2021-07-07

Life Advice: Become a Billionaire

In a certain view, billionaires are not merely wealthy, they are nearly god-like in their influence. As the New York Times op-ed Abolish Billionaires reads:

Billionaires should not exist — at least not in their present numbers, with their current globe-swallowing power.

One practical upshot of this view is that we ought to increase the marginal tax rate, break up tech monopolies, sharpen the pitchforks and so forth.

Yet an equally valid interpretation is this: if you truly see billionaires as all-powerful oligarchs who exert enormous control over world affairs, you should try very hard to become one of them.

How should you go about it? Conveniently, the NYT provides helpful–if Straussian–advice:

A few superstar corporations, many in tech, account for the bulk of American corporate profits… Artificial intelligence is creating prosperous new industries that don’t employ very many workers; left unchecked, technology is creating a world where a few billionaires control an unprecedented share of global wealth.

So there you have it. Work in tech, preferably artificial intelligence, and you’re well on your way to control an “unprecedented share of global wealth”. From there, the world is your sandbox.

At this point, a host of objections spring forth. I’ll eagerly greet them head on.

Is becoming a billionaire even worth it? Doesn’t wealth stop contributing to happiness after a fairly low threshold?

In a variety of landmark empirical results, happiness is shown to correlate with log(income). So the scaling is poor, but that’s very different from plateauing completely.

Khaneman and Deaton 2010 find that some measures of wellbeing plateau entirely, but Cantril’s Ladder, a measure of life satisfaction, does not.

Even more optimistically, a recent study by Killingsworth finds that both satisfaction and well-being continue to improve well past $75,000. [1]

This seems to bear out across countries as well. Per Stevenson and Wolfers (2008) via Dan Luu:

Note that all of these charts use log scales for income, so there are diminishing returns.

Okay, but even if it doesn’t plateau entirely, log scaling is really poor. What’s the point?

Log scaling is really poor, but as the NYT reminds us, we’re talking about really extreme levels of wealth here. Sure, you only gain a few more points of happiness between an income $75,000 and $160,000, but Jeff Bezos is sitting comfortably at a net worth of $211,000,000,000. Our intuitions just don’t apply very well here.

It’s hard to know how happiness scales at this extraordinary level of wealth, but we can at least make a rough estimate. The Khaneman/Deaton chart shows happiness increasing around 0.45 points (on a 10 point scale) each time income doubles. Naively extrapolating, we get that a $1,000,000,000 income would be 13 more doublings, putting you at around 13.35. Again, it’s a 10 point scale, so that’s incoherent, but the point is you still stand to gain a substantial amount of wellbeing.

Fine, maybe there are real benefits, but what about the cost?

If you’re living below the poverty line, there’s plenty of low hanging fruit you can pick to increase your happiness. Find stable shelter, consume enough calories, avoid illness, etc.

But what if you’re already an upper-middle class yuppie? Say your income is already at $160,000. How much are you sacrificing by striving for billionaire status?

Again, we’ll focus on the NYT’s suggestion that the financially ambitious pursue AI-driven tech companies. In that case, you might get paid below market for a few years while your startup gets off the ground, but the degree of financial risk is really not that great. Let’s say your income drops down to $80,000. That’s a halving, which loses you 0.45 points, but it’s only a momentary occurrence.

Though the data only reports effects on income, we can expect there to be a substantial contribution from stored wealth as well. So if you’ve been making $160,000 for a few years and have some savings, going a year without income while you pitch VCs doesn’t actually drop your quality of life by that much. The golden handcuffs were inside of you the whole time.

Even if the cost is much lower than the potential benefit, the odds are really bad. Isn’t it exceptionally difficult to become a billionaire? Commensurate, or even over-commensurate with the rewards? Given that capitalism functions as a finely tuned engine precisely to push people into the creation of market value, is getting another ideological shove ever needed or justifiable?

Again, I think this gets the Marxist critique precisely wrong. It’s not that capitalism—taken broadly as a set of socioeconomic and political devices—pushes people to be as wealthy as possible; it’s that it pushes people to become laborers renting their time to generate profits for others.

Or to continue abusing the Marxist jargon: We live in an unprecedented time where more people than ever have access to the means of production. What does it really take to start an AI startup? A laptop, free wifi, access to some open coursewhere and some AWS credits? That’s still some barrier to entry, but it’s easier than owning a factory or being lucky enough to inherit generational wealth.

In practice, we can easily generate a reasonable lower bound for the probability. Just take the number of billionaires and divide by a count of the total human population. You end up with a small number, but one considerably larger than 0.

But that’s unreasonably pessimistic. We can do much better. Surveying the top Y Combinator companies, I find that around the top 50 are valued at over $1,000,000,000. They won’t all exit successfully, and the founders won’t all own enough equity to emerge with tres commas to their net worth, but this already gets us to a much more practical and optimistic heuristic to life:

Try very hard to get into YC
Conditional on acceptance, try very hard to become a billionaire

Y Combinator has funded around 2000 companies ever, so at rough estimate, your odds are 1 in 40. Still low, but not unreasonably so.

But still, we can do better. Remember that many of the Y Combinator companies were very recently funded, and since batch size has increased over time, the total distribution skews young. Instead, let’s look only at companies funded before 2010. Of the top 50 companies, only 4 were founded after 2017. According to the YC Database, there were around 1400 companies founded before 2017. So that gets our odds up to 1 in 30.

You could object that there were other startup accelerators, and we wouldn’t have known at the time that YC was the right one to join. Or more broadly, that there were other viable career paths to become a billionaire, and startup founding was not as obviously among the surest paths to extreme wealth.

That’s all fair, and you should accordingly adjust the odds downwards, but even diluted by a factor of 10, the expected value looks pretty good.

What expected value? Is it even that good to become a billionaire? Maybe you get a few more points of “life satisfaction” or whatever, but it’s still a steep cliff.

Again, it’s true that wealth generates diminishing returns to happiness, but that’s not the whole story. Notably, wealth generates exponential returns to itself! As the NYT helpfully explains, wealth “serves primarily to perpetuate ever-greater wealth”. The upshot is, despite what you’ve seen for a single moment in time, it’s not at all clear what wealth-happiness scaling looks like in the long run.

Per the NYT as well, consider that “tech instills a winner-take-all dynamic across much of the economy”. That means power law returns, which convey exponential returns from startup rank to wealth:

So the relevant function is not really log(wealth). It’s more like log(wealth(wealth(startup rank)), or log(e^x^y). Does that end up being log scale? Linear? Exponential? Without knowing the specific parameters, it’s impossible to tell. Worst case we’re back to the Khaneman/Deaton case of diminishing returns, but it’s entirely possible trying harder actually generates exponential returns to happiness.

Fine fine fine. I’m sold that in some abstract theoretical sense the math works out. But you’re distracting from the much more salient reality that this is all horrible. People shouldn’t be selfishly exploiting economic inequality for personal benefit, they should be working to end it! If utility really is log(wealth), we can massively increase aggregate utility simply through distribution.

Actually, if you care about helping others, the case for becoming a billionaire is dramatically stronger. An egoist is stuck with shitty log returns, but a truly empathetic person can always give to the poor and thus models (aggregate) utility as a linear function of wealth. Barring really extreme cases, there are no diminishing returns.

Or taking the GiveWell analysis literally [2], a billion dollars could save 200,000 lives! It’s very hard to argue that your time could be better spent on any other cause.

Even that is somewhat unimaginative. You could instead try to start a Charter City that pulls a billion people out of poverty, or fund an anti-aging revolution, or give all your money to New Science and fundamentally change the way science is conducted. Whatever cliches you’ve heard about how “power corrupts”, there’s literally no actual reason you cannot do these things.

That’s fine in theory, but in practice you can only become a billionaire by exploiting others. As Anand Giridharadas put it, “the winners of our age must be challenged to do more good. But never, ever tell them to do less harm… The Aspen Consensus holds that capitalism’s rough edges must be sanded and its surplus fruit shared, but the underlying system must never be questioned.”

In a Marxist sense, this is literally and inescapably true. Profits are the result of exploitation, pure and simple.

Still, it’s worth asking who gets exploited. For example, it would obviously not be ethical to run a company based on slave labor, even if you end up donating the proceeds back to the slaves. In the most generous interpretation, this is at best morally neutral.

But what if you’re running a company that “exploits” wealthy tech workers, and the donate the proceeds to stop human trafficking? In some sense you are still guilty of capitalist exploitation, but it’s not a sense that matters.

Practically speaking, there are plenty of tech companies that de facto, albeit through several layers of abstraction, do exploit very poor people. They might be behind an API, but the poor are still exploited for mining your rare earth minerals, categorizing gore for your content filter, and getting disproportionately exposed to the impacts of climate change from your energy consumption.

My view is simply that slavery, including severely underpaid work and various forms of indentured servitude, is categorically wrong and you should not build a business that relies on it.

But even accepting this imperative, and accepting the Marxist ideological framework, it is not true that becoming a billionaire necessitates generating more suffering than you have the capacity to eradicate.

Hold on, that’s all just a shitty excuse. I can already picture the tech entrepreneur who claims they’re only becoming rich to give to the poor, but ends up donating their wealth at a meager trickle.

Look, that’s fine, but you’re not talking about life advice anymore, you’re just debating status.

I don’t care if we worship Elon musk or build statues to him on Mars. I’m making a specific claim about what you ought to do with your life.

If you’re more worried about optics and guilt by association than you are with reasoning about and acting on what’s good, you’re completely missing the point.

Okay, but it’s still not a coincidence that so many billionaires are shitty people right? Either you have to be shitty to get there, or getting there makes you shitty. Either way it’s a bad outcome and claiming higher motives is dishonest.

My view is that most people, in general, are pretty shitty, so it’s not really surprising that this holds true in populations that haven’t specifically been selected for empathy and altruism. That’s why I started with and devoted the bulk of this post to the self-interested case. This whole secondary discussion is predicted precisely on the condition that you want to know how to help other people.

Unless you seriously think there’s something fundamentally incompatible with being a decent person and becoming a billionaire (remembering that the future will be different than the past), this is all just a really stupid version of Newcomb’s paradox. The argument boils down to the thought experiment:

Box A is clear, and always contains $100,000
Box B is opaque. You’ve been told it contains $1,000,000,000.

Your choice is between taking Box A once a year, or spending, I don’t know, 6 months, to work on a startup, apply to Y Combinator and have a shot at Box B.

And your counterargument, with $1,000,000,000 sitting right in front of you, is that you have a vague sense based on anecdotes and selection bias that maybe taking the money is bad.

Come on. Obviously you should take the box. [3]

Thus far, I’ve mostly considered historical factors, assuming the future looks like the present. But the future could be very different! How does this strategy perform in different scenarios? I’d like to suggest it’s at least a reasonable hedge:

If income inequality continues to increase, it’s even more important to make sure you’re part of the oligarchical class
On the other hand, if inequality decreases and we move towards a socialist or UBI world, the opportunity cost goes down as the social safety net improves

That’s just the domestic version. A similar argument applies internationally as well:

If some countries remain very poor, you’ll always be able to retire with relatively modest savings and still have a very high quality of life
If all currently poor countries become rich, such an enormous amount of global wealth will be generated that it’s an even better bet to become a oligarch

I’m sort of kidding. It’s not that you literally have to go out and exert political influence. It’s just that in the latter scenario, there are such good returns on capital and so much wealth to go around that you ought to make sure you’re taking advantage of it.

This is a contentious topic prone to misinterpretation and heightened passions. So let me be clear:

The recommendation to become a billionaire is both serious and literal.
I understand that being a founder carriers a high risk of being failure, but it carries a low risk of actually ruining your life and personal finances
I understand that the odds of becoming a billionaire are low, but it doesn’t matter if you only consider the conditional probabilities. What’s the cost/benefit of taking 6 months off your day job to work on a startup? Conditional on being successful there, what’s the cost/benefit of trying very hard to seek out venture capital? Given that you’ve raised money, what’s the cost/benefit of trying to make a billion dollars? The bet sounds insane to begin with, but at each step you’re taking on a very reasonable level of risk.
I don’t endorse all facets of the current economic system, but unless you are actually involved in starting a socialist revolution, I recommend trying to do your best with the situation we have.
I’m not making any claim about the correct status of existing billionaires. But the closer your view is to “billionaires are god-like oligarchs who control everything”, the more seriously you should consider joining their class.

All things considered, the burden of proof is in your court. And I’d like to suggest that it will take a lot to make any other career choice even moderately competitive.

Appendix A: Arrogant Base Rates

Alexey Guzey will yell at me if I don’t acknowledge that people who actually succeed in becoming billionaires probably do not think in terms of base rates and conditional goods.

As a post-truth Aristotle might have said: “it is the mark of an educated mind to be able to entertain a thought without letting it become an infohazard.”

Alternatively, the argument laid out here demonstrates precisely that there’s nothing inherently conservative or modest about reasoning from base rates, and so the accusation falls flat on its face.

As for the conditional goods, I’ll add: no one should start a company just because they want to get rich, but it doesn’t hurt to be well positioned in the first place.

Appendix B: Aptitude and Asymmetric Uncertainty

In So Good They Can’t Ignore You, Cal Newport describes Steve Jobs’s humble beginnings:

At one point, he left his job at Atari for several months to make a mendicants’ spiritual journey through India, and on returning home he began to train seriously at the nearby Los Altos Zen Center.

…these are hardly the actions of someone passionate about technology and entrepreneurship, yet this was less than a year before Jobs started Apple Computer. In other words, in the months leading up to the start of his visionary company, Steve Jobs was something of a conflicted young man, seeking spiritual enlightenment and dabbling in electronics only when it promised to earn him quick cash.

Or recall from my review of The Making of Prince of Persia:

In the course of making Prince of Persia, Mechner:

Takes 6 months off to write a screenplay.

Drives out to Skywalker Ranch, meets George Lucas, fails to get his script acquired.

Applies to NYU film school, gets rejected.

In another series of anecdotes, Alexey Guzey illustrates that visionaries are not natural-born leaders:

Musk realized that he could have handled some of the situations with employees better. “I had never really run a team of any sort before,” Musk said. “I’d never been a sports captain or a captain of anything or managed a single person.”

That’s all to say: if your objection to all of the above is that you are not predisposed to becoming a visionary billionaire tech CEO, it doesn’t matter. Again, the billion dollars is sitting on the table and you’re choosing not to take it on the basis of a vague intuition.

Consider as well that the future is likely to undergo continual shifts in what constitutes the “right aptitude”. In Zero-to-One, Peter Thiel describes the PayPal Mafia’s shared childhood interest in building bombs. That probably would not have made for good founder material 20 or 50 years earlier, but in the late 90s it was highly profitable to be precocious, uninterested in personal safety, and willing to disregard legal concerns.

Yesterday’s leaders were charismatic strongmen who could dominate a room. Tomorrow’s leaders may be introverts better at writing than they are at speaking. Yesterday’s crypto billionaires were libertarians who flouted fiat and hated the FED. Tomorrow’s may be whoever has the closest ties to the SEC, or whoever best understands the shifting regulatory environment.

I’m not saying any of this is true, just that it’s possible.

You might object that uncertainty is deleterious to a bet. To a risk averse actor, higher variance is equivalent to lower expected value. But when the base rate of success is low to begin with, uncertainty provides an asymmetric advantage. Say your odds of becoming a billionaire are at 1% with SD of 0.5%. As variance increases, odds can’t go below 0, so the bet actually improves.

For these sorts of low probability high return scenarios, the question isn’t “is it likely”, but rather “is it at all possible?” The kind of uncertainty I’ve laid out here makes it less reasonable to rule out the possibility that you could become a billionaire.

Stop making excuses and take the box already.

Footnotes

[1] Michael Plant and Julian Hazell cast some doubt on the Killingsworth result, but the cause of the data discrepancy is unclear.

[2] GiveWell cautions against taking their estimates too literally as absolute measures, but it’s still an okay estimate.

[3] For what it’s worth, some version of this logic applies to the original problem as well.

Coda
https://www.youtube.com/watch?v=lI5w2QwdYik