Big Tech at the End of History

Byrne Hobart -

Welcome back to The Diff. Here are the subscribers-only posts you missed this week:

Subscribe now to receive paywalled issues, read the archives, and join today’s subscriber call at 2pm Eastern, where we’ll be discussing The End of History and more.

Subscribe now

This is the once-a-week free edition of The Diff, the newsletter about inflections in finance and technology. The free edition goes out to 14,867 subscribers, up 250 week-over-week.

In this issue:

  • Big Tech at the End of History
  • Belt & Road & Vial
  • CCP and SMBs
  • Hire Fast, Transfer Fast
  • Life Imitates Science Fiction
  • Koanic Saas Secrets
  • Interactive Brokers' Big Option Trade

Big Tech at the End of History

A few weeks ago, I gave a talk to a group of employees at a large tech company.[1] Topic: parallels between how tech companies grow and how countries go from poor to rich.

The two processes have some interesting similarities:

  • Most countries achieve sustained growth in GDP per capita through export-driven growth. As their exports rise, productivity per worker increases—they buy better capital equipment (and eventually make their own), they get better at managing complex supply chains, and they expand into adjacent industries where they can add more value. This happens with companies as well: early growth is often associated with a rising headcount and rising revenue per employee, as the company simultaneously scales its business and gets better at efficient scaling.
  • Countries accelerate their growth through “financial repression,” encouraging workers to consume less and save more, and recycling those savings into ever more investment. Sometimes, the policies behind this are fairly benign (a weak safety net and few worker protections will force workers to save much of their income to self-insure against unemployment), although some of them can be quite brutal (Stalin starved Russians to move resources away from agriculture and towards heavy industry). Startups also engage in a form of financial repression, by weighting pay towards equity instead of cash.
  • Rapid growth is often driven by top-down industrial policy. This approach is hit-or-miss; state-directed investment in Brazil and Nigeria has been disastrous (it took Nigeria 40 years to finish a steel plant). But in Japan, South Korea, Taiwan, and China, it produced rapid growth for decades. Startups, too, tend to be run in a fairly top-down way at first. Interestingly, growth in both industrializing countries and burgeoning tech companies often benefits from top-down policy and bottoms-up implementation; the South Korean and Japanese governments lavishly subsidized its industrial sector, but required companies to compete for access to those subsidies, generally by tying them to exports; Google gave employees “20% time” to launch new products, and Facebook’s “virtual gifts” product was built during a hackathon.
  • As countries' economies mature, productivity growth slows. There are some reasons for this that are tied to demographics—richer countries have better healthcare, and better healthcare means longer lifespans at retirement. And some are due to economic factors: as countries grow, more of their consumption takes the form of services rather than products, and it’s harder to drive productivity growth in services. Better textbooks, digital whiteboards, and an iPad-in-every-hand do not significantly change the number of teachers required for a first-grade classroom. For companies, too, growth slows and headcounts grow as they’re forced to engage in more sales, marketing, traditional recruiting, and customer service to support further growth. Small direct-to-consumer brands will happily spend five figures per month without talking to a human being. Procter & Gamble and McDonald’s are pickier, but they write some big checks.
  • Countries and companies also get more political as they mature. Taiwan and South Korea were military dictatorships during the early stages of their democracy. Japan was, technically, a democracy—but the function of its legislator was a) to propose bills for local spending, which invariably got shot down, and b) to rubber-stamp plans submitted by MITI.[2] Today, these countries have far more open political systems, and much more freedom of the press. Similarly, early-stage companies are often about to keep a lid on leaks and dissent (if you’re a fresh graduate with student loans, your equity is over 100% of your net worth, which is a good way to make you a team player). But more mature companies have to do things like moderating internal message boards— a decision Google and Facebook made recently.

The original point of that model was to say: this has happened before, just on a slower timescale. Growth starts out scalable and apolitical, but naturally slows over time. In a mature economy, wages are high, growth is low, the obvious high-return investments have all been picked over, and life is comfortable, boring, and more driven by politics than technological advances. So, too, at mature tech companies. Nobody talks about “resting and vesting” at the Series A level. (At least, not at companies that are going to have a decent Series B.)

But that’s not a story with a happy ending, because it doesn’t really have an ending. “At first, progress happens fast at the cost of hard work, uncertainty, and some material deprivation. Eventually, growth slows but most of the major problems go away, and everyone involved is quite well-off. The… end?”

The End of History, etc.

Fortunately, there’s a guidebook to what can happen next. Francis Fukuyama’s The End of History and the Last Man is all about the question of what comes after we’ve solved all of the big problems. Fukuyama’s book came out at an opportune time. Today, books about the broad sweep of recent history either talk about how bad things have gotten, or about why it’s contrarian to think the world is not so bad.[3] It’s very 1990s to write a book that makes the obvious point that everything is going absolutely great, and to ask the contrarian question of what could go wrong.

The End of History and the Last Man was and is a controversial book. Some of Fukuyama’s critics clearly didn’t read the whole thing, and some of them didn’t even make it all the way through the title. The term “The End of History” is pretty easy to interpret: all the big questions have been answered, capitalist democracy is the system everyone eventually adopts, and it’s hard to imagine a successor, so the big events are all in the past. “The Last Man” takes things in a darker direction: it’s a term from Nietzsche, referring to someone with no urge to create, no willingness to take risks, someone who has been spiritually poisoned by peace and comfort. The Last Man has lost all desire for recognition from peers, and is ruled by baser desires.

The book makes a smooth segue from the end of history to the last man, so reading it today is an increasingly enjoyable experience, as Fukuyama transitions from being a delusional fool who seems to have overdosed on episodes of The West Wing to sounding like a prophet. It may be a testament to pre-Internet attention spans that Fukuyama could put the caveats 300 pages after the arguments and expect people to read them.

Fukuyama’s end-of-history argument, even including the caveats, is in retrospect a little weak: he argues that only capitalist democracies can marshal the resources necessary for scientific advances, and that ultimately military threats force every country to converge on the system that works. So his model is: the more capitalist and democratic a country is, the more it can mobilize resources towards new technologies; some of those technologies have military applications; in the long run, everyone who fails to achieve military parity gets conquered, so anyone who doesn’t converge on a working model voluntarily will eventually do so at gunpoint.

This gives Fukuyama’s model of history some direction: science accumulates over time, so if science and capitalist democracy feed on themselves, science pushes history in only one direction. Science does indeed accumulate over time, but imperfectly: the cause and cure for scurvy was repeatedly discovered and forgotten, the US may have forgotten how to produce some nuclear weapon components, we still don’t know for sure what Greek Fire was, etc. And some stories that appear to have a moral about superior technology have something else going on entirely: the Spanish military did indeed have better equipment than the Aztecs, but Spain was outnumbered 100:1, and won by assembling a coalition, not through their own superior firepower. Military coalitions are not exactly a new idea; the earliest work of western literature, The Iliad, partly revolves around managing one.

Some technologies have negative side effects, and different countries can adopt uniquely effective subsets of technology in order to be more effective than more advanced countries. On paper, the US was far more powerful than Vietnam, Iraq, or Afghanistan, but a country that has fancy guns, fancy bombers, napalm, Agent Orange, and television and elections does not necessarily have an enduring advantage against a country that imports the guns but not the TVs and elections.[4]

And some of the evidence Fukuyama cites for a post-history world overfits the data. He notes the decline in Europe of nationalist wars after 1945, but as Tony Judt points out, nationalism in postwar Europe was less important because nation-states' borders largely corresponded to ethnic groups, through a combination of redrawn maps and wholesale forced migration. In Judt’s reading, Europe’s post-1945 peace largely vindicated Neville Chamberlain.

It’s extremely unfair to critique Fukuyama’s mid-90s viewpoint based on what happened since, especially since his book shows escalating concern over the possibility that the End of History does not mean indefinite utopia. For Americans, the End of History came to a spectacular end on 9/11. Bin Laden, among his other undesirable traits, was such a ham-handed literary symbolist he’d make the most pretentious MFA student blush. Crashing planes owned by a company called “United” and “American,” targeting a “World Trade Center”—we get it, the system ultimately contains deep and irreconcilable contradictions, give it a rest.

What 9/11 symbolized was not just the physical risk that a safe and prosperous country could face, but the ideological risk that modern liberal values could be thoroughly rejected, and the technological fruits of those values could be redirected towards mass destruction.

The rise of China, too, represents a serious challenge to the End of History. Prudently, Fukuyama draws attention to precisely this:

[D]espite the apparent absence of a systematic alternatives to liberal democracy at present, some new authoritarian alternatives, perhaps never before seen in history, may assert themselves in the future. These alternatives… will be created by two distinct groups of people: those who for cultural reasons experience persistent economic failure, despite an effort to make economic liberalism work, and those who are inordinately successful at the capitalist game.

Later in the same thought, he even hints at the origins of the 9/11 attackers, noting that fundamentalists backed with oil money can do a lot of harm, and that religious fundamentalism is strongest in modern countries. (The 9/11 hijackers were mostly well-off and well-educated; Mohammed Atta was born in Egypt but radicalized in Hamburg. And the operation was financed by Bin Laden’s family money, which derived from Saudi Arabia’s oil-funded construction boom.)

To Fukuyama, the core of the “last man” problem is thymos, which can be literally translated as “spiritedness” and more figuratively and vividly means the desire for recognition and glory. Thymos describes the urge to excel, even—especially!—at great risk to oneself. Fukuyama’s ultimate source, Hegel, spends a lot of time meditating on the original thymos-inducing practice of forcing another human being to submit to one’s will; modern societies find healthier and less direct ways to harness this desire. Plato divides the soul into reason, desire, and thymos. Reason and desire are enough for a pleasurable existence—you have wants, and you figure out how to meet them. Thymos is necessary to motivate sacrifices for public service and glory.

Fukuyama’s immediate predecessor in the end-of-history business, Kojève, thought the End of History arrived when everyone was perfectly satisfied. Kojève announces that at the end of history, humans will still have desires and reason, but they won’t have any higher desires; we’ll revert to a stage of very advanced, comfortable, happiness-seeking animals. To him, the End of History arrived when Napoleon crushed the Prussian Army at the Battle of Jena in 1806; everything since then has been an extended mopping-up of various remnant factions of humanity still stuck in history. The end result is a long march to universal satisfaction.

But it seems that successful people are perennially dissatisfied. They’ve retained some thymos. Growing up in the upper middle class, attending an elite prep school and then going to Harvard sounds like a guarantee of about as much material comfort as any human being could ever ask for, and yet two of the most valuable companies in the world were founded by people who rejected that life path. Arnold Schwarzenneger managed to be dissatisfied with his life as a monstrously fit millionaire property developer, and decided to become a movie star instead. Then he decided that wasn’t enough, either, and decided to be governor of the world’s fifth-largest economy. There are similar stories about plenty of other immensely successful people. Both Warren Buffett and Andrew Carnegie once memos-to-self in their mid-thirties, summing up their respective net worths and noting that this was more than anyone would need, and that it was enough for a life of quiet contemplation and charitable giving.

Interpreted with maximum charity, both Fukuyama and Kojève, with varying amounts of intellectual humility, describe life at the high point of an unsteady upward trend towards the end of history. Zoom in enough, and a sine wave looks like a straight line; zoom out enough, and a sine wave plus a linear term is a straight line for all useful purposes. History’s sample size is frustratingly small, and the data are noisy, so it’s possible to over-extrapolate a temporary deviation or to underestimate mean-reversion. Fukuyama’s model points to the possibility of exactly the sort of reversion that history promptly supplied.

Big Tech and Thymos

Where does all this leave Big Tech? Each of the largest tech companies has a very America-in-the-90s feel. They’ve conquered the big problems. They still have problems, but they’re relatively small-scale; the risk is a bad quarter, not bankruptcy. And yet, most of the well-paid tech workers I know feel deeply dissatisfied. When they look at how much revenue their work generates, and how valuable and profitable their employers are, they feel underpaid. When they consider the effort required to keep getting promotions, though, they feel overpaid. Life at a large tech company is, in Plato’s model, satisfying reason and desire, and ignoring thymos entirely.

Is there an escape? Of course. Some people quit completely, and retreat to monasticism, or perhaps podcasting. Others quit and form startups. That’s a serious problem for Big Tech, though: in The Republic, when Plato divides his ideal society into classes based on this model of the soul, the thymoeides are the ones tasked with defending the city from outsiders. If everyone at Facebook who could give their all to a company decides to quit and give their all to some other company instead, it’s ultimately defenseless.

Facebook has inspired thymos among its workers in the past. When Google+ launched, Mark Zuckerberg’s speech is described like this:

The contest for users, he told us, would now be direct and zero-sum. Google had launched a competing product; whatever was gained by one side would be lost by the other. It was up to all of us to up our game while the world conducted live tests of Facebook versus Google’s version of Facebook and decided which it liked more… It was the sort of nagging paternal reminder to keep your room clean that Zuck occasionally dished out after Facebook had suffered some embarrassing bug or outage…. Rounding off another beaded string of platitudes, he changed gears and erupted with a burst of rhetoric referencing one of the ancient classics he had studied at Harvard and before. “You know, one of my favorite Roman orators ended every speech with the phrase Carthago delenda est. ‘Carthage must be destroyed.’ For some reason I think of that now.”… Zuckerberg’s tone went from paternal lecture to martial exhortation, the drama mounting with every mention of the threat Google represented. The speech ended to a roar of cheering and applause. Everyone walked out of there ready to invade Poland if need be.

That’s some thymos! Carthage was destroyed. But now, Facebook’s threat isn’t an external enemy that wants to copy their features and take their users; it’s regulation, and internal conflicts over what, exactly, the company should accomplish.

Some companies find ways to maintain their thymos well into maturity. Good design can be a source of thymos: companies that produce nice-looking, well-functioning products have a stronger motivation to excel. You can feel some pride and esprit de corps from wanting your side to win over a different company, but it’s more deeply distressing to think that, if the product fails, everyone will use something ugly and broken instead of what you and your colleagues built. Employees at Stripe and Apple can be motivated by the thought that, if they don’t work hard and innovate, the world will get similar services that are uglier and less elegant.

Amazon seems to have some thymos left, too. When Jeff Bezos talks about having a single-digit share of retail sales, the intended audience is legislators worried about the company’s market dominance. But inside the company, the subtext might be: we can 10x this, and we still won’t be done.

But every success saps a little more thymos over time. High margins are the result of vanquished enemies, but if your company culture is built on repeated victories, what can you do when there’s nobody left to beat? Some people rest and vest, letting the hedonic treadmill ensure that $500k is just as insufficient as $100k was. Others quit.

On a national level, it’s hard to permanently exhaust a country’s thymos because it’s always being renewed. Some people emigrate, but emigrants can come home. A company isn’t home, though; if they lose their most ambitious employees, or select for the wrong kind of ambition, it takes extraordinary effort to turn things around.

Every big company should internally compare itself to the world that Fukuyama describes in The End of History and the Last Man. And then they should take note of the fact that history has emphatically refused to end.

Further reading: This piece owes a lot to The End of History and the Last Man. If you enjoyed that and want some more, Fukuyama’s book is an extension of, and response to, Kojève’s Introduction to the Reading of Hegel. And one interesting response to the end-of-history thesis, a reply to Kojève and others, is Peter Thiel’s “Straussian Moment” chapter in Politics and Apocalypse.

[1] The company in question is large enough that it has policies on exactly how to refer to such things—not “I was invited to speak at X,” but “I was invited to speak by a group of employees at X,” etc. I’d rather not accidentally break the letter or spirit of that agreement—and the phenomenon is not specific to them—so I’ll just note that it’s a large company and you’ve more likely than not used one of their products in the last thirty minutes. And, of course, that the invitation did not represent an endorsement of anything I say.

[2] Japan seems to be a democracy now, but sources in the 60s and 70s had no problem calling it a democracy back then, too. Japan seems to have democratized gradually and discreetly, as bureaucracies discredited themselves through poor planning in the 90s. But it’s hard to tell how much stories about, e.g. new PM Yoshihide Suga wearing out shoes from door-to-door campaigning represent a change in the system or just a more evolved form of political theater.

[3] I am fibbing a bit. Actually, the subjects of contemporary books on the state of the world tend to cover a wide variety of topics, ranging from Trump to Donald J. Trump to the current President of the United States to the guy who got Barack Obama’s job. Ironically Fukuyama name-checks Trump twice, as a typical example of an ostentatious rich guy.

[4] And there’s also the possibility of self-destruction. Robin Hanson points out that Fermi Estimates of life make it very implausible that intelligent life came into existence exactly once, and we’d expect to pick up radio transmissions from other species—unless there’s some “Great Filter” that wipes civilizations out once they reach a certain level. Maybe it’s a weapon—perhaps we’re one of a million species to invent nuclear weapons and the only one to get enough lucky breaks.

The anthropic principle is always a useful framework. The first time I read the Pentateuch, I was struck by how often the Israelites were threatened with annihilation, and had to rely on luck, mercy, or Moses' persuasive ability to survive. I wondered if I was reading the story of something that happened once, or the story of the first group of people to be chosen and to not mess it up.

Subscribe now

Elsewhere

Belt & Road & Vial

China is expanding vaccine testing to 18 countries ($, Nikkei), with an emphasis on participants in the Belt and Road Initiative. Vaccine diplomacy is a tool for closer diplomatic ties, as well as closer economic ties—it’s a way for China to directly help an entire country, and at a more limited scale a way to create green zones that include Chinese cities and overseas import and export partners.

Tyler Cowen has a more detailed look at vaccine diplomacy. He makes an important point about China:

It is the rich countries that are calling most of the shots when it comes to vaccines. In this respect, China may be unique: It has some properties of a rich country (a big, advanced scientific establishment), but it has a poor country’s willingness to take risks. That’s one reason China might end up leading on vaccines. The U.S. is ahead of China technologically, but Chinese priorities are more in sync with those of many other countries in the world.

An unbalanced economic model, with more emphasis on heavy industry (in the past) and high tech (right now) compared to domestic consumption has its downsides, but in cases like this it gives China distinct advantages.

CCP and SMBs

The Chinese government has announced plans to more tightly control small private-sector companies, by, among other things, setting a minimum quota for party members even at smaller firms. In the past, some entrepreneurs in China have been reluctant to acknowledge party membership, so in one sense this just represents a more open approach to an existing standard.

Hire Fast, Transfer Fast

Alex Kantrowitz profiles an Amazon effort to automate their pricing. An increasingly common phenomenon in white-collar work is a new spin on training your replacement: machine learning models generally rely on human-labeled inputs, so replacing a human-driven process with an automatic one generally requires those same humans to do a lot of the implementation. Amazon has found an effective workaround for the obvious morale problems this causes: instead of firing employees once they’re replaced by an algorithm, Amazon transfers them to new jobs whose replacement by algorithm is further off.

Life Imitates Science Fiction

A group of state-sponsored hackers working on extensive industrial espionage have been caught because they also hacked into video games to steal virtual items. This is, of course, the plot of Charles Stross' 2008 novel Halting State, and also the plot of 2020’s headlines. Video games represent a great honeypot for security issues: in-game money-laundering is trivial, so games are a tempting target, but the sums involved are small. As long as there’s an overlap between people who hack for a living and people who play videogames for fun, it’s a conveniently tempting target.

Koanic SaaS Secrets

How did Snowflake become the most valuable software IPO of all time? Their CEO has thoughts. Most of this essay is easy to nod along to but—as the author acknowledges—hard to implement. But it’s worth keeping in mind: the management techniques behind extraordinarily successful companies may not be especially complicated, but that doesn’t make them easy.

(Previously on The Diff, I wrote up Snowflake’s S-1. One surprising thing about Snowflake is that it’s a tech infrastructure company that, even prior to its IPO, was more famous to investors than to tech people.)

Interactive Brokers' Big Option Trade

Institutional Investor has a writeup of the traders alleging that negative oil futures prices were the result of market manipulation. This will be hard to prove; the futures contract in question was relatively illiquid, because most traders had moved to the next month’s contract, and illiquid assets often do crazy things during times of high uncertainty. There were certainly traders who would profit from a deeply negative price, but since every trade has a counterparty, there was plenty of money at stake in manipulating the price up. Inevitably, lawsuits are brewing. One interesting detail:

After taking a closer look at the Interactive Brokers settlement documents [in which IB released traders from paying for their full losses from negative oil futures], Institutional Investor can reveal that traders receiving compensation from the company must not only agree to release it from any legal claims arising from the negative-oil-price events, but also transfer any of their own related claims against third parties to Interactive Brokers, assisting the company with its own litigation and providing testimony; attending hearings or trials; and procuring paperwork, if needed.

This is a very interesting trade. IB’s system broke down during the oil crash, since it couldn’t handle negative prices, so it makes sense that they’d compensate traders for their losses. But if IB thinks it’s possible that the market was manipulated, and that they can prove it in court, getting customers to sign over the rights to legal claims is a good way to capture that upside. Depending on how fines are calculated and how many parties get fined, it’s hypothetically possible for the total value of the settlements to exceed the losses. Usually settlements like that are eaten by lawyers, especially if there are many parties with relatively small losses, but since IB has acquired so many claims, it can negotiate better economics with lawyers. Like any good-but-unlucky trader, they took their losses, took stock of the situation, and found an opportunity to reenter the trade at a more favorable price.