How Elon Musk, Charlie Munger & Aristotle Had You Fooled
First Principles Thinking, Heuristics, Intuition
First Principles Thinking, Heuristics, Intuition
Elon Musk & Charlie Munger are wrong about first-principles thinking.
Reasoning from first principles is not the most optimal framework for thinking.
And if you take it too far, dig for too long in places where good, long-standing heuristics already exist, it may lead you to a dark place.
Infinite Jest
We live in a vast universe, according to science.
While few of us have personally ventured into space to witness it, it would be folly to argue that this statement was false.
And yet, we might.
Why don’t we break everything down to the core, to understand the small pieces?
How far should we go?
What is empirically known among those pieces, and what can be challenged?
What is empiricism?
Evidence that justifies beliefs?
Evidence that confirms or disconfirms scientific hypothesis?
Who’s vantage point for those beliefs are we taking here — our individual view, or are we taking someone else’s as a proxy?
As you can imagine, we can go down this rabbit hole for a very long time. The tree of our inquiry will grow and grow, giving us a sense of purpose for a while. This sort of inquiry can feed a philosopher for years. It requires rigour and will.
Should everyone stare into this abyss?
Karen Green, the wife of the writer David Foster Wallace, found her husband dead one day. He had hung himself.
Having struggled with depression for years, the author tragically succumbed to its pressures at a young age of 46.
Whatever Foster Wallace wrote was filled with the sense of getting to the core of things, of searching for the essence of a phenomenon or experience.
There may not be a direct correlation between first-principles thinking and the writer’s suicide. But in the book “All Things Shining: Reading the Western Classics to Find Meaning in a Secular Age” the authors Hubert Dreyfus and Sean Dorrance Kelly open with a critique of this premise in Wallace’s writing. It’s this idea that “through the strength of human determination, we can strike meaning and find certainty in the universe on our own.”
This sort of maniacal search for the root in everything can be so daunting that in the end, you’d be so overwhelmed and lonely and empty, that you would possibly find it hard to go on.
A team I worked with took on a challenge.
They worked it all down to the first principles. What was a product? Who was a persona?
Such a stripped down approach pared down to the core need was overkill.
This reductionism in the problem space has lead to endless navel-gazing & pondering of life’s deep questions. How did it happen?
Right Method, Wrong Principles
Daniel Kahneman’s book “Thinking Fast, and Slow” is on every intellectual’s book shelf.
The System 1 & System 2 model from the book relates to this idea that we generally think on a spectrum from thinking in whole blocks & patterns, to reasoning from first principles.
A baby has to learn everything from scratch, and an adult can lean on progressively more complex shortcuts as they go through life. Some of those shortcuts come to us in the form of collective wisdom, and some of these shortcuts we figure out ourselves.
But what shortcuts will an abused baby learn?
This is a Medium article that you’re probably skimming through semi-seriously, so no harm here.
But what if you were to base your whole life philosophy & on the wrong premise?
Cedric Chin, the author of “How First Principles Thinking Fails”, puts it this way:
If you pick the wrong set of base principles — even if they’re all true — you are likely to end up with the wrong conclusion at the end of your thinking. In other words, the only real test you have is against reality.
When applying first-principles thinking, we’re relying on the processor (us) to do the hard rigorous work of digging up the first principles.
But who’s to say that it’s only a matter of time spent doing this? Who’s to say that the answers are already there, in pure form?
And what is the cost of delay of doing this digging, if we still can’t find the perfect answer? Is the promise of perfection worth it?
It’s ironic because as I wrote this, I realised that I myself could take this argument further down, to the essence of what it meant to be the “first principle.” I was staring into the abyss.
(But the article is here now.)
I recall a story a user researcher told me.
He was questioning the value of his profession (by the way, a very life-threatening thought for many).
He said that after a lot of good, deep, quality research, their conclusions still didn’t lead to a successful product. How could it be? Was it all a lie? Where was the promised land?
Intuition Takes Me Everywhere
In my line of work, we’re really careful not to “solution-ise” in a given situation.
Yet, the brain offers us solutions, hypotheses, positions almost immediately. The quality of those will depend on this little non-scientific thing some would call “sense”, “intuition”, maybe even “experience”.
In my earlier story about a team who went too far back, the senior people saw it & had an intuition for what needed to be done. But like well-meaning leaders, they held back and gave the team space to find the right question.
(We can argue that they should have been clearer in their intent, but that’s a story for another day.)
It was hard to argue for intuition and against principled reasoning.
As impostor syndrome takes over, we start to doubt the things in our gut that we can’t express how we know. It’s not data-driven. Haven’t we read multiple articles about bad products conceived by an exec that was going on “sense”, “gut”? No one wants to come off as a micromanager, or worse, a rookie “solution-iser”?
But could we be taking it too far and mistrusting our deeply embedded truth?
In “How Doing What I Feel Like Made Me a Better Product Leader”, Noa Ganot writes:
Intuition is not an opinion. […] It is something that you know without conscious reasoning. In other words, intuition is knowledge, not a feeling, but we don’t know how we learned it.
In Gary Klein’s “Sources of Power”, a brilliant book on the decision making of professional firefighters, policemen & other professions who deal with tough situations all the time, he puts it simply:
Experience counts. The different sources of power are ways of drawing on experience. In natural settings, perceptual learning takes many cases to develop.
How Do You Decide?
It’s the battle between “if it’s classical, status quo — doesn’t mean it’s true” vs. “if it’s classical, it endured over time and is likely to be true, and endure for longer”.
It’s the lack of agreement around what constitutes knowledge that is objectively known vs. subjectively known, multiplied by the longevity of that knowledge. What’s common for all vs. common for some.
We’re debating the general model of the reality. We’re trying to balance model fitness with model bias.
Overfit, and your model is so precise that it works amazingly well in just 1 particular scenario.
Underfit, and your model is generally useless, as it is too broad.
Intuition may hold just as many “first-principles” nuggets. It is not shared knowledge.
Your intuition is not my intuition.
There are probably areas where they intersect, but it stands in stark contrast to, say, a law in physics that has the qualities of being shared + proven empirically or analytically, multiplied by all the times it’s been applied in practicality (the reality test).
Intuition is often something learned the hard way, too, “via negativa” as NN Taleb would put it.
It’s the entrails of the comet of loss; post-hoc analysis driven by regret that settles as a lesson in our memory.
This lesson is now sitting in the back of our brain, waiting to be tested. It’s not knowledge that is visible to all, though some in similar situations may share it. It’s not tested exactly, which is why sometimes we don’t trust our intuition, just to find out — “via negativa” — if we should have trusted it or not.
So this argument will always resolve based on your personal appreciation of the reality.
It’s my way of saying, “it depends”.