The world sucks at finding the right work for engineers.

This is directly in response to Matt Aimonetti’s “Engineers Suck at Finding the Right Jobs“, because I disagree that the problem resides solely with engineers. Rather, I think the problem is bilateral. An equally strong argument could be made that there’s an immense wealth of engineering talent (or, at least, potential) out there, but that our contemporary business leadership lacks the vision, creativity, and intelligence to do anything with it.

Don’t get me wrong: I basically agree with what he is saying. Most software engineers do a poor job at career management. A major part of the problem is that the old-style implicit contract between employers and employees has fallen to pieces, and people who try to follow the old rules will shortchange themselves and fail in their careers. In the old world, the best thing for a young person to do was to take a job– any job– at a reputable company and just be seen around the place, and eventually graduate into higher quality of work. Three to five years of dues-paying grunt work (that had little intrinsic career-building value, but fulfilled a certain social expectation) was the norm, but this cost only had to be paid once. The modern world is utterly different. Doing grunt work does nothing for your career. There are people who get great projects and advance quickly, and others who get bad projects and never get out of the mud. Our parents lived in a world where “90 percent is showing up”, whereas we live in one where frequent job changes are not only essential to a good career, but often involuntary.

Software engineering is full of “unknown unknowns”, which means that most of us have a very meager understanding of what we don’t know, and what our shortcomings are. We often don’t know what we’re missing. It’s also an arena in which the discrepancy between the excellent and the mediocre isn’t a 20 to 40 percent increase, but a factor of 2 to 100. Yet to become excellent, an engineer needs excellent work, and there isn’t much of that to go around, because most managers and executives have no vision. In fact, there’s so little excellent work in software that what little there is tends to be allocated as a political favor, not given to those with the most talent. The most important skill for a software engineer to learn, in the real world, is how to get allocated to the best projects. In fact, what I would say distinguishes the most successful engineers is that they develop, at an early age, the social skills to say “no” to career-negative grunt work without firing oneself in the process. That’s how they take the slack out of their careers and advance quickly.

That’s not the same as picking “the right jobs”, because engineers don’t actually apply to specific work sets when they seek employment, but to companies and managers. Bait-and-switch hiring practices are fiendishly common, and many companies are all-too-willing to allocate undesirable and even pointless work to people in the “captivity interval”, which tends to span from the 3rd to the 18th month of employment, at which point leaving will involve a damaging short job tenure on the resume. (At less than 3 months, the person has the option of omitting that job, medium-sized gaps being less damaging than short-term jobs, which suggest poor performance.) I actually don’t think there’s any evidence to indicate that software engineers do poorly at selecting companies. Where I think they are abysmal is at the skill of placing themselves on high-quality work once they get into these companies.

All of this said, this matter raises an interesting question: why is there so much low-quality work in software? I know this business well enough to know that there aren’t strong business reasons for it to be that way. High-quality work is, although more variable, much more profitable in general. Companies are shortchanging themselves as much as their engineers by having a low-quality workload. So why is there so little good work to go around?

I’ve come to the conclusion that most company’s workloads can be divided into four quadrants based on two variables. The first is whether the work is interesting or unpleasant. Obviously, “interestingness” is subjective, so I tend to assume that work should be called interesting if there is someone out there who would happily do it for no more (and possibly less) than a typical market salary. Some people don’t have the patience for robotics, but others love that kind of work, so I classify it in the “interesting” category, because I’m fairly confident that I could find someone who would love to do that kind of work. For many people, it’s “want-to” work. On the other hand, the general consensus is that there’s a lot of work that very few people would do, unless paid handsomely for it. That’s the unpleasant, “have-to” work.

The second variable is whether the work is essential or discretionary. Essential work involves a critical (and often existential) issue for the company. If it’s not done, and not done well, the company stands to lose a lot of money: millions to billions of dollars. Discretionary work, on the other hand, isn’t in the company’s critical path. It tends to be exploratory work, or support work that the firm could do without. For example, unproven research projects are discretionary, although they might become essential later on.

From these two variables, work can be divided into four quadrants:

Interesting and Essential (1st Quadrant): an example would be Search at Google. This work is highly coveted. It’s both relevant and rewarding, so it benefits an employee’s internal and external career goals. Sadly, there’s not a lot of this in most companies, and closed-allocation companies make it ridiculously hard to get it.

Unpleasant and Essential (2nd Quadrant): grungy tasks like maintenance of important legacy code. This is true “have-to” work: it’s not pleasant, but the company relies on it getting done. Boring or painful work generally doesn’t benefit an employee’s external career, so well-run companies compensate by putting bonuses and internal career benefits (visibility, credibility, promotions) on it: a market solution. These are “hero projects”.

Interesting and Discretionary (3rd Quadrant): often, this takes the form of self-directed research projects and is the domain of “20% time” and “hack days”. This tends to be useful for companies in the long term, but it’s not of immediate existential importance. Unless the project were so successful as to become essential, few people would get promoted based on their contributions in this quadrant. That said, a lot of this work has external career benefits, because it looks good to have done interesting stuff in the past, and engineers learn a lot by doing it.

Unpleasant and Discretionary (4th Quadrant): this work doesn’t look good in a promotion packet, and it’s unpleasant to perform. This is the slop work that most software engineers get because, in traditional managed companies, they don’t have the right to say “no” to their managers. The business value of this work is minimal and the total value (factoring in morale costs and legacy) is negative. 4th-Quadrant work is toxic sludge that should be avoided.

One of the reasons that I think open allocation is the only real option is that it eliminates the 4th-Quadrant work that otherwise dominates a corporate atmosphere. Under open allocation, engineers vote with their feet and tend to avoid the 4th-Quadrant death marches.

The downside of open allocation, from a managerial perspective, is that the non-coercive nature of such a regime means they have to incent people to work on 2nd-Quadrant work, often with promotions and large (six- or seven-figure) bonuses. It seems expensive. Closed allocation enables managers to get the “have-to” work done cheaply, but there’s a problem with that. Under closed allocation, people who are put on these unpleasant projects often get no real career compensation, because management doesn’t have to give them any. So the workers put on such projects feel put-upon and do a bad job of it. If the work is truly 2nd-Quadrant (i.e. essential) the company cannot afford to have it done poorly. It’s better to pay for it and get high quality than to coerce people into it and get garbage.

The other problem with closed allocation is that it eliminates the market mechanic (workers voting with their feet) that allows this quadrant structure to become visible at all, which means that management in closed-allocation companies won’t even know when it has a 4th-Quadrant project. The major reason why closed-allocation companies load up on the toxic 4th-Quadrant work is because they have no idea that it’s even there, nor how to get rid of it.

There’s no corporate benefit to 4th-Quadrant work. So what incentive is there to generate it? Middle management is to blame. Managers don’t care whether the work is essential or discretionary, because they just want the experience of “leading teams”. They’re willing to work on something less essential, where there’s less competition to lead the project (and also a higher chance of keeping one’s managerial role) because their careers benefit either way. They can still say they “led a team of 20 people”, regardless of what kind of work they actually oversaw. Middle managers tend to what little interesting stuff these discretionary projects have for themselves, placing themselves in the 3rd-Quadrant, while leaving the 4th-Quadrant work to their reports.

This is the essence of what’s wrong with corporate America. Closed allocation generates pointless work that (a) no one wants to do, and (b) provides no real business value to the company. It’s a bilateral lose-lose for the company and workers, existing only because it suits the needs of middlemen.

It’s common wisdom in software that 90 to 95 percent of software engineers are depressingly mediocre. I don’t know what the percentage is, but I find that to be fairly accurate, at least in concept. The bulk of software engineers are bad at their jobs. I disagree that this mediocrity is intrinsic. I think it’s a product of bad work environments, and the compounding effect of bad projects and discouragement over time. The reason there are so many incompetent software engineers out there is that the work they get is horrible. It’s not only that they lack the career-management skills to get better work; it’s also that good work isn’t available to them when they start out, and it becomes even less available over time as their skills decline and their motivation and energy levels head toward the crapper.

I don’t see any intrinsic reason why the world can’t have ten, or even a hundred, times as many competent software engineers as it has now, but the dumbed-down corporate environment that most engineers face will block that from coming to fruition.

There’s an incredible amount of potential engineering talent out there, and for the first time in human history, we have the technology to turn it into gold. Given this, why is so much of it being squandered?

The end of management

I come with good news. If I’m correct about the future of the world economy, the Era of Management is beginning to close, and will wind down over the next few decades. I’ve spent a lot of time thinking about these issues, and I’ve come to a few conclusions:

  1. The quality gap between the products of managed work and unmanaged work has reversed, with unmanaged work being superior by an increasing– at this point, impossible to ignore– amount. For one notable example, open-source software is now superior to gold-plated commercial tools. Creativity and motivation matter more than uniformity and control. This was not always the case, but it has become true and this trend is accelerating.
  2. This change is intrinsic and permanent. It is unnatural for people to manage or be managed, and the end of the managerial era is a return to a more natural motivational framework.
  3. Approaches to business that once seemed radical, such as Valve‘s open allocation policy, will soon enough be established as the only reasonable option. Starting with top technical companies, an with the trend later moving into a wide variety of industries, firms will discard traditional management in favor of intrinsic motivation as a means of getting the best quality of work from their people.

What’s going on? I believe that there’s a simple explanation for all of this.

“We will kill them with math”

Consider payoff curves for two model tasks, each as a function of the performance of the person completing it.

Performance | A Payoff | B Payoff |

5 (Superb)  |      150 |      500 |
4 (Great)   |      148 |      300 |
3 (Good)    |      145 |      120 |
2 (Fair)    |      135 |       40 |
1 (Poor)    |      100 |       10 |
0 (Awful)   |       50 |        0 |

What might this model? Task A represents easy work for which an average (“fair”) player can achieve 90 percent of the highest potential output:135 points out of a possible 150. An employee achieving only 50 percent of that maximum is clearly failing, and will probably be replaced, and there won’t be much variation between the people who make the cut. Task B represents difficult work for which there’s much more upside, but for which the probability of success is low. Average performers contribute very little, while the difference between “good” and “superb” is large. Task B’s curve might be more applicable to high-yield R&D work, in which a person would be considered highly successful if she had success in even 30 percent of the projects she set out to do, but it increasingly applies to disciplines like computer programming, where insight, taste, and vision are worth far more than commodity code. What matters, mathematically, is that Task A’s input-output relationship flattens as performance improves, while Task B’s accelerates. Task A’s curve is concave and Task B’s is convex. For Task A, the difference in return between an excellent and an average performer is minimal, but for Task B, it’s immense.

Does excellence matter? At most jobs, the answer has traditionally been “no”. At least, it has mattered far less than uniformity, reliability, and cost reduction. The concave behavior of Task A is more appropriate to most jobs than a convex one, and that’s largely by design. The problem with creative excellence is that it’s intermittent. Creativity can’t be managed into existence, while reliable mediocrity can be. As much as we might want managers to “nurture creativity”, the fact is that they work for companies, not subordinate employees, and their job is largely to limit risk. If we expect managers to do anything different, we’re being unreasonable. For Task A, performance-middling behaviors like micromanagement are highly appropriate, because bringing the slackers into line provides much more benefit than is lost by irritating high performers, and most industrial work that humans have performed, over our history, has been more like A than B. Getting the work done has mattered more than doing it well.

One of the interesting differences between concave and convex work is the relationship between expectancy (average performance) and variance. For traditional concave work, there’s a lot of variation at the low-performing end of the curve, but very little among high performers. To consider variance uniformly bad, therefore, will not be detrimental, the upside of variation being so minimal. Managerial activities that reduce variance are generally beneficial under such a regime. Even if high performers are constrained, this is offset by the improved productivity of the slackers. For convex work, the opposite is true. In a convex world, variation and expectancy are positively correlated. It turns out to be much easier, for a manager, to control variance than it is to improve expectancy. For this reason, almost everything in the discipline of “management” that has formed over the past hundred years has been focused on risk reduction. In a concave world, that worked. Reducing variance, while it might regress individual performances into mediocrity, would nonetheless bring the aggregate team performance up to a level where no one could reliably do better with comparable inputs. For most of industrial humanity’s history, that was enough.

Variance reduction falls flat in the convex world. Managerial pressures that bring individual performance to the middle don’t guarantee that a company has an “average” number of high-performing people, but make it likely that the firm has zero such people, and the result of such mediocrity is an end to innovation. In the short term, this damage is invisible, but in the long term, it renders the company unable to compete. Its prominence and market share will be snapped up in small pieces by smaller, more agile, companies until nothing is left for it but dominance over low-margin “commodity” work. Contrary to the typical depiction of large corporate behemoths being sunk wholesale by a startup “<X> killer”, what actually tends to happen is a gradual erosion of that company’s dominance as new entrants compete against it for something more important, in the long run, than market share: talent. Talent is naturally attracted to convex, risk-friendly work environments.

For a digression into applied mathematics– specifically, optimization– I would like to point out that since maximizing a concave function (such as bulk productivity) is equivalent to minimizing a convex one, we can think of management in the concave world as somewhat akin to a convex optimization problem. This is more of a metaphor than a true isomorphism, with one being abstract mathematics and the other rooted in human psychology, but I think the metaphor’s quite useful. I’ll gloss over a lot of detail and just say this: convex optimizations (again, akin to management of concave work) are easier. A convex minimization problem is like finding the bottom of a bowl (follow gravity, or the gradient). However, if the problem is non-convex, the surface might be more convoluted, with local valleys, and one might end up in a suboptimal place (local minimum) from which no incremental improvement is possible. The first category of problem can be solved using an algorithm called gradient descent: start somewhere, and iterate by stepping in the direction that, locally, appears best. The second category of problem can’t be solved  by simple gradient descent. One can fall into a local optimum from which some sort of non-local insight (I’ll return to this, later) is required if one wants to improve.

Concave and convex work are, in kind, also sociologically different. When the work is concave, the optimization problem is (loosely speaking) convex, and the one stable equilibrium (or local optimum) is, roughly speaking, “fairness”. On average, you’ll get more if you focus your efforts on improving low performers (who will improve more quickly) than by making the best even better. A policy that often works is to standardize performance: figure out how many widgets people can produce, and develop a strategy for bringing as many people as possible to that level (and firing the few who can’t make it). Slackers are intimidated into acceptable mediocrity, incorrigible incompetents are fired, and the bulk of workers get exactly the amount of support they need to reliably hit their widget quota. It’s a “one-size-fits-all” approach that, while imperfect, has worked well for a wide variety of industrial work.

Management of convex work is, as it were, a distinctly non-convex optimization problem. It’s sociologically much more complicated, because while the concave world has a “fairness” equilibrium, convex work has multiple equilibria that are usually “unfair”. You end up with winners and losers, and the winners need to be paired with the best projects, roles and mentors, although one might argue that the winners “don’t need them” from a fairness perspective. For convex work, you don’t manage to the middle. The stars who get more support and better opportunities will improve faster, and the schlubs’ mediocrity (whether a result of inability or poor political position) will persist. The best strategy, for a managed company, would be to figure out who has “star” potential and invest heavily in them from the start, but the measurements involved (especially because people have such strong incentives to game them) are effectively impossible for most people to make, both for intrinsic and political reasons.

For convex work, excellence and creativity matter, and they can’t be forced into existence by giving orders. Additionally, the value produced in convex work is almost impossible to measure on a piece-rate basis. Achievements in concave work tend to be separable: one can determine exactly how much was accomplished in each hour, day, and week, so it’s easy to see when people are slacking off. Work that is separable by time is usually also separable by person: visible low performers can be removed, because the group’s performance is strictly additive of individual productivity. For convex work, this is nearly impossible. A person can seem nonproductive while generating ideas that lead to the next breakthrough– the archetypical “overnight success” that takes years– and a colleague who might not be publishing notable papers may still contribute to the group in an important, but invisible, way.

If your tools are traditional management tactics, then convex work is intractable, and management is often counterproductive. I think the best metaphor that I can come up with for managers and executives is “trading boredom”. There are many traders out there who could turn a profit if they stuck to what they knew well, but get bored with “grinding” and start to depart from their domains of competence, adding noise and room for mistakes, and burning up their winnings in the process. Poker players have the same problem: the game gets so boring (at 2000-3000 hours per year) that they start taking unwise risks. The 40-hour work week is so ingrained in modern people that there’s often a powerful guilt people face of feeling useless when there is no work for them to do (even if they achieve enough within 10 of those hours to “earn their keep”) and this often leads to counterproductive effort. This, I believe, explains 90 percent of managerial activity: messing with something that already works well, because watching the gauges gets boring. Whenever an executive comes up with a hare-brained HR policy that the company doesn’t need, trading boredom, and the need to still feel useful when there is no appropriate work to do, is the cause.

At concave work, this managerial “trading boredom” is a hassle that veteran workers, who have been doing the job for decades, learn to ignore. They already know how to do their jobs better than their bosses do, so they show enough compliance to keep management off their backs, but change little about what they’re actually doing unless there’s a legitimate reason for the change. They keep on working, and the function of the team or factory remains intact. For convex work, on the other hand, managerial meddling is utterly destabilizing. The pointless meetings and disruptions inflicted by overmanagement take an enormous toll. In a convex world where small differences in performance lead to large discrepancies in returns, spending 2 hours each week in pointless meetings isn’t going to reduce output by a mere 5 percent, as one might expect from a linear model (2 hours lost out of 40). It’s probably closer to 25 percent.

The corporate hierarchy: an analytical perspective.

The optimization metaphor above, I believe, explains certain functional reasons for the typical three-tiered corporate hierarchy, with executives, managers, and workers. The workers are just “inputs”– machines made of meat, with varying degrees of reliability and quality, and for which there exist well-studied psychological strategies for reducing variance in performance in order to impose as much uniformity as possible. A manager‘s job is to focus on a small region over which the optimization problem is convex (which implies that the work is concave) and perform the above-mentioned gradient descent, or to iterate step-wise toward a local optimum. The strategy is given to the manager from above, and his job is to drive execution error as close to zero as possible. As variance will, all else being equal, contribute to execution error, variance must be limited as well. The job of an executive is to have the non-local insight and knowledge required to find a global optimum rather than being stuck at a local one. Executives ask non-local “vision” questions like, “Should we make toothpaste or video games?” Managers figure out what it will take to get a group of people to produce 2 percent more toothpaste.

This hierarchy is becoming obsolete. Machines are now outperforming us at the mechanical work that defined the bottom of the traditional, three-tier hierarchy. They are far more reliable and uniform than we could ever hope to be, and they excel at menial, concave work. We can’t hope to compete with them on this; we’ll need to let them have this one. So the bottom of the three tiers is being replaced outright. In addition, specialization has created a world where there is no place for mediocrity, and, therefore, in which the individual “worker” is now responsible for finding a direction of development (a non-local, executive objective) and planning her own path to excellence (a managerial task). The most effective people have figured this out by now and become “self-executive”, which means that they take responsibility for their own advancement, and prioritize it over stated job objectives. As far as they’re concerned, their real job is to become excellent at something, and they will focus on their career rather than their immediate job responsibilities, which they perform only because it helps their career to do so. As far as self-executive people are concerned, their employers are just along for the ride– funding them, and capturing some the byproducts they generate along the way, while they work mostly on their real job: becoming really good at something, and advancing their career.

Self-executive employees are a nightmare for traditional managers. They favor their career growth over their at-moment responsibilities, have no respect for the transient managerial authority that might be used to compel them to depart from their interests, yet tend at the same time to be visibly highly competent, which means that firing them is a political mess. They’re the easiest to fire from an HR “cover your ass” perspective (they won’t sue if you fire them, because they’ll quickly get an external promotion) but the damage to morale in losing them is substantial. In concave work, a team could lose its most productive member with minimal disruption; at convex work, such a loss is crippling. Managers who want to peaceably remove such people have to make it look like something the group wanted, and so they create divisions between the self-executive and colleagues– perhaps by setting unrealistic deadlines and then citing the self-executive person’s extracurricular education as a cause for slippage– but these campaigns are disastrous for group performance in the long run.

From a corporate perspective, a self-executive employee is the opposite of a “team player” and possibly even a sociopath, but I prefer to call the self-executive attitude adaptive. What point is there in being a “team player” when that “team” will be a different set of people in 36 months, and where one can be discarded from the team at any time, often unilaterally by a non-productive player who’s not even a real part of it? None that I see. The “team player” ethic is for chumps who haven’t figured it out yet. Additionally, because the working world is increasingly convex, self-executive people are increasingly good (if chaotic good, to use a role-playing analogy) for society. They sometimes annoy their bosses, but they become extremely competent in the process and, in the long term, they will advance the world far more than anyone can do by following orders. Self-executives tend to “steal an education” from their bosses and companies, but twenty years later, they’re building superior companies.

Self-executive employees want to take risks. They want to tackle hard problems, so they get better at what they do. While managers want to reduce variance, almost obsessively, self-executives want to increase it. Also relevant is the fact that managerial fictions about intrinsic “A”, “B”, and “C” players don’t exist. Stack-ranking– the annual “low performer” witch hunt that companies engage in to scare their middling crowd– doesn’t actually do much good in personnel selection. (It excels at intimidation, which is performance-middling and thus reduces variance, but the desirability of this effect is rapidly declining.) What does exist, and seems to be intrinsic, is that there are low- and high-variance people. Low-variance workers can kick out an acceptable performance under almost any circumstances– long hours, poor health, boring work, difficult or even aggressive clients and managers. They’re reliable, not especially creative, and tend to do well at the war of attrition known as the “corporate ladder”. They make lousy executives, but are most likely to be selected for those sorts of roles. High-variance people, on the other hand, are much more creative, and tend to be self-executive, but are much less reliable in the context of managed work. Their level of output is very high if measured over a long enough timeframe, but impossible to read at the level of a single day, or even a quarter. This distinction of variance, much more than the A- and B-player junk science, seems to be intrinsic or, at the least, very slow to change for specific individuals. Unfortunately, traditional managers and high-variance individuals are natural enemies. Low-variance people tend to be selected for management positions, are easiest to manage, and (most importantly) are less likely to make their bosses insecure.

What is changing

Why is work moving from concavity to convexity in output? There are a few answers, all tightly connected. The first of these is that concave work tends to have a defined maximum value: there’s one right way to perform the task. If we can define a target, we can structure the task as a computation, and it can be automated. Machines win, no contest, at reliability as well as cost-reduction. They’re ideal workers. They never complain, work 168-hour weeks, and don’t have hidden career objectives that they place at a higher priority than what they’re asked to do. As we get better at programming machines to perform the concave work, it leaves us with the convex stuff.

Second, the technological economy enhances potential individual productivity. The best programmers deliver tens of millions of dollars per year in business value, while the worst should probably not be employed at all. The capacity to have multiplier effects across a company, rather than the additive impact of a mere workhorse, is no longer the domain of managers only. The best software architects and engineers are also multipliers, because their contributions become infrastructural in nature. I don’t think that this potential for multiplicative impact is limited to software, either. As software becomes more capable of eliminating menial tasks from peoples’ days, there’s more time available for the high-yield, high-risk endeavors at which machines do poorly. What this enables is the potential for rapid growth.

When studying finance, one often learns that high rates of growth (8% per year) in a portfolio are “unsustainable”, because anything that grows so fast will eventually “outgrow” the entire world economy, which grows at only 3 to 5 percent, as if that latter rate were an immutable maximum. This might also apply to the 10-15% per year salary growth that young people expect in their careers– also unsustainable. Wall Street (in terms of compensation) has been called “a bubble” for this reason: even average bankers experience this kind of exponential compensation growth well into middle age, and it seems that this is unreasonable, because even small-scale economies or subsectors “can’t” grow that fast, so how can a person? Can someone actually increase the business value of his knowledge and capability by 15% per year, for 40 years? It seems that there “must be” some limiting factor. I no longer believe this to be necessarily true at our current scale. (There are physical, information-theoretic upper limits to economic prosperity, but we’re so far from those that we can ignore them in the context of our natural lifespans.) Certainly, rapid growth becomes harder to maintain at scale; that is empirically true. But who says that world economic growth can’t some day reach 10% (or 50%) per year and continue at such a rate until we reach a physical maximum (far past a “post-scarcity” level at which we stop caring)? Before 1750, growth at a rate higher than 0.5% per year would have been considered impossible: 0.1 to 0.2 percent, in agrarian times, was typical. If we view our entire history in stages– evolutionary, early human, agricultural, industrial– we observe that growth rates improve at each stage. It’s faster-than-exponential. I don’t believe in a single point of nearly-infinite growth– a “Singularity”– but I think that human development is more likely than not to accelerate for the foreseeable future. In the technological era, rapid improvements are increasingly possible. Whether this will result in rapid (30% per year) macroscopic economic growth I am not qualified to say, and I don’t think anyone has the long-term answer on that one, but we are certainly in a time when local improvements on that order are commonplace. Many startups consider user growth of 10% per month to be slow.

Rapid growth and process improvements require investment into convex work, which often lacks a short-term payoff but often provides an immense upside. It’s this kind of thinking that companies need if they wish to grow at technological rather than industrial rates, and traditional variance-reduction management is at odds with that. That said, traditional management is quite a strong force in corporate America. Most companies cannot even imagine how they would run their affairs without it. For sure, the managerial and executive elites won’t go gently into that good night. The private-sector career politicians who’ve spent decades mastering this inefficient, archaic, and often stupid system are not going to give up the position they’ve worked so hard to acquire. The macroscopic economic, social, and cultural benefits to a less-managed work world are extreme, but also irrelevant to the current masters, who have a personal desire to keep their dominance over other humans. The people in charge of the current system would rather reign in hell than serve in heaven. So what will give?

There won’t be an “extinction event” for managerial dinosaurs and the numerous corporations that have adopted their mentality, so much as an inability to compete. First, consider the superior quality of open-source software over commercial alternatives for an expanding set of software. That’s indicative. Open-source projects grow organically because people value them and willingly contribute, with no managers (in the industrial-era sense) needed. Commercial products die unless their owners continue to throw money at them (and sometimes even then). Open-source contributors are intrinsically motivated to be invested in the quality of their software. They’re often users of the product, and they can also improve their careers by gaining visibility in the wider software world. They have real, technological-era, self-executive motivations for wanting to do good work. For a contrast, most commercial software products are completed at a standard of “just good enough” to appease a boss and remain in good standing. It’s software written for managers, but from a product-quality standpoint, bosses themselves rarely matter. Users do. The quality gap between non-managed work and managed work is becoming so severe that the value of managed work is (albeit slowly) declining, out-competed by superior alternatives. This is bringing us to a state where “radical” cultures such as Valve’s purportedly manager-free open allocation policy become the only acceptable option. I would be shocked, in 30 years, if open allocation weren’t the norm at leading technology companies.

The truth is that managed work and variance reduction, which made the Industrial Revolution possible, are capable of producing growth at industrial (1 to 5 percent per year) rates, but not at technological rates (and a venture-funded startup must grow at technological rates or it will die). Compared to the baseline agrarian growth rate (0.05 to 0.3% per year) of antiquity, the industrial rate was rapid. Traditional management still works just fine, if your job is to turn $1.00 into $1.03 in twelve months. If you’re already rich and looking to generate some income from your massive supply of capital, this might continue to work for you indefinitely. If you’re poor, or looking to compete in the most exciting industries, and you need to unlock the energies that turn $1.00 into $2.00, you need something different.

Does this mean that there will no longer be no role for managers? It depends on how “manager” is defined. Leadership, communication, and mentorship skills will always be in high demand. In fact, the increasing complexity of technology will put education at a premium, and the few people who can lead groups of self-executive workers are becoming immensely valuable. Although the most talented workers will evolve into self-executive free agents, they will need some way of learning what efforts are worth their time, and they’ll be learning this from other people. Some aspects of “management” will always be important, but to the extent that management lives on, it will have to be about genuine leadership rather than authority.

Fossil fools

What has a one-way ticket to the tarpit (and almost no one will miss it) is the contemporary institution of corporate management: the Bill Lumbergh, who uses authority by executive endowment to compensate for his complete lack of leadership skills.

Leaders are chosen by a group that decides to be led, whereas corporate managers are puppet governors selected by external forces (or “from above”) as a means of exerting control. They don’t have to have any leadership skill, because the people being led have no choice. They’re hand-picked by their bosses: higher-level managers and executives. Leadership often requires handling trade-offs of peoples’ interests in a fair way, but that’s impossible for a corporate manager to do. Executives will never select a manager who would support the workers’ interests at an equal level to their own. Managers play a variety of roles, but their main one is to be a buffer between two groups of people (executives and workers) who would otherwise be at opposition because, even if for no other reason, they get dramatically different proportions of the reward. Managers legitimize executives by creating the impression that the separation between the company’s real members and its workers is a continuous spectrum (and thereby support the company’s efforts to mislead people regarding their true chances of upward mobility) rather than a discrete chasm, but they also form, because of their own increasingly divergent interests, a weak link that is increasingly problematic.

The corporate structure is effectively feudal. Just as medieval kings never cared how the dukes and earls treated their peasants, as long as tributes were paid, managers generally have unilateral power over the managed (as long as they don’t get the company sued). Managers are trusted to execute the corporate interest. It seems like this should create a weak link, giving managers the power to force workers to suit the manager’s career goals rather than the corporate objective. Perhaps surprisingly, though, in a concave world this doesn’t cause a major problem for the company. Managerial and corporate interests, at concave work, are aligned for reasons of sociological coincidence.

Managers have high social status and status is positively autocorrelated in time (that is, high status tends to reinforce itself) so a manager will “drift” into higher position as the group evolves, so long as nothing embarrasses him. Workers have to prove themselves in order to stand above the masses, but managers can coast and acquire seniority. In other words, a manager’s career is optimized not when he maximizes group productivity (which may be impossible to measure) but when he minimizes risk of embarrassment. A subordinate who breaks rules, no matter how trivial, embarrasses the manager– even if he’s highly productive. It’s better, from a manager’s career perspective, to have ten thoroughly average reports than nine good ones and one visible problem employee. Great employees make themselves look good, while bad ones are taken to reflect on the manager who failed to keep them in line. The consequence of this is that managers are encouraged toward risk-reduction. In a concave world, this is exactly what the company needs. So what happens in a convex world?

The convex world is different. If a company “gets” convexity (which is rare) it will begin to make allowances for individual contributors to allocate time to high-variance, high-reward activities which are often self-directed. This gives workers the opportunity to achieve visible high performance, and it’s good for the (expected) corporate profit, but managers lose out, because the worker who hits a home run will get most of the credit, not the manager. They find their subordinates increasingly interested in “side projects” and extra-hierarchical pursuits that “distract” them from their assigned work. There’s a conflict of interest that emerges between what the worker perceives as managerial mediocrity and the quest for the larger-scale excellence that can exist in convex pursuits. Because self-executive workers think non-locally (extra-hierarchical collaboration, self-directed career advancement) the appearance is that they’re jumping rank.

In the concave world, managers were the tactical muscle of the company. They drove the workers toward the local optimum in the neighborhood that the globally-oriented executives chose. In the convex world, managers are pesky middlemen. If they operate according to self-interest (and it’s unreasonable to expect otherwise of anyone) then their best bet is to use their authority to coerce their reports to prioritize the manager’s own career goals, which have now diverged from the larger objective. In other words, they become extortionists. That’s not a business that will live for much longer.

What seems clear is that middle management will decline in power and importance over the next 50 years. Increasingly convex work landscapes will decrease the use for it, and people will have less desire to fill such positions (which, in concave work, are coveted). What’s less clear is what will replace it. If this corporate middle class disappears within the current framework, what’s left is a two-tier system with workers and executives. That’s a problem. Two-class societies are extremely unstable, so I don’t think that arrangement will thrive. What’s more likely, in my (perhaps overly optimistic) opinion is that the functions of workers, managers, and executives will all blend together as individuals becomes increasingly self-executive. In many ways, this is a desirable outcome. However, it dramatically changes the styles of business that people will be able to form, making companies fairer and more creative, but also more chaotic and probably smaller. If the result of this is a macroscopic increase in creativity, progress, and individual freedom over one’s work, then a truly technological era might begin.