The key missing piece to Thiel's analysis of the design vs. statistics dichotomy is the fact that it's best to have both: a strong vision combined with a lot of low-risk experiments to shape your understanding of the best path to achieving that vision.
From what I understand, Apple does exactly this: they do a lot of prototyping before settling on a product design, even though the broad strokes of what they're trying to build are clear from the get-go. I expect the future of product development to keep moving in this direction as prototyping becomes progressively less expensive.
Speculating on history a bit: the transition from determinate to indeterminate beliefs happened because people figured out that a lot of strong proclamations about the future turned out to be very wrong. So the pendulum swung the other way for a while, and now we're finally starting to settle on an equilibrium between the two.
The analogy I like best is "Work to increase your surface area for luck." When dissolving a substance, if you pulverize it, you increase the surface area, which allows for greater interaction with the surrounding water molecules, and thus enables a faster dissolving rate.
In the same way, the more interactions one has, the more opportunities for fortune to strike. The result may look like "luck" from the outside, but in reality it was a confluence of both stochastic elements and fortuitous prior decision making.
Isn't that an example of a "diversification" strategy, similar to the "join many clubs/activities in high school" approach? The fact that it's a strategy (and therefore involves prior decision making - specifically the decision to follow the strategy) doesn't make it a "determinate" strategy.
I believe the lecturer (Thiel) is advocating for determinate strategies - which, practically speaking, reduce the surface area for luck since time is a finite resource. My interpretation is that he is advocating for "moon shot" approaches (examples given: Tesla, Space-X, AirBNB) where the plan/vision is laid out from the beginning.
This contrasts with the iterative "lean startup" approach whereby an initial product is launched followed by an iterative customer feedback loop. The evokes raises an interesting question: is the "lean startup" methodology an artifact of "Indeterminate Optimism"?
There is an intimate link between determinism/indeterminism, optimism/pessimism and local/global maxima, ectropy/entropy.
Optimism is fundamentally a belief that we are not currently at a local/global maxima. Determinism is a belief that the path to local/global maxima is knowable through human intelligence.
This is why A/B testing, as pointed out in the article, is fundamentally incapable of leading to anything other than a local maxima. It is optimistic, but without determinism, it will never sacrifice traversal through local minima to achieve a greater maxima.
This is personally why I have such a reaction of disgust to "ideas are worthless" and "nobody knows anything until you build it," because I am adamantly optimistic deterministic at heart and do believe in design, increasing ectropy, and vision. A side effect of this is that in our pessimistic indeterminate societal POV you are characterized with delusions of grandeur until you prove out your vision. Even then though, you may be discounted as lucky.
I think this idea that the US has moved to the optimistic indeterminacy quadrant coincides nicely with the rise of the internet and cable news where there is a huge bias towards focusing on perceived bad news, no matter how important any of it actually is. There isn't money to be made in NOT scaring people, so CNN/Fox News/MSNBC/Huffington Post/Politicians/Pretty much anybody trying to make a name for themselves sell fear and fear sells well.
Twenty years ago most things that dominate the 24-hour news channels wouldn't be conversation worthy but now people are addicted to bad news and have come to believe at any given time that things are worse than they've ever been.
“I’m a great believer in luck, and I find the harder I work the more I have of it.”
I just wanted to point out for those who were as inspired by this "Thomas Jefferson" quote as I was, that it was not actually TJ who said this, but a man named Coleman Cox [1].
The ultimate source of luck as a factor seems to me to be your origin:
- Were you randomly selected by the genetic lottery to be smart, attractive, healthy, etc.? Other traits might also matter, such as charisma and alpha personality characteristics. I've often noted that individuals with a slightly manic personality are actually very well adapted for high-speed industries. I've experimented with using things like modafinil to push myself a little further into this state with extremely positive results on my career. (My natural state is more "beta" and contemplative than "alpha" and high-energy/assertive.)
- Were you born into high or low socioeconomic status?
- Were you born in a rich and relatively free country or a poor or corrupt/despotic one?
- When were you born? Not only does this matter on huge scales (e.g. it's better to be born today than in 1500), but it also matters in a cyclic sense. Do your life span periods of highest work productivity overlap with an economic high note or with a recession?
Finally there is the obvious factor of accidents. If I get hit by a car and get brain damage, obviously I'm less likely to be successful.
These lead to a few questions for me:
(1) To what extent is it statistically in the best interest of most people then to support some form of social safety net?
(2) Is it in my statistical best interest to support some level of social safety net given that by doing so I might gain additional value due to higher rates of cooperation? (Assuming that I am in the top, say, 25% of the ability curve and might not need a safety net as much as many.) A safety net can also be viewed as insurance. What if I am injured, get cerebral meningitis and get brain damage, etc.?
(3) To what extent can these be hacked? Look at my first point about genetic luck of the draw, and how I've managed to kind of hack this a bit with chemistry and make myself more of an alpha personality when it's advantageous to be one.
Personally I think that (1) and (2) are nonzero, which is why I do support some level of social safety net as opposed to naked capitalism. I also think that (3) is a definite 'yes' based on personal experience. Makes me wonder what would happen if you gave the poor nootropics. (Before the PC police whack me for that last comment, consider that I don't just mean IQ. I also mean assertiveness, energy level, ability to focus, and so forth.)
At the end of the day I disagree with Peter Thiel. You are a lottery ticket. You are just a sentient, intelligent lottery ticket. You get to choose when you are played, as well as use whatever hacks you can get your hands on to change the numbers beneath the scratch pads. You also belong to a community of other lotto tickets and together are capable of conspiring against the rules of the game.
Boy, this is what Thiel teaches in his classes? The only useful, actionable insight was in the last paragraph and then it ended. Lots of abstract, thoughtful stuff, but nothing that could make me more successful.
Luck is code for "variables I don't see." The problem with defining a luck/skill separation is that it's subjective which of those factors should be a person's responsibility, and which are impossible to anticipate. So we'll never resolve that one. It's a fairly subjective call. They're two sides of the same coin. It's a "skill" variable if managing that feature is part of the job, and "luck" if not so (perhaps that feature is unmanageable.)
I take a different and much more negative approach. Once a company or ecosystem (such as VC-funded technology, increasingly shifted toward marketing-driven light tech) ceases to add value, it reaches a state in which neither luck nor skill is the main variable, but pre-existing social class and status. Indeed, one might argue that the appearance that it's either luck or skill is an intentional deception.
Take college admissions. There's some merit (it's non-trivial to crack 1500+/1600 and get good grades in high school) in the equation, but there's a huge extracurricular component (cough racism, classism cough) that is impossible (for a middle-class parent or high school student) to predict or navigate. It looks "random" to those people. That's how it's supposed to look. Some cluelessly conclude, "It wasn't in the cards." Others, equally cluelessly (and more pathetically) conclude, "I guess Johnny isn't cut out to be a leader" (like anyone can fucking tell at 17). It's not luck or skill at all. It's deeply classist which extracurriculars are valued and which are not, and it's made to look random to one set of people, and skill-driven to another, because the reality is socially unacceptable.
The same applies to Silicon Valley. It has ceased generating real value. It's a Disneyfied farce of entrepreneurship designed to make rich twerps like Evan Spiegel and Lucas Duplan appear to have earned it. Some people believe that this is based on superior skill. Others are misled into believing it's luck. It's just random "serendipity" that some people get and some people don't. Not so. It's corruption. It looks like something else, because that's the ruse. The luck/serendipity aspect is put there to muddy the water so people can't see what is really going on.
Possible alternative explanations to what you're suggesting:
1. Growing up in a higher socioeconomic status environment does genuinely prepare you better for academic and career success, so the correlation in university admissions is rational given that universities can only change a person to a limited degree.
2. 90% of everything is crap, and startups are no exception. But that doesn't mean they all are.
> (cough racism, classism cough) that is impossible (for a middle-class parent or high school student) to predict or navigate
Ron Untz has concluded there is heavy admissions discrimination against whites from rural areas and asians. Interestingly, Jews are quite over-represented.
From what I understand, Apple does exactly this: they do a lot of prototyping before settling on a product design, even though the broad strokes of what they're trying to build are clear from the get-go. I expect the future of product development to keep moving in this direction as prototyping becomes progressively less expensive.
Speculating on history a bit: the transition from determinate to indeterminate beliefs happened because people figured out that a lot of strong proclamations about the future turned out to be very wrong. So the pendulum swung the other way for a while, and now we're finally starting to settle on an equilibrium between the two.