ESSAY 10: ABSTRACTION VS FIRST PRINCIPLES

The path less (and most) traveled; the only answer is there is no answer

Abstraction vs First Principles

At the beginning of the year, I was really interested in two ideas: Abstraction and First Principles. These both are necessary modes of thinking for human beings, but they often counter each other. I will say that I went looking and looked very hard. The only ultimate truth I found is that there are no ultimate truths.

I can’t really imagine a brain that could only do one or the other, they are both necessary.

When I think about abstraction, to me it means creating heuristics or rules about a collection of data points. It is a practice of pattern recognition, extrapolation or interpolation or however the brain works that I don’t understand. Humans are simple creatures, and we package these heuristics into symbols that we reference for later use.

Convolutional Neural Networks Work Similarly

There is also the idea of first principles, which is zooming in on each of the data points and questioning our overarching theories. Since these heuristics are imperfect, they can cause misinterpretations or miss out on key nuances. It is necessary for an engineer to approach problems in first principles, and we should approach problems in fields like literature and politics all the same.

Tend to First Principles – The Resource Tradeoff

First principles is theoretically the best course of action. But we are faced with resource and time constraints, so humans often defer to abstraction or mimetically deferring to others assumptions. It would be hard to imagine waking up and efficiently going to work every day without having these unquestioned, instinctual behaviors.

Problems Due to Abstraction - Rules were meant to be broken.

Many problems in our society are due to deferring to symbols of understanding and abstraction. If you think about it, the Constitution is largely used for the purpose of setting standard laws that should remain true over time, while leaving room for change. As the country evolves over millennia, some of these are going to become outdated. Many heuristic dissolve over time.

We maybe understand that we like a certain ideology and so we vote for a party that falls under the abstraction of democrat or conservative. That’s not why we vote though? The party is just a way to divide a spectrum of policies into two buckets. However, many don’t really investigate whether the candidate’s stance mirrors their own. Further, these stances might be cloudy takes that don’t address the real problem. “I am going to stop all wars” avoids the perfectly reasonable fact that wars are sometimes necessary to survive. Politics is plagued by this issue because the population of interest / target market you are selling to isn’t knowledgeable. Democracy has its flaws.

In chess, we make rules like keep the knights in the center of the board, get rooks on open lines and control the center. Play chess against a computer engine, Magnus, Kasparov or Tal and you will see powerful counterintuitive moves. A dim knight on the rim has made my life miserable on multiple occasions.

Similarly in business, you may learn general rules like never compromise on hires, the customer is always right, and to always align all stakeholders. Just to make an example with one of these, the customer is not always right.

Every Situation is Unique

I think the largest realization I’ve had recently is that I spend so much time learning from others from any source I can find, looking desperately for “ultimate truths of business.”

Business isn’t mathematics. The reality is there are no ultimate truths of business. Every industry, person, and problem is inherently distinct. A quote that I will reapply from Peter Thiel here is,

“The next Bill Gates will not start an operating system. The next Larry Page won't start a search engine. The next Mark Zuckerberg won't start a social network company. If you are copying these people, you are not learning from them.”

Heuristics are still helpful. Jeff Bezos used two heuristics that generally hold up. The assumption that customers will always want their items to arrive faster and cheaper. And that if you improve on those two metrics you can never go wrong.

Working hard tends to be pretty good advice as well. I’ve spent a lot of time learning from those who have found success in software and finance, but I think that the necessary next step is to unlearn many lessons and create my own framework that will work well for me.

This problem goes much deeper and permeates how we think about everything

I think this problem is sneakily much larger than even the above lets off. Humans are goal- and purpose-oriented creatures. The problem is that humans no longer use religion as the great north star of our internal compass, we fill it with other concepts or identities. Whether that is to be a good person, to get rich, to be a good student, etc.

These questions actually get infinitely complex when you break them down into cells and atoms. What is the essence of why we actually want to get rich? Why do we work? What does it mean to be a good person and whose ethical framework do we want to abide by? Is being a good person dependent on what we do in the moment or do the ends justify the means? Once again, if we were actually original in every way of life, it would be paralyzing to try to analyze everything, so we rely upon heuristics.

People often look to others who have become rich and try to emulate their strategies. The problem is that these strategies may not always work the same in the next generation. In the gold rush, wall street boom, internet bubble, the software revolution, and now crypto and AI, these waves are idiosyncratic and distinct in their own ways. We can keep saying that we’re in the next internet software revolution, and that’s a helpful way to think about it, but that same playbook will never work perfectly again.

The founders that compete well in each wave also have different skill sets and personalities. Within those waves, there are different types of businesses that are born, which require different lines of thinking. Being different is your asset, not a liability. There are things we can learn from each of these instances, but no time is ever the exact same. Possibly, the only constant is to be different, unique, individual, and rather than shape to the crowd, shape to the problems you face.

Dig Deeper

So, when you say you want to get rich, what does that mean? Does having money allow you a lifestyle that you can live, or people you can associate with? Does it buy you your time? If you view money to buy you time, what would you do with that time? Why not just do that right now?

When people say they want to be an entrepreneur, what they are often saying is “I want to be rich and famous,” because there is a certain image of an entrepreneur that we have as a society.

I’m trying to just touch grass more, looking internally rather than externally.

After I graduated high school, even though I was interested in science, I studied finance, because I wanted to start businesses and make money, only to find out that you can start many different types of businesses and that finance isn’t really a necessary skill to start businesses. It’s just a skill to understand the flow of capital—“business” is a broadly stroked blanket abstraction that hides the fact that each business is unique. You can start an infinite number of businesses with different business models. I looked outward. Now I ended up in biotech anyways and boy would that science degree have been helpful…

I then matriculated into investment banking because at the University of Illinois, it seemed like the only available class of people who went on to be successful. A problem of availability bias that I then pegged to my understanding of the world. Once there, I realized that my personality is not one that fits in well. I was trying to fit a square box in a round hole.

How to Approach Problems?

What became apparent in my first real job, which is both a strength and a weakness is I don’t like convention—I like the big picture. Banking was extremely process oriented and rooted in convention, in addition to requiring granular attention to detail for things that I often that were de minimus. It’s interesting, if you ask someone who works in banking at another company, their day-to-day always almost completely reflects yours—managing trackers, making profiles, etc. The industry is mature and quite set in its ways. There was a lot that I learned from it, but I have always been someone who tries to fully understand every angle, possibility, the essence of a problem and think through the best way forward, then executing. The thing that constantly hurt me was having certain boxes I needed to check without anyone really questioning or understanding why we were checking them, and then having to execute quickly constantly, without having much time to question the process. I never liked school for this reason, and it felt like my first job was to do homework professionally.

I have been increasingly changing my view of the world to one of nuance. Maybe partly due to growing up with multiple choice questions, I had begun to believe the world had answers and they were determinate. The earth is a giant ball of confusion, and it is an infinitely multi-dimensional gradient.

This is of course a super hard way to conceptualize problems, but that’s exactly how they all work. Problems are faced with multiple input variables, multiple output variables, entropy, irrationality, non-linearity, etc. Now, instead of picking a heuristic, I try to identify multiple opposing modes of thought and wrestle on the gradient between them. I try to question convention. Don’t listen to convention. Color outside the lines. They’re often set arbitrarily.

For instance, the heuristic I previously mentioned has two takes: (1) the customer is always right and (2) if Ford asked a customer for what they wanted they would have said faster horses.

(1) Some say the idea doesn’t matter and startups are about execution, but (2) of course the idea matters because there are constraints and different sized pockets of value that may never work.

We shouldn’t defer to optimism or pessimism; we should be both a hardcore optimist and pessimist and constantly wrestle between the two. We shouldn’t be an insider or an outsider, both have advantages at different times. We shouldn’t be either creative or purely executive, we should wander between the two, there are times for both.

There are many permutations of the above. We must recognize them, like different lines of play in chess, and see as far out as reasonable possible. There are heuristics that help us approach uncertainty, but we shouldn’t rely on them. I hope to live my life not choosing binary modes of action, but by seeing n variables of the spectrum and strategically straddling the line, finding the global point of maximization.

I don’t actually know where I was going with this or if I characterized it well.