Let’s explore why rich societies are more economically productive, and why both Kantian and utilitarian ethics forbid exploit development and SEO, as explained by Bill & Ted’s Excellent Adventure.
Suppose almost everyone in the world worked really hard to get better at football as a hobby, instead of wasting their time watching non-football TV or getting angry about politics. The level of professional football games would improve, but there would still be about the same number of professional football games; maybe a little more diversity, but just as today, most people would watch the matches of the top clubs and not the much larger number of bush-league matches. So the overall well-being of the world would increase, but only a little. Or maybe it would decrease, if more people tore their knee ligaments. Either way, it would be pretty much the same.
Suppose that, instead, almost everyone in the world worked really hard to get good at medicine as a hobby — doing paramedic training, reading case studies, predicting patient outcomes, doing Gwern-style double-blind self-experiments, performing recreational plastic surgery on their pets, that kind of thing. The level of health care would improve dramatically — deaths due to drug interactions the prescribing physician failed to notice would go way down, people’s decisions about when to go to the emergency room would be much more accurate, people’s lifestyle choices would become slightly healthier, and medical science would advance much more rapidly. The overall well-being of the world would increase substantially.
On the gripping hand, suppose that almost everyone in the world worked really hard to get good at fighting with kitchen knives. Perhaps the level of Olympic fencing would improve substantially, which would make for enjoyable TV — though perhaps the sports are far enough apart that it wouldn’t transfer — but it’s also likely that a larger fraction of fights would involve knives instead of just shouting or fists, usually killing one or more people. The training, too, would probably involve a certain number of accidental fatalities, and some fraction of those would lead to revenge killings. More children would be orphaned. The overall well-being of the world would diminish.
On this basis, I claim that it would be better for almost everyone to take up medicine as a hobby than football, and better football than knife fighting; and this is because knife-fighting is a negative-sum game, football is a zero-sum game, and medicine is a positive-sum “game”.
As I remember, sage Dave Long at some point unpacked the Bill and Ted philosophy as follows: “party on” means not to engage in negative-sum “games” — in the very general sense of “interactions with other agents” — and “be excellent” means to engage in positive-sum games. To the extent that you can control which “games” you “play”, it is better to invest your limited time in positive-sum games. (Sage Shakyamuni reportedly took the alternative position that you should not waste your time playing any games, because you are going to die soon, while the sages of Wyld Stallyns instead played games to defeat death itself.)
Of course, a person may find themself in a situation where it is more advantageous to them personally to direct their effort to a zero-sum game or a negative-sum game. Consider the thief who spotted my neighbor, an old woman, at the bus stop last month, and decided to knock her down and beat her to steal $300 from her (about US$7). He chose to direct his efforts to robbing her, and although he injured her in the process, he presumably considered that a worthwhile tradeoff, because it satisfied his desires for food or crack or whatever. A pyramid-scheme participant who recruits downstream members is doing the same thing, but in a less courageous way, and the injury is delivered via deception to the victim’s mind, rather than via a stick to her knee.
From a consequentialist point of view, this is a bad outcome if we consider the thief and their victim to be equally important, so that the loss of money by one is precisely canceled by the gain of money by the other, while the putative injury or deception has no such counterbalancing benefit. If the consequentialist considers the thief’s profit to be more important — perhaps because the thief is poorer and therefore benefits more from the money, or perhaps because they subscribe to a worldview where it is better for the money to go to a brave thief rather than a cowardly victim, or a smart thief rather than a stupid victim, or simply because they are racist — then the consequentialist can justify the theft as morally correct. But an egalitarian consequentialist cannot so justify theft or any other negative-sum game.
Neither can a deontologist who subscribes to Kant’s categorical imperative, which is of course (XXX) one of Kant’s first examples; the thief cannot at the same time will that the victim should steal their money back, delivering the same beating to the thief, because that would leave the thief bruised but no richer. Only by some kind of special pleading can the Kantian thief save his livelihood, and the same kinds of special pleading that the consequentialist could use to justify the theft can be used to exempt the thief from the maxim that would otherwise strip his stolen money from him. Again, this logic applies equally well to any negative-sum game.
For an egalitarian consequentialist, the imperative to “be excellent” — to play positive-sum games rather than zero-sum games — is just as strong as the imperative to “party on”, that is, not play negative-sum games. Kant’s theory (like most deontologist theories) also regards it as a “perfect duty”, and thus obligatory, to “party on” in every possible way — at least in the specific examples Kant gives — but regard it as a supererogatory “imperfect duty” to “be excellent”.
XXX this description of Kant’s incoherency criterion is itself incoherent and needs reworking
In theory, since computers only do what you program them to do, you could decide not to program them to accept arbitrary commands from random people anywhere on the internet, instead only programming them to accept arbitrary commands from their actual owners. That is, there is no need for software to contain security holes. Computer security violations are not inevitable. They are the result of undiscovered programming mistakes, which is why the most significant forum for announcing and fixing security holes in the 1990s was called BUGTRAQ.
In the late 1990s I spent some effort finding security holes in software and reporting them to get them fixed. Unfortunately, it became increasingly apparent that the current economic and intellectual environment was going to introduce new security holes faster than we could remove them, so rather than a gradually decreasing pool of still-undiscovered security holes, we have a gradually increasing one, with the disastrous effects on the human right to privacy documented by the Snowden revelations.
In such an environment, exploit users and defenders are in a sort of arms race, a literal race to see who can respond faster. If the defender is faster to patch a newly discovered hole in deployed systems, they win that round; if the exploit user is faster to acquire and employ an exploit for it, they win that round. So whenever one side or the other increases their commitment of resources and gets ahead a bit, the other side has the option of matching that increase to get back to the previous equilibrium.
This sounds like a zero-sum game, but in a larger context, it’s negative-sum: the human effort spent on the vulnerability treadmill is taken from the time available for making music, discovering new algorithms, cooking food, meditating, painting pictures, writing poetry, reading poetry, watching movies, making love, or raising children. Every sysadmin on call who has to patch the production systems within an hour so that exploit users won’t break in with the newly announced vulnerability is spending the irreplaceable minutes of their life on a Red Queen’s race with the spies at the NSA or the FSB. They are falling short of partying on.
Of course, neither side can unilaterally scale back its efforts; that would amount to surrender. If it’s your job to keep your employer’s public-facing systems up to date so they don’t get popped, you can’t just decide it’s not worth the effort. But you can find a new job.
There are other drawbacks as well: it’s no longer a viable option to continue running outdated software unless it’s in a very unusual isolated environment, like a non-networked video game, and the increasingly rapid and frequent response required to new vulnerabilities has the effect of centralizing both patching and exploit use in large organizations. This means everyone is exposed to the risks of deploying untested new software on short notice and to having strangers administer their most intimate computer systems, such as Android hand computers. Also, sometimes patching a vulnerability unavoidably introduces incompatibilities, causing bitrot.
So I stopped spending effort on that.
The nuclear arms race during the Cold War had a similar dynamic: each side constantly worked to preserve a second-strike capability (by, among other things, building enough warheads that some would be likely to survive a first strike from the other side) and to find ways to remove the other side’s second-strike capability. Neither side could opt out of the game, but the result of both sides playing it harder and harder was the decades-long threat that civilization could end at any time, with 20 minutes of warning.
Fortunately there were also people on both sides like Jonas Salk, Norman Borlaug, and Andrey Kolmogorov who were able to dedicate their lives to positive-sum games instead of the negative-sum game of the nuclear arms race.
Commercial competition is, in theory, a positive-sum game, though eventually only slightly so — as more and higher-frequency participants in a stock market means fairer prices, with retail participants having to pay much narrower spreads to the market-makers, more competition in consumer-goods markets should result in goods priced just above the lowest possible marginal cost of production. One reason this doesn’t happen in practice is advertising: consumers buy goods that are advertised rather than equivalent goods that are not, and in many cases are induced to buy categories of goods they wouldn’t have bought at all without advertising.
This puts advertising in the negative-sum category: to survive, firms are forced to push positive information about their products in front of customers, whether that’s by traditional display ads and flyers, by getting newspapers to write articles about them, by commissioning shill research (such as rigged Gartner product comparisons), by buying product placements in supermarkets and music videos, or by doing SEO on their web pages. Moreover, they cannot afford to be much more forthcoming about the drawbacks of their products, or less enthusiastic about their benefits, than their competition is, unless consumers are turned off by their immodest hucksterism.
To the extent that consumers can find objective research about the relative merits of different products and distinguish it from the advertising, all these strategies will be ineffective, so firms must work harder and harder to disguise their hustle and puffery as objective research, making it harder and harder for customers to find objective information about their products and, ultimately, about anything at all.
And so it is that if you search for almost any commercially relevant topic on Google today, the results are the Wikipedia article and nine shill pages.
You might reasonably think there’s a prisoner’s-dilemma-type global-local incentive conflict when it comes to playing negative-sum games: even if it is not in the interest of society as a whole that you beat up old women to rob them, you might perceive it as in your own interest, however you define that. And surely for some definitions you will sometimes be correct. But this happens less often than you might think.
First, let’s consider negative-sum games like retail day-trading, in which all the participants are voluntary (as contrasted with the game of robbing old women at the bus stop, in which my neighbor was an involuntary participant). This is a mildly negative-sum game because of broker commissions and the spreads paid to market makers, which are siphoned off of the otherwise-zero-sum transactions between retail day-traders. (The advent of penny pricing and algorithmic HFT has enormously diminished the magnitude of the spreads in the last 15 or 20 years.)
Presumably most of the participants in the game believe it is in their interest to participate, although there might be a few acknowledged addicts who just haven’t managed to quit even though they know it’s bad for them. But many of them are wrong, more than half in this case. That means that if you think it’s in your interest to participate because you will make money, you’re likely just mistaken. In this market you really are behind a Rawlsian veil of ignorance, not knowing if you are predator or prey. I’ve watched smarter people than myself waste fortunes on such mistakes.
But there’s another, subtler problem. The people who participate in some activity, whether it’s football, day trading, knife fighting, medicine, SEO, or beating and robbing old women, form an affinity group — a “community of practice”, it’s sometimes called. They tend to talk to each other, sharing information about their shared activity, and often they engage in other transactions with one another — cellphone thieves need cellphone fences, for example, who in some sense form part of the same community of criminal practice. So if you decide to spend time day trading, you’re also implicitly deciding you’re going to spend some time hanging out with day traders, talking with them, maybe buying them a beer from time to time. And similarly if you do any kind of medicine — even if you’re just taking CPR training — you’re going to spend some time hanging out with medical people.
So, we can reasonably ask, how might hanging out with medical people differ from hanging out with cellphone thieves? And there are a lot of answers that are specific to these fields of endeavor (for example, medical people tend to be from Cuba, while cellphone thieves tend to own motorcycles) but one difference that’s common across all these fields is that the kind of people who choose to play negative-sum games are (however slightly) the kind of people who choose to play negative-sum games, valuing their individual interests over others’, while the kind of people who choose to play positive-sum games are less so.
We can renormalize this and get a stronger result: if you beat up old women to rob them, then the people who choose to hang out with you knowing this will tend to be people who think that’s a reasonable choice, but who aren’t afraid you will beat them up and rob them — often because they think they’re stronger than you, which (refer to previous lemma) they believe justifies robbing you. And similarly for SEOs, advertisers, day traders, and soldiers, with appropriate variations.
So, if you go out drinking with doctors and with day traders, you should expect the day traders to stick you with the tab more often. Maybe not much, but detectably. And if you hang out with people who beat up old women to rob them, you’re likely to get robbed, one way or another.
(Of course, it’s also possible that someone could take revenge on you. But that’s true in positive-sum games too — positive-sum games aren’t always win-win, though zero-sum and negative-sum games are never win-win.)
This doesn’t mean you’re better off if you hang out with people who advocate playing positive-sum games and not playing negative-sum games. A noticeable fraction of them are just trying to talk you into giving them your money, or your volunteer time, or your vote, or leaving your monogamous partner alone with them, or whatever.
“Social capital” is often invoked to explain why rich countries experience vastly higher levels of economic productivity from workers with similar levels of education employing similar levels of capital intensity than poor coutries do; the effect is so strong that poor people from poor countries can greatly improve their income simply by working in rich countries.
After 12 years of living in Argentina and plenty of opportunity to contrast the beliefs and practices of different social groups, one of my conclusions is that much of what is known as “social capital” or “high trust societies” amounts to a self-reinforcing tendency for people to play positive-sum games rather than negative-sum games, or to play positive-sum games in a wider context (not, for example, only within their own family); and this accounts for the otherwise puzzling overwhelming dominance of Wikipedia contributions from highly developed countries, far out of proportion to any difference in literate populations.
Here in Argentina, two centuries of often-militarized factionalism has created a profound cynicism about politics and about any attempt to carry out positive-sum projects; politicians get and maintain power by polarizing their constituents against other politicians and their constituents and by using the state to extract wealth to give to their own supporters, with the unsurprising result that everyone has concluded that all politicians are thieves and liars — because would-be politicians who are not thieves and liars are not successful at gaining support.
The popular belief here that all games are zero-sum is so strong that an Argentine woman told me once that Argentine cars are of poorer quality than American cars because American companies send their defective parts to Argentina, as if statistical process control worked by producing several times the needed quantity of goods and discarding most of them. This belief is strengthened by the Reagan-like deployment of positive-sum rhetoric by politicians to promote negative-sum policies that enrich the rich at the expense of the poor; Macri, our current president, is especially guilty of this, and his economic policies have been disastrous.
We can also argue for a subtler effect, similar to the renormalization argument in the previous section. Leaders of factionalist ignoramus societies, such as Donald Trump, Hugo Chávez, or Mauricio Macri, cannot afford to take advice from the intelligentsia, or put them in charge of policy, even if they could figure out who they were; they are forced to assume that wiser heads from other factions are looking for ways to defeat them, and will use delegation of authority or even openness to advice as a way to undermine their leadership. (See Notch scorn for pompous bloviation on this topic.) Instead, they are forced to delegate authority to people loyal to their own faction, and listen only to their advice, with the predictably disastrous consequences we are seeing today in Venezuela.
(Given that Trump is president of the country with half of the world’s top universities, it may seem strange to call it a “factionalist ignoramus society” — but Trump has apparently spent his entire life playing zero-sum and negative-sum games, and is certainly an ignoramus who surrounds himself with ignoramuses, and almost half the population of the country voted for him anyway. So clearly the factionalist ignoramus element of that society, always significant, is today its dominant element, just as in Germany after 1933.)
You might think that the pervasive belief in zero-sum-ness would make it impossible to motivate voters to participate in politics at all, since clearly whatever time you spend handing out campaign fliers or getting gassed by the cops (a nearly universal experience among the lower and middle class in Argentina) isn’t going to be compensated by the marginally increased chance of your party winning at the polls and putting in place whatever policies you favor, even presuming those policies benefit you and not just the politicians you elected. But humans are not primarily motivated by such calculations, and never have been; any who were would fail Newcomb tests and be cast out of society to die in the wilderness.
Suppressing factionalism is no panacea either — whenever those in power pursue foolish policies, they can reasonably accuse anyone who criticizes those policies of fomenting factionalism, and this commonly happens in modern China, in the modern Bahá’í ecclesiastical hierarchy, and in the medieval Catholic church, for example, all of which have or had strong taboos against factionalism. Just as extreme factionalism deprives the leadership of the benefit of the collective cognition of the wise, so too does conformism.