As I get further into the organizational problem– with the hope of, one day, working a way out of it– it becomes increasingly clear that there is no simple solution. There are a variety of principles that sometimes contradict. Consider bridge bidding, with various guidelines on how to send signals to one’s partner and find a winning contract, but with ambiguity and even contradiction. There are a variety of strategic concerns that must be balanced, and what makes it interesting is the fact that the right thing to do is not at all obvious. If bridge were deterministic and solved, there’d be no such thing as a good bridge player. It’s the difficulty and chaos of the game that makes it fun. Then, the more complex game of business formation will even be harder to “solve”. I expect that, when I finally am ready to “solve it”, I’ll arrive at twenty principles that should, by then, seem obvious– not one “closed-form” model. Certainly, I think the previous 11 essays have given us some insight into organizational corruption, without falling back on any prevailing pessimism (that organizations always tend toward failure). We have a good sense of why the MacLeod tiers emerge. We still need the core ideas that will help us build something different.
There are a few concepts that deserve further exploration: growth, risk, and chaos.
Growth, and how it influences behavior
Macroscopic economic growth is a fairly modern phenomenon. Before 1700, global economic growth was never faster than 0.3% per year, and generally slower. Most people who were rich had made others poor to get there. A zero-sum approach to human problems made sense. Conservative, religious institutions had a lot of power, and for understandable reasons. If large-scale material growth is not to be expected, then people need to organize themselves in a way that does the best with what is available. Religion gave people access to social plans that had been tested by time. Some of its dictates were nonsense; some were incredibly valuable. On the whole, it probably did more good than harm from a single-time perspective (its retardation of progress requiring another discussion).
A late 18th-century Anglican clergyman named Thomas Malthus is notorious for having been “wrong” in his thesis about population sustainability– he predicted a mid-19th century calamity in England. In truth, he wasn’t that wrong with most assumptions. By 1798, economic growth had accelerated to just under 1 percent per year. He asserted that exponential population gains would outpace economic growth, which he modelled as linear. On that, he was incorrect. On the conclusion– that the economy was not growing fast enough to support human population growth, and that the latter would be checked by famine, disease, or war– he was still right. Malthusian catastrophes are all over the place in history. One was to see it is that the Industrial Revolution intervened. A darker view is that the 19th-century English catastrophe was outsourced to Ireland.
Religious institutions, being conservative while expansive both in time and space, were storehouses of “big picture” knowledge about a society’s history and evolution. Typical people saw rises and falls: frenzied victories, hideous defeats, tribal and racial hatreds, and the formation and dissolution of societies and cultures. There’d be spells of peace and war, and control would pass from one set of hands to another, but global progress was rare until the 18th or 19th century. Religion– and, later, the narrative-driven nation-state– emerged to create order in an almost zero-sum world where it was inconceivable that any economy could grow at a rate comparable to the human desire to populate. It didn’t solve the problem, but it provided explanations and planning. Over time, religion’s central role was eclipsed when competent outsiders and specialists began attacking human problems. One of those specialties was economics, truly a “dismal science” in a time of sub-1 percent growth, explanatory of famine and war.
Clearly, we live in a different time. Economic growth is faster than population increases, so the average person’s standard of living is improving. We hit a phase change at some point between 1725 and 1925, the exact date of it still being debated, at which point economic growth stopped being subsumed by population increases, enabling genuine long-term improvement. The zero-sum outlook no longer makes sense. Progress and growth are now expected. So what causes economic growth? Where does it come from?
“Improvement of processes” is a common first answer. The Industrial Revolution enabled us to get better at stuff. It’s the culmination of thousands of years of progress. Writing gave us stable history. Mathematics gave us precision and mechanisms for solving problems. Empiricism and science gave us the ability to measure things. Technology made us more powerful, more productive. Each invention or innovation we kicked out pushed us forward, and the economic value of our knowledge base has been a faster-than-linear function of its size, while that knowledge base itself grows exponentially. We’ve had an ongoing story of faster-than-exponential growth going back to the advent of sexual reproduction 1.2 billion years ago due to a self-accelerating improvement of processes, but it’s still a bit vague what that means. We should find “improvement of processes” to be intensely suspect. Why aren’t processes already in the improved state?
Almost always, improvements come from outsiders, so the macroscopic change is a “creative destruction” driven by replacement. Most of China’s lasting intellectual accomplishments came from the “outside elite” of its scholarly civil-service structure, and not the upper-crust mandarins. Greek philosophical advancements are not owed to reputable clerics of that time, but to heathens, many of whom disavowed the traditional gods and were persecuted for doing so. In medieval Europe, a despised class of merchants grew richer and, eventually, more powerful than the hereditary aristocracy. Large companies are often defeated by innovative small ones run by people the conglomerate would never hire. Social substructures seem to have a “natural growth rate” that, to the extent that it evolves, decreases over time as the substructure reaches saturation or even decline. What drives the faster-than-exponential growth seen thus far is the generation of new ones with faster growth rates. Where do those come from? There is, metaphorically, a place where genuine creation occurs. Chaos. Nothing begets something because there is no nothing. The closest this world (possibly the universe, possibly any universe) has to nothing is pure chaos, which is still fundamentally creative. I’ll get back to that concept.
In 2013, most genuine growth– of the economy, of human knowledge, of technology– comes not from established entities but from new ones trying to force their way into existence as they emerge out of chaos. Large, established organizations have given up on progress, by cutting R&D funding and focusing on next-quarter profits. Progress has also given up on them. The smartest and most ambitious people, who wish to harness chaos for new creation (or, to see it more cynically, enrichment beyond the pittance judged fair by some massive organization) generally lack interest in the corporate behemoths with entrenched processes and slow growth. Don Draper, in departing from a sclerotic corporation (in which his advertising agency had become a subsidiary) to form a new one, said it best:
Who the hell is in charge, a bunch of accountants trying to make a dollar into a dollar ten? I want to work. I want to build something of my own.
Draper (a MacLeod Sociopath, without question) doesn’t want the restrained, conservative growth of a bulky corporate enterprise. He wants something more rapid, personal, and fun. So he left an existing substructure to form a new one.
That was 1963, when the first hints of a technological era just appearing. In 1863, industrial growth actually was the most exciting game in town, and people would have been thrilled at the prospect of turning “a dollar into a dollar ten”. For millennia, humanity had been in a zero-sum arrangement where being powerful meant dominating and controlling other people. From the Egyptian pyramids to the Colosseum to the aristocracy of the Southern United States was a trail of monuments that required slavery. Industry provided a way out: the possibility to turn $1.00 into $1.10 without hurting anyone. Compared to the zero-sum squabbling, that was immensely progressive. With industry catching on, nations all over the world abolished slavery in the mid-19th century. Getting rich no longer required making others poor, or unfree, or dead. The downside of this was turning work into something often too stable and boring to excite the Don Drapers of the world.
Why is it so common to call someone like Don Draper a sociopath? The state of society delivers a prevailing growth rate. For agrarian societies, it was near zero. For industrial ones, it was slow (1 to 5 percent per year). As we move into the technological era, it might become higher. Whatever that rate is, evidence strongly suggests that there will always be people who want to grow their fortune at a faster clip. One way to do this is to steal. That’s what war (ending lives to rob the dead) and slavery (stealing freedom and autonomy) were about. The other, much harder, way is through new invention and the creation of value that didn’t exist previously. It might like these are being put forward as vice and virtue, eternally separate. It’s not so. It’s a lot more complicated. Is mining for gold, diluting its financial value in the same way as counterfeiting does, a zero-sum theft of financial value that delivers nothing to the world? Or is it, since the yellow metal has some hedonic value in so far as people like to look at it, a productive activity? This is hard to answer. Additionally, how does an enlightened or altruistic industrialist (circa 1750) ensure deliverance of value when powerful forces will divert any produce to zero- or negative-sum pursuits, such as warfare? How does the engineer make sure his efforts are used to build more plowshares rather than swords? I don’t think he can. In any case, that’s not important from a macro perspective. New invention and raw theft both come from a class of rule-breakers who aren’t content to have their fortunes grow at the prevailing rate.
The problem of chaos
Often, “chaos” is used to describe squalor or malfunction, but its original meaning is closer to abyss, or the formless void from which the universe emerged. It’s not “nothingness”, because things (such as the universe) come from it. One might think of it as a pregnant silence, or a blank canvas. It’s creative emptiness.
Experientially, we know that the best way to create a chaos is to clear something away. Many meditation practices can be viewed this way. The principled and mindful attention, with a calming intent, to thought processes leads us away from the toxic, repetitive, and mostly negative thoughts that occur in conditioned life. We create a chaos into which new forms of thoughts and experience– which wouldn’t have otherwise existed– can come into being. Sensory deprivation, sleep, and dreaming are also forms of chaos, in so far as they induce experiences that seem not to be produced by the objective world.
Chaos, as a source of something, we welcome so long as we trust our ability to filter the positive from the negative. Chaos, as a state of nothing, we view negatively. We strive to differentiate ourselves from the scarcity, formlessness, and indifference of primordial chaos. It is an open question how far chaos is to be desired. It’s chaos that creates the need to build, and that gives us the tools to do it. Law, order, and structure with intent toward permanence are built by some to protect people against the pain of chaos. However, an alternative approach (more common in Eastern religions) is to embrace chaos and impermanence– to recognize that it is better to adapt to chaos than to cling to the flimsy things that we invent to protect ourselves from it.
It is reductive and useless to call chaos “good” or “bad”. It’s neither. Nor is it random. In fact, some aspects of it (such as unexplored mathematics) are quite structured. It is, however, unexplored and mostly incomprehensible. People find it discomforting, the ultimate chaos (from a human viewpoint) being death. There are, however, degrees to which people accept or avoid it. Those who engage chaos directly will be more creative, but also more volatile. Stepping back from the metaphysics, we have (in chaos) a way to understand economic progress. In pre-industrial times, a person wanting to become rich without harm to others might search for gold. The earth itself was once a chaos. The Greeks acknowledged this by assessing ownership of all material wealth to the temperamental god of the underworld: Hades.
In 2013, gold is a effectively a commodity. It is expensive per ounce, but because a large amount of human service is required to produce that ounce (reflecting the metal’s rarity). Nothing special distinguishes that service. So, those who aspire to wealth (or growth) in 2013 are not likely to dig for gold. The technocrats don’t have a special proclivity for digging the earth in search of a yellow metal. They leave that to the specialists who have the equipment. Instead, they mine chaos.
Entrepreneurs and innovators scan various chaoses with the hope of putting something together that is of value to other people. Transfer occurs from one chaos (unrealized ideas) to a painful one that there is benefit in filling (unmet human desires). It might be called “magic”. It isn’t. However, it’s unpredictable and intermittent by nature. Once that chaotic transmission is completed, the work involved is no longer inherently exciting. It’s a commodity. (Making easier the jobs of those who have to do it, on the other hand, remains potentially fruitful.) Once the chaos is taken out of the equation, what remains is dull labor.
From chaos, we can understand the nature of convexity. There are some people who have the skill to go “into” chaos (or, at least, a subchaos related to a specialty) and find something useful. To the unskilled, chaos simply looks “random” and dark. To continue the metaphor of chaos as a space, the improbability of a find is (to an unskilled person) an exponential function of how deep into chaos it lives, because more “lucky” steps are required for an unskilled person to get there, and the multiplication of low probabilities has that effect. In other words, the farther a find or job is from the well-ordered and easily understood territory, the more that personal skill and knowledge matter, it’s likely an exponential relationship. Things being valuable in proportion to their rarity, we see where convexity comes from.
The risk thing
Industrial businesses begin as chaos-mining operations, but after carving out a space of order, seek to protect it. Standardization becomes the goal. We can’t exactly quantify (or even perfectly define) chaos, but we can quantify risk. Risk and chaos are fairly related. Risk pertains to how interactions with chaos might affect an entity’s economic health. It’s not concerned with the whole of chaos, but only about what threats might come out of it. Financial risk is even measurable, to some degree, enabling portfolio managers to discuss “how much risk” the company has. We’ve now seen a mature commoditization of that kind of risk, with markets able to price assets not only based on their expected (average-case) yield but to account of risk according to what the principal players find desirable.
All else the same, risk is considered undesirable. Most people would rather have $5 million than a 50% chance at $10 million. Let’s separate the value of an asset into its mean value and a zero-mean risk variable. The first possibility is that one has $5 million. The second is that one has $5 million plus a 50% chance of winning $5 million and a 50% chance of losing $5 million. Most people would value that zero-mean risk negatively. A person with the latter portfolio might be inclined to “sell” that risk for -$10,000– that is, to pay $10,000 for someone else to take that variation and have a solid $4.99 million instead. Much of finance is about figuring out fair prices at which to transfer risk.
I’ve previously discussed law and chaos in the context of alignment. Are these connected with the chaos described above? Absolutely. Risk and chaos aren’t the same thing; the latter is an injection from chaos known to have an effect on one’s well-being. Computationally, we process risk because chaos is beyond what we can quantify. Civil alignment correlates to a person’s attraction or repulsion to chaos. Lawful people tend, in general, to have faith in the infrastructure that humans have created to hold chaos at bay. They prefer by-the-book solutions to problems because they fear the chaos of improvisation. Chaotic people, on the other hand, see chaos as potentially beneficial. They want to mine it. Measured in terms of risk, lawful people are, in general, going to be more risk averse. Chaotic people tend to be risk seekers.
This is not necessarily true, however, in terms of financial risk, on account of its commoditzation. Financial risk can be separated into an expected value and a zero-mean random variable whose variability itself can be measured. This enables the commoditization (measurement, trade) of financial risk. That’s not to say that there aren’t black swan, out-of-context, risks out there (that rarely follow Gaussian distributions) but those are often placed in a different category. Because of this commoditization, financial risk has largely been divorced from a personal law/chaos bias. Lawful people with means will pursue investment strategies with high volatility in order to get high returns; those with less means or likely to need liquidity soon might favor less volatile strategies. However, it has little to do with a person’s often visceral reaction to chaos.
If financial risk is a commodity, it can be allocated. Traders have risk allotments based on past performance and seniority that represent how much risk they can take on behalf of the firm. Such rules are necessary to resolve the conflict of interest that exists. Traders, usually paid on commission, have an upside-biased risk profile: making $40 million is twice as good as making $20 million, but losing $20 million and losing $40 million are identical– both result in getting fired, but there’s no further consequence. Without risk limits, traders would have the incentive to take on risks that the firm (absorbing wins and losses) would not want.
Typical business organizations do not have as well-formed an understanding of risk as trading desks, because financial risk has been mostly commoditized, while the performance and chaotic risks that businesses deal with cannot be. However, that mentality is still in force. The organization earns a profit because it takes on risk: otherwise, in a competitive market, there should not be profits. There is, therefore, a certain amount of risk to be expected, and little more should be tolerated. As a trading desk would distribute risk allocations among its traders, a standard business organization attempts to create a risk allocation regime for its people. The firm must allow some small set of people to take chaotic risks, because the world is lawless and volatile and a firm that ignores chaos outright will struggle to thrive. Those people are called executives. Then there are people trusted with financial and performance risks (such as assessing people to be hired and fired). Those categories of risk are seen as “less dangerous” because financial risk is easy to measure, and performance risk over concave labor, while not directly tradable, falls within a tight bell curve. Those people are called managers. Workers, in this model, should not be trusted with any risk at all. As the organization sees it, they already bring too much risk by walking in the door to have any right to ask for more.
In the optimization model put forward for the corporation, I discussed the idea that a manager’s job is to hill climb to the top of a neighborhood (gradient ascent) and find a local maximum, while executives are trusted with non-local explorations that might lead to finding better hills. Non-locality implies that the executive is going into uncharted territory, or engaging directly with chaos. He’s not, however, typically allowed to go very far in, but he has some non-zero chaotic risk allocation.
The human side
One way to view the organization’s miserly risk allocation protocol is to say that it’s inhumane and demoralizing. Engagement with chaos is part of what makes us human, is it not? Most Americans participate in activities that are more industrious and difficult– such as picking their own fruit, open-source programming, and independent writing– than their pointless, subordinate office jobs. Even risk-taking is a hobbyist activity, if sometimes a destructive one, in the form of gambling (engagement with an otherwise uninteresting chaos). “Work”, for most people, is a boring and unhealthy psychological monoculture, leading to the question: why do people tolerate it in the first place?
Risk and chaos are the forces at play. Deep into chaos is somewhere that most people don’t want to be. It’s lonely, unsettling, and weird. Without financial constraints, most people would still fall into routines over time: people, places and work that make them happy. So there is an inherent willingness for those who are more chaos-averse to enter the ordered zone of a facile subordinate position. There are psychological reasons for people to take the MacLeod Loser deal. Many people would rather have the comfort of a stable group than attempt to lead it and risk rejection or group dissolution. Organizations exist to diminish chaotic risk for themselves, but in doing so, create a realm that is highly ordered and allow the chaos-averse to make a home there.
There are also the financial aspects, and that discussion becomes a bit less humane. People make the financial Loser deal because most of them have no choice. They need a stable monthly income. That trade, both from a micro- and macroeconomic perspective, tends to get worse over time. The low savings of lifetime wage-takers forces them to continue making this risk-reducing trade, limiting their leverage, and consigning them to take deals that are increasingly risky (end of corporate loyalty) but increasingly costly to them. It’s a feedback loop that keeps “the 99%” tied in to a certain pattern where they are forced to buy risk reduction, even if it brings them down to a subsistence wage.
There are some, however, who react with a certain insubordination. They get what business corporations are about and learn quickly how to play them. Initially, they will not be allocated chaotic risk by their firms. They just take it. These are the MacLeod Sociopaths.
Convexity and chaotic acceleration
Business corporations exist to create a process that reliably generates income. Their initial architects might glance around in chaos early on to find a source of profit, but once that is accomplished, they are almost all about law. What little engagement with chaos the firm needs is handled by a nerve center containing a small set of people called “executives”. Everything else lives in, or is forced to live in, the concave, far-from-chaos world of “another day, another dollar”.
The industrial world began when scientific advances altered the labor model. I’ve discussed concavity and convexity, but what’s the labor model that has been with humanity for most of its existence, before the industrial era? A binary one, in which there are compliance and noncompliance. That’s the world of slave labor. A noncompliant person was beaten, a compliant one was not. This judgment of compliance might not have had any connection with reality, of course, and hardworking people frequently got the “noncompliant” treatment. It was about emotion and perceived loyalty. The industrial world, in which productivity was derived from systems of conditions rather than exertion alone, encouraged people to look into concerns like morale and quality of training as “hidden” force multipliers that mattered at scale, far more than individual perceptions of loyalty. It no longer matted how many compliant people one had, but how they were arranged. The concave model replaced the binary loyalty-based model of slave labor, and it became clear that coercive labor was no longer tenable. Semi-coercive wage labor, with the worker financially dependent but free enough to change bosses, won out.
Since the 1940s, the binary model of labor has returned, but in a different and benign form: computing. Given a job, properly specified, a computer will do it without complaint. Such machines are extremely good at following deterministic laws. In almost all cases, a computer program will get the right answer to a well-formed problem either 100% of the time, or 0% (a bug). Exceptions (non-deterministic “Heisenbugs”) exist and are extremely painful to deal with, but they tend to be rare in critical components, at least by human standards. Almost no human could multiply two 100-by-100 matrices without mistakes. If the labor is intrinsically binary in value, we can specify requirements and usually program to them. If it is concave, we can often specify what perfect completion (or, at least, an arbitrarily close approximation) is and program for that. Work that lives in the lawful world that human society has already explored is all being done, or will soon be done, by machines.
What’s left for us to do? Convexity, which will require us to move away from semi-coercive labor to a fully free system based on intrinsic motivation. The industrial world saw risk as a commodity that could transferred, and allocated the right to take risks to a small number of people. That works for the typical financial (or performance) risk, since it can be separated into a constant expected value and zero-mean random variable, the latter of which can be traded (often “synthetically” through derivate contracts). In the convex world this separation of risk cannot be performed. It doesn’t make sense. Workers not allowed to take creative risks won’t create. It won’t be useful to employ them, then. The three-tiered corporation ceases to be functional.
What remains, once the machines have conquered the concave world, is the chaos of an unsolved problem. Can we handle it, as humans? Sure. We always have. But how will corporations survive it? They exist to produce law but, in the technological era, the rate at which a company will need to grow to be competitive is one that is innately chaotic. Some companies claim (most, without meaning it) that they want every employee to participate in the growth process. Not so far from now, that will be reality, but that requires a dramatically different view of the organization.
Image may be NSFW.
Clik here to view.
Image may be NSFW.Clik here to view.
