are usually taken by people who already have a degree, rather than as an alternative to a degree.

It states that the barrier to entry is low enough for basically anybody who can code.

It raises no single point to indicate that spending 4+ years and hundreds of thousands of dollars on a degree is in any way a sane choice.

And it ends with….

So…a degree is pointless, right? Of course not. In fact, many would argue that the college experience is about much more than just getting a job when you graduate. We wouldn’t encourage anybody who is already in college to drop out, or anybody who wants to earn a degree to skip it and go straight to work instead. There are many reasons why, even if you don’t need one to get the job you’re after, you might want to pursue a college degree.

In other words:

“All the data shows that the common understanding of the system, which says a college degree is a signaling mechanism required to get a good job in the field, is wrong”

BUT

“Even though the less common understanding - that a degree as a social signaling mechanism or ,fart, a way of learning programming is not cost or time efficient - seems correct, it’s scary and we don’t recommend trusting it.”

There’s other examples of this uncommon vs common paradigm I can derive from the same survey. No, I can’t have the pleasure of quoting the very people who collected the data while they disagree with what their own data shows, in order to appease their internal representations.

For example, when people ask me career-related questions, I always try to cram in the advice: “Try your utmost best to network with and work for companies in the US, they almost always offer better pay”. This is a conclusion I’ve drawn from both the data I’ve seen (example) and from anecdotal evidence.

Yet most people will probably give advice about what programming language to learn next, or what field you should try to work in. It seems that the highest paid type of position internationally still earns less than the lowest pay type in the United States. The same is true for the programming language you work with: location is king.

But "learn X language" is a common type of understanding when it comes to a “career as programmer”, whilst "apply to companies based in Y country (even if the jobs is remote)" isn’t.

Why is that ?

I don’t know, I don’t have enough data here to prove that the uncommon understanding is better. The aforementioned surveys hint at it, but I think there’s less definitive answers on this subject than on college degrees.

However, I have a hunch that tells me this falls into the same trap; we are used with understanding the system as “based on what you know, you will find a nicer or worse job”, and understanding that says “what you know is not that relevant, where you live and where the companies you work for reside is much more important” seems weird, because it contradicts the common understanding.

But I digress, I don’t want to turn this article into “why getting a BA is probably a bad choice”. I reserve that for another time.

I also don’t want to sound like I am saying: “Any model of a system that seems nonsensical is bad and should be abandoned”. There’s a good argument to be made for why argument’s shouldn’t make us abandon common knowledge. But in that case, if we decide a piece of common knowledge is too sacred to touch, we shouldn’t strive to understand it.

If you live on a planet where everyone does {X} and doing {Y} will get you killed, you shouldn’t bother learning why this is. At best, you learn {X} indeed makes sense and, at worst, you gain the useless and potentially dangerous knowledge that {Y} is actually better. Accept the sacred common knowledge fo what it is.

Granted, curiosity for the sake of curiosity is often drenched in a veil of “fun” or “interesting”. I will never design a soda can factory, but I enjoyed seeing how one operates nonetheless. But unless we view this “common” understanding as (usually) useless, we risk falling into the trap where all the systems we understand are understood by everyone; we might have as well just followed the crowd and learned nothing.

At the end of the day, unlike most of my articles, I do have a strong takeaway from this one:

If you think you understand a system, ask yourself:

Will this understanding change the actions I take during my life in a good and significant way?
Will this understanding help me understand other systems for which #1 holds true?

If the answer to both is no, I don’t care what system you are modeling, it’s either the wrong system, a bad model, or both, and you should throw it away.

It could be a model of your national politics. It could be a model of why a given out-group is bad. It could be a model of how you’d run the world if you were god-emperor. It could be a model of traffic in your city and why the roads are designed terribly. It could be a model of the rat race which shows why you can never escape it.

It can be a model that’s fundamental to society and held by almost everyone, or it could be something you constructed for whatever internal reasons and only you or a small group hold.

The same rule applies: if you are certain it’s not actionable in any way and it won’t become actionable in any way, you should throw it away.

If it’s actionable but only insignificantly, you should at least try to simplify it.

The stoics tell us all things but ourselves are or can be fundamentally outside of our control, so we should learn to be content no matter the circumstances. I say that with the right mindset, we can all find thousands of things that are actually under our control (in a good way), no matter the circumstances.

If you enjoyed this article you might also like:

If you wish to discuss this article, feel free to use this reddit thread for doing so. There's a comment section bellow, it works, but it lacks a lot of functionality.

Published on: 1970-01-01

Reactions: