Facets of Greatness - Knowledge - Meta and Epi

λ  (  §µßƍŁ¥þĦ ԠƎ  )

I’m going to start off by discussing a concept that may seem hard to understand and a bit philosophical, but that I believe to be irreplaceable. Then I’ll describe how to use models of the world around us that we all form. However, what most people miss is that it isn’t just how you understand the world or some concept that is important. You must understand how other people have modeled this concept because that is where you’ll find the fuel for change.

Metaphenomena and Epiphenomena

Metadata and Epidata

Knowledge is Power

Harnessing the Knowledge Differential

Competitive Advantage and Ouroboros

Baller Blockin’

You’re Going to Do What to My Dividends?

There’s a quote from someone like Socrates or Einstein and I can’t find it. But basically, this person stated that the most intelligent people don’t just understand some topic, they understand how other people understand it. You’ll see this when someone who is highly intelligent debates someone else on some controversial topic because they will be empathetic towards how that other person sees the world. This is because that this intelligent person understands that other persons model and viewpoint. This shows up in politics quite often because we’re all truly trying to solve the same problems: poverty, disease, etc. So it helps to be able to understand how someone debating against you models some problem and it’s solution. You’ll also find that the most intelligent, but biased people quite clearly comprehend the differences between their solution and others because they strategically embue that understanding into their rhetoric, all the while refusing to lose ground to that other person. These people are incredibly smart, but not likely to listen.

Metaphenomena and Epiphenomena

There’s a concept that I believe is incredibly important to understanding the world around us. It’s the meta, which is a prefix based on the greek word for “after.” Aristotle used it to distinguish his work Physics from the concepts discussed in Metaphysics, meaning that which arose after physics. The greeks did not use meta in the sense that we do with english. We instead refer to meta as within or a denotation of self-reference: metaprogramming is about programs that write programs. Metadata is data about data.

This concept of the meta is fairly well-known, but I believe that meta has a dual concept which is so similar, it’s usually overlooked. And that is the concept of epi. So, where meta drills down into the phenomena which are self-referential in some phenomenon, epi goes in the opposite direction. It’s a concept which is so similar that it’s often just referred to as meta. It’s very difficult to distinguish, so I’m not sure I’ll nail this correctly, but the concept of epi and it’s relationship as dual to meta is necessary.

Metadata and Epidata

The metaphenomena of some phenomena are really the mappings between it and itself that establish some whole or partial self-similarity – it’s kind of like an automorphism in category theory. It’s a bit easier for me to explain some of these concepts in those abstract math terms, but I’ll refrain for the sake of retaining simplicity – even though that’s hopeless, amirite? The epiphenomena of some phenomena are instead transformative. That is, the epiphenomena maps a pair or set of phenomena. It requires mapping some phenomena to another, while striving to preserve self-similarity by minimizing the number of changes required for and induced by the transformation.

In this sense, some kinds of metadata are not metadata at all. That is a misnomer and they are epidata, which maybe seems trivial, but it’s important because as an indication of the higher-level distinction between meta and epi, the exact definition of which I am still searching for. Metadata for a set of data cannot involve mappings to other sets of data. The metadata for a text file could contain aggregates that only involve data from that text file: a count of words, characters and sentences. Another example would be the topics contained in the text file. The epidata for a text file could contain the set of differences between it and another text file. An example of something that is traditionally thought of as metadata are creation and modification timestamps. But is this really data about that specific data and how so? It’s truly the data about that piece of data as it relates to some other external system: time. In the same sense: would the number of shares and tweets for this blog article be metadata? IMO, it is epidata because it involves data about the relationship between that specific piece of data and some external system.

Establishing a distinction between metadata and epidata is crucial because it emphasizes the nature of the dependency graph required for producing some piece of data, which today would be simply known as metadata. That is, the piece of epidata that indicates the aggregate number of shares for some blog requires access to the data for each share. If you don’t have some level of access to that entire data set and some mechanism for temporary storage, then you can’t produce that piece of epidata. This notion of metadata and epidata can be applied at various levels: you can have metadata about some set of data, which would really be epidata about some piece of data in that set.

The reason I’m even discussing metadata and epidata in this article is because it’s a great example for drawing a distinction between the concepts of meta and epi. The utility of these two concepts is irreplaceable in terms of making sense of our world. The ultimate truth is that every concept can be defined in terms of it’s relationship to other concepts and their aggregates. So, if you understand how to look at some thing and draw parallels to some other things that you already know, then you can use these parallels as an incredibly useful tool to understand that new thing.


The concept of meta and epi maximize your ability to make inferences about some new thing, using only your existing set of knowledge. The only requirement is a familiarity with your mind and what you already know. And optimally, you’ve also acquired a valuable set of knowledge that propels you forward. But even if you don’t have it, you always have a chance to learn. I cannot emphacize my enthusiasm for access to free knowledge online. Tools like Coursera and random lecture videos on YouTube have been invaluable for me. I would not understand the world as I do if I had never taken Coursera’s Machine Learning and Cryptography classes in 2012. Those classes, amoung others, provided me with tools for reasoning that I just wouldn’t comprehend if I hadn’t taken them.

Knowledge is Power

These great people people value knowledge and systematize it’s acquisition. They’re intimately aware of what they don’t know. They understand how to apply knowledge for gain. They have the wisdom to deftly perform when they don’t have the knowledge.

They construct strategies based on what they learn. Then they enumerate the factors and dependencies of the factors in that strategy. They understand that in some industries or trades, there are factors that are critical and there are factors that seem critical, but fall away over time.

Most people believe that particular indicators and statistics drive markets, prices, etc. Usually the commonly understood model of something is mostly true, but it’s a facade that hides the deeper, more complex machinery that truly drives the secret motion of things. The commonly understood factors that drive business or drive human behavior are too easy. There isn’t much value to be derived if modeling a business on an oversimplistic model – if everyone understands it, where’s the opportunity? If that’s all you had to know to win, then everyone would do it – whatever it is: business, music, sales, acting, etc.

The commonly understood factors are just the basics and math is often the key to modeling systems faster and more accurately. Usually the common model is based on a limited set of math. It’s accuracy and predictions are limited because of this. Finding true success means that you have to understand everything at a much deeper level. Once you have this deeper understanding, then you can exploit that knowledge differential towards your success.

Harnessing the Knowledge Differential

This knowledge differential is the difference between your deeper understanding of a subject against that of the commonly understood model for your subject of interest. But you need to be intimately familiar with each model! This is so you can understand the opportunities created by the differences in your model vs. the common model. And more important, you need to know what leads someone to believe the common model accurately models a system and why. You need to know how people perceive errors in their model of a system and why. Often this deeper understanding of a trade, a skill or a industry comes from experience. Thousands of hours on a subject.

Furthermore there are accurate models of subjects that just aren’t commonly understood yet. So here, your more complete understanding gives you an advantage. You can profit from the disparity in knowledge here – but don’t be evil, lol! All joking aside, evil behavior here includes misleading other people and network marketing schemes. Pure evil, unless your cousin pushes CrackStar or something.

Competitive Advantage and Ouroboros

But you have to know that, usually, competitive advantage in this case doesn’t last forever, if you depend on some better understanding of how a market, product or industry works. If other people manage to find a better solution to the problem your system solves or if they find a better model to the system themselves, then your opportunity begins to fade. So, unless you have IP that protects you, your advantage here might only last until a more accurate model of that system becomes the commonly accepted one. Social media and SEO are good examples here. Some techniques work until they are commonly known. And in these cases, you need to posture yourself and your organization for a healthy transition. It’s your responsibility to your employees and your organization.

Also, there are factors which are perceived as being driving forces and strong indicators which aren’t really that useful. If you know how people determine what is important in a marketplace, then you can profit from the disparity between how they think it works and how you know it works. That is, again, you profit from the knowledge disparity there – but what is crucial is knowing when certain factors in a model are going to be proper indications. If you have experience in knowing when a factor will determine a system’s behavior and you understand how this affects people’s actions, then you can profit.

A general example: Factor A is usually a good indicator of demand for Product X from Industry Y. It’s commonly accepted knowledge for stock brokers. Maybe it’s a dip in price for a particular commodity. But your model says: this is true, except when Factor B and C supercede A. In which case, it’s actually an indication of lower demand for Product X. Then, you know the price is going to rise from people’s expectations and their misguided purchase of stock from Industry Y. So you know to sell your stock when the price rises from the aggregate effect of purchases. Then, you wait for the inevitable drop in price that your model predicts. Boom. Purchase more shares. Make more money. Simple as that (not really)

All of this “factor” and “model” bs is very generic, i know. But it’s very usefull in more than just the stock market. Everyone wants to participate in Activity X or Trade Y wants to build a mental model of how it works, especially if they want to participate competitively. Some of these models are good and some are bad. And some people just exhibit irratic behavior. But:

★ Understanding what people believe

★ Inferring their belief structure

★ Comparing it to others

★ Generalizing belief structures

★ Comparing and contrasting them

★ And understanding how that affects actions and results

These are crucial to identifying opportunity and mitigating risk in nearly any market, any subject, any skill, any activity.

Furthermore, the effect these factors and models have grows and wanes. People who are successful understand why Factor A seems dominant now, but that it is Factor B which will be dominant in a decade. So now is a good time to prepare a venture that expects to profit from Factor B. While everyone else is expecting Factor A to determine the market for decades, this signals an opportunity to prepare for profiting from the disparity in market forces there.

Yet, the only way you can capitalize here is by having the resources to invest. If you don’t have resources to invest while the opportunity is open, you can’t exactly reap the rewards, nor can you insulate yourself from a changing market. In other words, you are like a leaf in the wind, subject to forces beyond your control.

This is also why I’m so frustrated with being ignored when I had so many startup ideas over the years. It’s only really an opportunity until someone dominates the market. I believe while there is a plethora of opportunities now, the barrier to entry for technical startups is going to rise slowly over the next decade.

Baller Blockin’

They Straight Up Baller Blockin’

And it’s crucial to diversify your portfolio when market conditions seem likely to change. There’s an important conflict of interest here that many larger entities have: usually they don’t want to compete with themselves. Why would they develop Product B which would return 20% the profits now when they could just shove their heads in the sand. Many businesses at this point pursue other strategies to mitigate a new technology or product because they are not incentivized to develop it. In fact, it’s quantifiably more profitable – knowing this technology is inevitable – to stave it off as long as possible than it would be to develop it themselves. That is, if the profit rate for Product A is really that much higher than Product B. This is a complicated scenario with several other courses of action, including purchasing the competitor’s IP and snuffing that baby in the cradle.

But, my point is – heads. in. sand. In this case, it’s a hard sell to a board of directors to invest in this risky new technology that might slash through profit margins in 5 years or so. Why would they do that when they got a good thing goin on? And convincing a public company to do this, when the public shareholders have considerable influence over the direction of a company?

You’re Going to Do What to My Dividends?

Truthfully, the situation described above, where a company stagnates on innovation because of conflict of interest probably doesn’t play out that way. But often it can. Yet, this signals more opportunity. If you know that a new technology is going to revolutionize an industry and that the juggernauts of that industry can’t shift gears fast enough to capitalize on it, that’s an opportunity.

Basically, to sum it up: you gotta know which way the wind blows to understand how things are going to play out. Knowledge is key to constructing an accurate vision of how the world will change. Instagram was an opportunity partially created because Apple updated iOS to include the ability to apply filters to photos using OpenGL, AFAIK. Basically, if you know what new technologies are coming down the pipeline, you can prepare to take advantage of them later. So if you know that adding OpenGL to iOS will enable features like instant photo filters, you can prepare to code that shit as fast as you can out of the gate. But the key here is to stay in tune with upcoming tech and understand how it will affect people’s use of technology.

And knowing which way the wind blows means knowing which market forces are going to be dominant and which are going to fall away. It’s similar to complexity theory (aka Big-O Notation) for algorithms. Some aspects of an algorithm might seem important and expensive at first glance. But in the end, when you apply that algorithm at scale, those aspects just don’t matter. And this is one lesson to be learned by employing economies of scale: some factors, like cost of distribution or whatever, just fall away when your enterprise is structure to employ economies of scale.

And understanding what market forces are dominant means sorting through the sea of data and sorting through the noise. People have vested interest in getting you to believe something is true. You have to sort through all that and understand their incentives, behavior, beliefs and signaling to find the truth.

Next Up:


Next up on Facets of Greatness: Knowledge, I’m going to deconstruct a youtube video I made some time ago on how to become intelligent. In the meantime, I’m going to write about some of the other facets, such as communication and signaling.