I think Paul is over-generalizing. Young, fast-developing industries tend to produce rich ecosystems of startups due to low cost of entry. But over time, as industries mature, companies are naturally starting to consolidate into larger and larger entities in order to survive. Paul thinks that just because we're seeing this early growth in web software, it applies to all economic activities of mankind.
Using Pauls's main arguments (historical perspective) it's not hard to "prove" the opposite: world is constantly moving towards consolidation.
Look: Desktop software development world looked very much like web scene today: one person could invent (and implement) an electronic table or an editor or a basic interpreter or an interesting game and do very well financially (has been done thousands of times). Yet in the early-mid 90s most software got prohibitively expensive to build for a small firm.
Also you can go back to early automotive boom in the US: there were myriads of automotive startups in mid-west (and in Europe too) and look what happened to all of them later. Same can be said regarding telecom, oil and railroad industries.
Developing industry = more startups
Mature industry = very few startups
I would expect an essay talking about smaller is better to discuss two things I didn't see there (although I'm pretty tired from having driven through the Alps, again):
* Coase and lower transaction costs. I think this generally favors pg's argument.
* Capital requirements for new companies. With internet startups, these are extremely low. Is this an anomaly that will disappear soon? Is it a new trend that will spread to other fields? Is anything else that has 'gone before' comparable to this field's low capital requirements?
I think Paul's point is based on the fact that new technologies are coming along faster, and thus new industries are being born faster than ever before.
And as you say, "Young, fast-developing industries tend to produce rich ecosystems of startups".
These two points imply that the trend is to have many young industries (many more than there are old), and thus many many more startups than big companies.
Further, he mentions how new technologies are game changers. Big companies cannot adapt as fast as smaller ones, and therefore their "death-rate" will be higher.
At the end of the day, I think you're not taking into account the fact that the rate of technological innovation is accelerating. This fact is going to be a large cause of much social and economic disruption (a good thing).
I think the other piece that's missing is that the role of corporations in society has changed drastically. I don't think they're fundamentally for the same thing as startups in the modern world. Paul's essay mentions large companies being less successful in the immediate future, but I don't think they're playing by the same rules, fundamentally.
Perhaps the closest example I can think of here is the transformation that happened in both Christianity and Islam with the rises of the Catholic church and the Caliphate, respectively. At a certain point their religious function became incidental. They were by and large instruments which brokered in power and influence.
You can see a similar pattern in governments founded through populist revolutions based on ideologies as they've transformed into world powers. Liberty, as such, has for quite a while ceased to be the primary function of the US or French governments.
In the same sense, I believe that the modern trans-national corporation has ceased to be an entity that exists primarily for the propagation of products and services: they are also brokers of power and influence. There was a first colonial phase starting at the middle of the last century where corporations became powerful political influence outside of their own home territories and since the 70s or 80s they've extended to becoming powerful forces even within the machines of the modern super-powers. Drawing another analogy, what I believe the modern renewal of small companies is effectively a reformation akin to what's been seen in religious institutions, where there's a sense of things getting back to the ideals. I feel like it's still too early in history to see if this will be a fundamentally disruptive change, or merely a blip in the growth pattern. Again, history provides some examples there for example in the Wycliffe-ian or Hussite movements, which preceded the more dramatic changes of the reformation and counter-reformation (and before those movements themselves began creeping towards instruments of power). The real reformation may still be yet to come.
In addition to this trend in individual industries, there's also an underlying trend across industries. The emergence of organized VC firms to supply funding for new startups (regardless of industry) is evidence for it. The funding of new companies used to be more haphazard, because it was a comparatively rare event. Starting in the 1960s it became more common, and more deliberate.
When I read the sentence "An ambitious kid graduating from college now doesn't want to work for a big company.", I stopped reading. Over-generalizations doesn't work.
Most of creative work at 'big companies' involves designing business processes. Which is in fact programming, only using people and resources. Another task involves optimizing the processes and minimizing the associated cost.
The whole raison d'être of a large organizational pyramid is cost of processing and transmitting information and handling the associated noise. Middle layers of management filtering and processing information for the upper decision-makers. Keeping the noise down by designing company 'policies', 'strategies' and 'missions' that can be easily communicated to every employee.
A manager who designs the most efficient and scalable organization structure wins over the competition or conquers a new market.
Here comes the internet with zero cost of communication. Here come high level languages that make it cheap to describe the most efficient processes.
Any large organization can in principle be substituted by a code because the organization itself IS a code written in job description language.
I think this is a great point, but I have to disagree with two parts.
Here comes the internet with zero cost of communication - information still costs time (opportunity cost) to produce and consume. It can be designed to be easier to produce/consume, but that costs design resources. The Internet gives zero cost to transmit information.
Here come high level languages that make it cheap to describe the most efficient processes. - this won't happen because people are not interchangeable like Silicon is. You can hire two people that fit the same job description but will fit differently into the organization. They will work differently, interact differently with management and coworkers, produce different results (even if equivalent).
I think the closest we can come to high level business languages are
1) something like design patterns that can't be plugged in directly but can guide implementation of a business process with known trade-offs OR
2) design processes that can be completely specified and automated so human judgement isn't involved
Interestingly, #1 favors talent cultivation (a la Google) and the second leads to outsourcing.
If programming is analogous to planning business processes, you won't be able to make a "high-level language" for it. High-level languages make the planning easy at the cost of making the process itself inefficient, which is the opposite of what you want.
It's amazing how quickly things change. Paul mentioned the 70s as a heyday for big organizations, but I remember this going well into the 80s. You can see it in the culture. Remember how many movies were made that glorified climbing the corporate ladder? "Working Girl" with Meg Ryan, "The Secret of My Success" with Michael J Fox, "Big" with Tom Hanks. Those are just a few. There was an entire hollywood genre of climbing the ladder - movies that made it seem that making presentations with little arrows going up and down on a chart was the apex of human achievement.
I was in my teens during the latter half of this decade - I'd guess that at least half of the people who read hacker news are too young to have seen them when they came out. If you do watch them now, they'll probably seem as quaint as "Hair".
Ever notice any 80s movie characters who were computer programmers? I can remember one - it's the weirdest thing to watch. He's socially awkward, which is reasonable, but he also has no money! All the cool kids in the movie make more than him. You could not have such a character today - completely implausible.
You're right. I remember when those movies came out. Even 1991's "Don't tell Mom the Babysitter's Dead" follows that pattern.
Ironically, Tron was this way also. Flynn sought to run the big tech corporation instead of that arcade. Now, I'd rather be running that arcade than working for the big corporation.
How often does it happen that a rule works for thousands of years, then switches polarity?
I'm not a historian, but as far as I know from reading various books by Drucker, large corporations didn't exist for much of the last thousand years. In fact, I remember he made the point that looking at the turn of the last century, even the largest business of that day would be considered small to medium by today's standards (and goes on to ascribe that to a lack of management knowledge - it's impossible to manage a huge corporation without management).
If "larger is better" has only been with us for a hundred years, it may well slink back into the shadows sooner than we think.
I think the order in which this will unfold will be highly dependent on what kind of business you're in. Car manufacturers and pharmaceutical companies are unlikely to get small any time soon, at least not until we have perfect simulated models of the human body to get rid of all the human trials, and until we have instant, near-free manufacturing processes available to all. By the time those happen, it seems unlikely that "corporations" will look like they do now....
I don't think 'corporation' is a required title for the concept to apply. Large command-oriented economic drivers have existed for a long time. The Roman army was one, it's economic function the looting of conquered territory. Notice that when it could no longer accomplish that function, things began to turn south for Rome.
We are observing the decentralization of economic authority, common throughout history, but usually a side-effect of economic decline and/or societal collapse; maybe more correctly a side-effect of the movement from specialized occupations back to subsistence farming.
This decentralization seems to be different, and appears to be driven by the unprecedented communication, access to information, and wide audience granted by the Internet.
If you mean a "join-stock company", large corporations are probably less than a thousand years old. The first were alledgely shipping combinations organized in London and Dutch coffeehouses, but I have also read suggestions that Medieval Japan and China may have had something like a modern joint-stock corporation.
But what Graham means, is corporation in it's true sense of the word as a generic "body". The Catholic Church, Roman and Ottoman Empires, Hanseatic League, various hybrid religous orders such as the Teutonic Order and etc, can all be considered corporations. And in general, larger was better, with occasional shifts and retreats.
I think car manufacturing can be done by much smaller organizations. I suspect that pharmaceutical companies will have to change a lot; for all they spend on research, they simply don't prolong human life enough to justify being 14% of the economy.
There are lots of small pharmaceutical companies already; they are a major focus of current VC funding, in fact. The majority of "biotech" firms are actually working on pharmaceuticals, not food crops or spider-goats or biological weapons.
"Now it turns out the rule 'large and disciplined organizations win' needs to have a qualification appended: 'at games that change slowly.'"
The pattern I see throughout history is that large organizations endure until some disruptive technology takes them down. (Greek hoplite armies vs. light infantry and archers, armored knights vs. longbows, cavalry corps vs. tanks.)
The modern age has simply accelerated the whole process. Technological change occurs increasingly quickly now, and so these "overturnings" happen more and more often. Because large organizations take time to form, the balance of power has shifted to small entrepreneurial organizations.
true, you might want to look at empires that lasts for centuries after rapid expansion
alexander the great, napoleon, the japanese (1940ish) don't count because they cannot sustain the expansion
ironically, mongol empire (and its vast expansion) lasted for 3.5 centuries. there must be some other reasons that neither cruelty nor barbarism can explain (in fact, empires based on cruelty/barbarism don't last long)
Perhaps it's an east coast thing but I was just back for Thanksgiving and of the many Ivy League superstars I grew up with, almost all are working on Wall St, in management consulting shops, or they have become lawyers. I'm not saying that Ivy League == best and brightest, but it is probably a good indicator of ambition.
These three industries all have one thing in common, they pay well and they only require a Liberal Arts degree. My sister graduated from Harvard a few years back and I'd say that every single one of her friends I've met works in one of these three industries. Their conversations of others I haven't met indicate that these three professions are the norm for all their classmates.
Perhaps Stanford is different and perhaps CS majors are as well. But the major trend I see is that the most ambitious people from the most prestigious Universities are going to the places that pay them the most.
the trend i see is that the most ambitious people do startup and degree(s) matter less as time goes
maybe a quantitative measure like Z = asset_from_work_now / cost_before_work_then where cost_before_work_then include tuition, rent, etc ... work can be defined as wall street, lawyer, plumber, startup, etc
Z for googlers can be in the thousands to millions, while Z for wall streets might be in the tens or hundreds ... big difference
or course Z can be negative (bankrupt+debt) ... but the huge discrepancy of Z value only reinforces my point, only the brave and ambitious dare to walk the uncertain path; the rest only follow thru the proven, safe path
Part of the reason—possibly the main reason—that startups have not spread as broadly as the Industrial Revolution did is their social disruptiveness.
I think a more abstracted version of the problem would be that, at least in western civilization, products spread much faster than best practices. If Nintendo releases a new videogame system they'll sell millions the first day, and yet 2/3 of American children still don't get their daily RDI of calcium.
The problem is that processes are much more important than products in terms of quality of life, health care, child development, business, education, etc. And while technological progress is increasing exponentially, the rate of best-practice adoption remains flat.
It seems like the American dream is being able to purchase the solution to any one of life’s problems in a big f*ing box at Wal-Mart for less than 200 bucks. Consumers are already really well trained at this, and we need to figure out a way to leverage this behavior to get them to adopt new processes. The problem as I see it is threefold:
1) People don't know about best practices
2) They know about a best practice but they aren't sold on it
3) They're sold on it but it's too hard
So far as I can see it, the only way for America to remain competitive with the rest of the world is to make the adoption of best practices an order of magnitude easier. We might be temporarily ahead of everyone else in terms of getting the smartest kids to go out and join or start their own businesses, but in terms of nearly every other best practice we're falling drastically behind.
The reality is that we live in a society where doctors don't wash their hands before surgery, where shoddy farm practices cause excessive soil erosion, where pregnant women eat fish high in mercury and PCBs, etc. America is certainly getting more high-res in response to change, but is the catalyzing change really technological? Corporations are great at buying stuff, it's the best practices they are really slow at adopting. The fact that corporations are breaking down into smaller units seems to be a synechdoche of society at large falling apart due to the same set of failures.
I think a major reason for the high-res society as Pg so eloquently calls it, can be found in Ronald Coase's classic The theory of the firm from 1937.
In it he tries to describe firms and why they exist. His main finding is that the reason that firms exist as large entities instead of just individuals trading labour and services with each other is transaction costs. In an industrial society the transaction costs of many individuals coming together to create and sell a product would be prohibitive. By creating firms the transaction costs are lowered - it is cheaper to create and maintain the structure needed to uphold the firm than it is to locate and purchase services in the marketplace when you need them .
The transaction costs are coming down with the advent of the global information flow - it is much easier for me to locate and purchase the services of an individual coder than it would be to locate and purchase the services of a welder 20 years ago. The transaction costs are especially low in industries that sell immaterial services and goods that can easily be moved around the globe, such as web businesses.
This makes firms increasingly less competitive with individuals, it also explains why some sectors such as the auto and aero industries are dominated by big companies and probably will be in the future. The transaction costs of individuals building a Boeing 747 are just too prohibitive.
Isn't there some danger that you are over-generalizing from an industry where there are exceptionally low barriers to entry to a global trend? What if you want to make a new automobile, or even worse, a new airliner?
However, I think there is another trend that reinforces what you say here. Previously there was a strong incentive for risk-averse people to work in large corporations. I'm a risk-averse person and I once worked at a large corporation. With the change in organizational behavior so that layoffs have become a first resort action, instead of an action only appropriate in desperate circumstances, there is little risk-aversion benefit to large organizations. So other than health insurance, large corporations have become a generally inferior class of employers. This is true even if you are not entrepreneurial.
I believe that as a general trend, we are in a period of time where the large organization is collapsing, across wide areas of human activity.
Without writing a thesis on it, I offer the following disconnected observations:
* The US Military, one of the largest human organizations, was successfully attacked at it's headquarters for a total expenditure of less than $400,000 and a few lives, and has not and may never catch the attacker
* The Roman Catholic Church's sex scandals
* To be fair to all religious factions, punch in "baptist financial scandal" into google, or any other large religous denomination
* The disappearence of any sort of financial security in corporate employment or retirement
* The fact that there is no "corporate ladder" that PG refers to, and it has likely existed only in people's hopes and fantasies for two decades or so. The way to advance in most corporations is to have the corporation grow beneath you, or to leave for another corporation and come back to a higher position. Promotions and raises inside one corporation do not meet inflation when averaged over all employees. Loyalty is punished, almost as if in some sense the organization knows it is a bad thing and is trying to kill itself.
* The collapse of joint financial organizations that were once emmensely powerful and efficient:
* We used to all put premiums into large insurance corporations, and those insurance corporation used their economy of scale to invest those premiums, such that they paid out something like 109% of premiums as claims and still profited. Today most big insurance corporations pay out less in claims than what they take in via premiums.
* We used to all put money into banks, which then made loans and paid interest to the depositors. Now we don't save, and as of a few months ago, banks don't lend.
* The US Federal Government, arguably the largest human corporation, is bankrupt and disfunctional -- it was never a model of efficiency, but in times past it got some things done, built dams and won the wars it fought and etc. Now it does nothing productive at all.
* IBM, Microsoft, GE, GM, Exxon, etc -- the readers here don't need to be told of their failures, but I would additionally point out the trend that they don't want new additions to their organization, and attempt to hire mainly contract, outsourced, non-employee employees these days.
* The United Nations
One thing to note, is that I think if the trend is away from big organizations, the trend of integrating Europe into one big EU is a bad one. The trend of increasing Federalization in the United States is probably also generally the wrong direction.
You need to separate the stability of individual large organizations over time to the relative importance of all large organizations. The Mormon faith is growing and other religions are dieing, but on average about the same number of people are part of some religion. Google grows as GM dies, but the average number of people working at large 5,000+ person companies seems fairly constant. Individual government organizations grow and die but overall the government workforce seems fairly constant. Companies with less than 5 people are a wold apart from company's with 50 but after 5,000 it's all about the same mess.
PS: The Pentagon is just an office building NORAD is far more important.
I think the trend here was missed. The fast change is not a new event but rather a continuous trend that moved laterally into business.
It's decentralization, and it's been occuring throughout history. For example, equality is a form of decentralization that has taken quite a while to manifest. You've also got decentralization in information, labor (specialization), and housing.
The next one that we desperately need is energy decentralization: we need to have smaller, agile, and more incremental methods of generating energy. The tech is coming along so we'll be able to do this, and some of the tech will only work well if it's not scalable.
- The Internet is different from other fields because of low barriers to entry and the ease of collaboration between groups (low transaction costs). This means that even if our field grows very large (and I think it will), the principles underlying its success might not successfully spread to other industries.
- Another thing that might happen is that large organizations gain enough power that they can control the Internet. If they're able to do so and prevent the rise of alternative, more free networks, then hegemony could come to dominate this industry too, as it did all the others.
I think the limiting factor to growth of the startup culture is the fact that big organizations are good at providing commodity goods and services and the freedom and unconventional nature of startups doesn't seem to scale up to big organizations. At some point I'm sure even companies like Google and Genentech, which tried to maintain the startup feeling, start to feel pretty corporate.
So I would agree insofar as to say that startups will become the de facto standard method of commercializing new technology, but I don't think they will ever represent a significant portion of the economy as far as the number of people employed. Having such a small percentage of the population involved in startups doesn't afford them much opportunity to affect social norms on a large scale.
Short of complete automation, big organizations will always be best at doing commodity things like picking up your trash and making your shirts. Even once those particular tasks are automated, big organizations will grow (often from startups) elsewhere to tackle new commodity goods and services. Once something is no longer cutting edge, it'll be provided by a big organization.
Startups will always disrupt the way big organizations currently do things, but by being successful they will just become big organizations themselves.
"If Internet startups offer the best opportunity for ambitious people, then a lot of ambitious people will start them, and this bit of the economy will balloon in the usual fractal way."
With the advent of open hardware and 3D printers, I think you can take out the word Internet and have the statement still be true.
The conclusion overreached a little bit. And there was something disturbingly California centric about it.
In my opinion there is not such a huge cultural gap that's responsible for the different startup rates.
Few (if any) cultures are so deeply authoritative as to be anti-success.
The valley's lead is much more the result of rather simpler causes. People - a whole lot of smart ones, two recession proof large smart people producers (the universities), and funding lots and lots of funding.
That's where the lead comes from and it will be difficult to close it. Culture doesn't have anything to do with it.
"Few (if any) cultures are so deeply authoritative as to be anti-success."
I would argue that most cultures, even in California, are too authoritarian and inflexible to really cultivate productive, disruptive change. For all the talk about innovation, it is genuinely difficult to argue against the status quo. It is very, very hard to change one's frame of mind, and people tend elsewhere to defer to expertise and experience over ideas with are promising, make sense, but are not so fleshed out. In the Bay Area you will at least find an audience, if not easily convince them. Being anti-authority is obviously insufficient for startup success, though.
A good quote to keep in mind:
"...I would design my own, fresh, without knowing how other people do it. That was another thing that made me very good. All the best things that I did at Apple came from (a) not having money and (b) not having done it before, ever. Every single thing that we came out with that was really great, I'd never once done that thing in my life." -- Steve Wozniak
We are currently encountering much existing knowledge in engine design. Some of it is undoubtedly wise, but a lot of it is probably out of date. We have the chance to experiment. But it would be genuinely impossible to do such a thing in most auto companies today.
There's a famous economics essay about the size of the firm by Ronald Coase (winners of the Nobel prize for economics), in which he discusses why firms (companies, organizations) exist at all (instead of individual contractors), and the factors influencing their size.
I dont quite understand if the article was to convince me about something or was it to predict the future?
"An ambitious kid graduating from college now doesn't want to work for a big company. They want to work for the hot startup that's rapidly growing into one."
Thats a generalization and I believe its untrue and biased towards talented, ambitious young people Paul meets. Sorry if I'm wrong, but if that was to convince me, I'd need a proof. From my point of view, the rest of the world is still at sort of peak of belief in the "corporate world" (but I dont have any research doc to reference as well :)).
"Fifty years later, startups are ubiquitous in Silicon Valley and common in a handful of other US cities, but they're still an anomaly in most of the world."
Well thats what I dont understand the most. If the whole article talks about the huge positive impact that startups are going to have on economy, let me ask Paul: what has changed since then ("fifty years [ago]") that makes you believe that NOW is the moment startups are going to change the whole world so much? What was the reason you wrote (published) the article NOW?
Large organisations are modelled on the military. So were schools. And many more countries had compulsory national service. Now schools are evolving and university with all its freedoms is more likely experience than military service for most new entrants to the workforce, it's not surprising that a huge, rigid, hierarchical structure is no longer relevant or even tolerable to many people.
Another point to bear in mind: more small companies means more outsourcing, particularly for legal, HR, IT support, and so on. We should also see masses more companies using shared business premises, with shared receptions and so on.
Small companies should also mean more teleworking, since (1) people in small companies are generally more open to the idea of it, and (2) it's easier to keep track of what everyone in a smaller company is doing, even if they are working remotely, and (3) it means small companies can have international representation which may be one guy in a specific Asian country working from a bedroom-office.
It's interesting to read this with Taleb's "The Black Swan" in mind (which I just reread this weekend), specifically because it made me think of Taleb's "barbell strategy":
"... your strategy is to be as hyperconservative and hyperaggressive as you can be instead of being mildly aggressive or conservative. Instead of putting your money in 'medium risk' investments, you need to put a portion, say 85 to 90 percent, in extremely safe instruments, like Treasury bills—as safe a class of instruments as you can manage to find on this planet. The remaining 10 to 15 percent you put in extremely speculative bets, as leveraged as possible (like options), preferably venture capital-style portfolios."
It seems that going to work for a startup is the exact opposite of this.
I.e., it would be better to be an investor in many startups rather than working solely on/for one startup.
[+] [-] old-gregg|17 years ago|reply
Using Pauls's main arguments (historical perspective) it's not hard to "prove" the opposite: world is constantly moving towards consolidation.
Look: Desktop software development world looked very much like web scene today: one person could invent (and implement) an electronic table or an editor or a basic interpreter or an interesting game and do very well financially (has been done thousands of times). Yet in the early-mid 90s most software got prohibitively expensive to build for a small firm.
Also you can go back to early automotive boom in the US: there were myriads of automotive startups in mid-west (and in Europe too) and look what happened to all of them later. Same can be said regarding telecom, oil and railroad industries.
Developing industry = more startups Mature industry = very few startups
Has always been like that.
[+] [-] davidw|17 years ago|reply
* Coase and lower transaction costs. I think this generally favors pg's argument.
* Capital requirements for new companies. With internet startups, these are extremely low. Is this an anomaly that will disappear soon? Is it a new trend that will spread to other fields? Is anything else that has 'gone before' comparable to this field's low capital requirements?
[+] [-] briprowe|17 years ago|reply
And as you say, "Young, fast-developing industries tend to produce rich ecosystems of startups".
These two points imply that the trend is to have many young industries (many more than there are old), and thus many many more startups than big companies.
Further, he mentions how new technologies are game changers. Big companies cannot adapt as fast as smaller ones, and therefore their "death-rate" will be higher.
At the end of the day, I think you're not taking into account the fact that the rate of technological innovation is accelerating. This fact is going to be a large cause of much social and economic disruption (a good thing).
[+] [-] wheels|17 years ago|reply
Perhaps the closest example I can think of here is the transformation that happened in both Christianity and Islam with the rises of the Catholic church and the Caliphate, respectively. At a certain point their religious function became incidental. They were by and large instruments which brokered in power and influence.
You can see a similar pattern in governments founded through populist revolutions based on ideologies as they've transformed into world powers. Liberty, as such, has for quite a while ceased to be the primary function of the US or French governments.
In the same sense, I believe that the modern trans-national corporation has ceased to be an entity that exists primarily for the propagation of products and services: they are also brokers of power and influence. There was a first colonial phase starting at the middle of the last century where corporations became powerful political influence outside of their own home territories and since the 70s or 80s they've extended to becoming powerful forces even within the machines of the modern super-powers. Drawing another analogy, what I believe the modern renewal of small companies is effectively a reformation akin to what's been seen in religious institutions, where there's a sense of things getting back to the ideals. I feel like it's still too early in history to see if this will be a fundamentally disruptive change, or merely a blip in the growth pattern. Again, history provides some examples there for example in the Wycliffe-ian or Hussite movements, which preceded the more dramatic changes of the reformation and counter-reformation (and before those movements themselves began creeping towards instruments of power). The real reformation may still be yet to come.
[+] [-] pg|17 years ago|reply
[+] [-] pg|17 years ago|reply
[+] [-] greyman|17 years ago|reply
[+] [-] asdf333|17 years ago|reply
Mature industry = chance for disruptive change = developing industry
[+] [-] 1gor|17 years ago|reply
The whole raison d'être of a large organizational pyramid is cost of processing and transmitting information and handling the associated noise. Middle layers of management filtering and processing information for the upper decision-makers. Keeping the noise down by designing company 'policies', 'strategies' and 'missions' that can be easily communicated to every employee.
A manager who designs the most efficient and scalable organization structure wins over the competition or conquers a new market.
Here comes the internet with zero cost of communication. Here come high level languages that make it cheap to describe the most efficient processes.
Any large organization can in principle be substituted by a code because the organization itself IS a code written in job description language.
[+] [-] pchristensen|17 years ago|reply
Here comes the internet with zero cost of communication - information still costs time (opportunity cost) to produce and consume. It can be designed to be easier to produce/consume, but that costs design resources. The Internet gives zero cost to transmit information.
Here come high level languages that make it cheap to describe the most efficient processes. - this won't happen because people are not interchangeable like Silicon is. You can hire two people that fit the same job description but will fit differently into the organization. They will work differently, interact differently with management and coworkers, produce different results (even if equivalent).
I think the closest we can come to high level business languages are
1) something like design patterns that can't be plugged in directly but can guide implementation of a business process with known trade-offs OR
2) design processes that can be completely specified and automated so human judgement isn't involved
Interestingly, #1 favors talent cultivation (a la Google) and the second leads to outsourcing.
[+] [-] marvin|17 years ago|reply
[+] [-] kds|17 years ago|reply
[+] [-] geebee|17 years ago|reply
I was in my teens during the latter half of this decade - I'd guess that at least half of the people who read hacker news are too young to have seen them when they came out. If you do watch them now, they'll probably seem as quaint as "Hair".
Don't ask.
[+] [-] herdrick|17 years ago|reply
[+] [-] jedberg|17 years ago|reply
http://www.imdb.com/title/tt0096463/
[+] [-] umjames|17 years ago|reply
Ironically, Tron was this way also. Flynn sought to run the big tech corporation instead of that arcade. Now, I'd rather be running that arcade than working for the big corporation.
[+] [-] swombat|17 years ago|reply
I'm not a historian, but as far as I know from reading various books by Drucker, large corporations didn't exist for much of the last thousand years. In fact, I remember he made the point that looking at the turn of the last century, even the largest business of that day would be considered small to medium by today's standards (and goes on to ascribe that to a lack of management knowledge - it's impossible to manage a huge corporation without management).
If "larger is better" has only been with us for a hundred years, it may well slink back into the shadows sooner than we think.
I think the order in which this will unfold will be highly dependent on what kind of business you're in. Car manufacturers and pharmaceutical companies are unlikely to get small any time soon, at least not until we have perfect simulated models of the human body to get rid of all the human trials, and until we have instant, near-free manufacturing processes available to all. By the time those happen, it seems unlikely that "corporations" will look like they do now....
[+] [-] gills|17 years ago|reply
We are observing the decentralization of economic authority, common throughout history, but usually a side-effect of economic decline and/or societal collapse; maybe more correctly a side-effect of the movement from specialized occupations back to subsistence farming.
This decentralization seems to be different, and appears to be driven by the unprecedented communication, access to information, and wide audience granted by the Internet.
[+] [-] RobGR|17 years ago|reply
But what Graham means, is corporation in it's true sense of the word as a generic "body". The Catholic Church, Roman and Ottoman Empires, Hanseatic League, various hybrid religous orders such as the Teutonic Order and etc, can all be considered corporations. And in general, larger was better, with occasional shifts and retreats.
I think car manufacturing can be done by much smaller organizations. I suspect that pharmaceutical companies will have to change a lot; for all they spend on research, they simply don't prolong human life enough to justify being 14% of the economy.
[+] [-] pg|17 years ago|reply
[+] [-] kragen|17 years ago|reply
[+] [-] nebula|17 years ago|reply
If you can't ship your products to large distances, you can't have large manufacturing facilities.
Without all these wonderful communication channels, you can't have your workforce distributed across the globe.
[+] [-] sdurkin|17 years ago|reply
The pattern I see throughout history is that large organizations endure until some disruptive technology takes them down. (Greek hoplite armies vs. light infantry and archers, armored knights vs. longbows, cavalry corps vs. tanks.)
The modern age has simply accelerated the whole process. Technological change occurs increasingly quickly now, and so these "overturnings" happen more and more often. Because large organizations take time to form, the balance of power has shifted to small entrepreneurial organizations.
[+] [-] hs|17 years ago|reply
alexander the great, napoleon, the japanese (1940ish) don't count because they cannot sustain the expansion
ironically, mongol empire (and its vast expansion) lasted for 3.5 centuries. there must be some other reasons that neither cruelty nor barbarism can explain (in fact, empires based on cruelty/barbarism don't last long)
[+] [-] uuilly|17 years ago|reply
These three industries all have one thing in common, they pay well and they only require a Liberal Arts degree. My sister graduated from Harvard a few years back and I'd say that every single one of her friends I've met works in one of these three industries. Their conversations of others I haven't met indicate that these three professions are the norm for all their classmates.
Perhaps Stanford is different and perhaps CS majors are as well. But the major trend I see is that the most ambitious people from the most prestigious Universities are going to the places that pay them the most.
[+] [-] hs|17 years ago|reply
maybe a quantitative measure like Z = asset_from_work_now / cost_before_work_then where cost_before_work_then include tuition, rent, etc ... work can be defined as wall street, lawyer, plumber, startup, etc
Z for googlers can be in the thousands to millions, while Z for wall streets might be in the tens or hundreds ... big difference
or course Z can be negative (bankrupt+debt) ... but the huge discrepancy of Z value only reinforces my point, only the brave and ambitious dare to walk the uncertain path; the rest only follow thru the proven, safe path
[+] [-] bokonist|17 years ago|reply
[+] [-] Alex3917|17 years ago|reply
I think a more abstracted version of the problem would be that, at least in western civilization, products spread much faster than best practices. If Nintendo releases a new videogame system they'll sell millions the first day, and yet 2/3 of American children still don't get their daily RDI of calcium.
The problem is that processes are much more important than products in terms of quality of life, health care, child development, business, education, etc. And while technological progress is increasing exponentially, the rate of best-practice adoption remains flat.
It seems like the American dream is being able to purchase the solution to any one of life’s problems in a big f*ing box at Wal-Mart for less than 200 bucks. Consumers are already really well trained at this, and we need to figure out a way to leverage this behavior to get them to adopt new processes. The problem as I see it is threefold:
1) People don't know about best practices
2) They know about a best practice but they aren't sold on it
3) They're sold on it but it's too hard
So far as I can see it, the only way for America to remain competitive with the rest of the world is to make the adoption of best practices an order of magnitude easier. We might be temporarily ahead of everyone else in terms of getting the smartest kids to go out and join or start their own businesses, but in terms of nearly every other best practice we're falling drastically behind.
The reality is that we live in a society where doctors don't wash their hands before surgery, where shoddy farm practices cause excessive soil erosion, where pregnant women eat fish high in mercury and PCBs, etc. America is certainly getting more high-res in response to change, but is the catalyzing change really technological? Corporations are great at buying stuff, it's the best practices they are really slow at adopting. The fact that corporations are breaking down into smaller units seems to be a synechdoche of society at large falling apart due to the same set of failures.
[+] [-] mixmax|17 years ago|reply
In it he tries to describe firms and why they exist. His main finding is that the reason that firms exist as large entities instead of just individuals trading labour and services with each other is transaction costs. In an industrial society the transaction costs of many individuals coming together to create and sell a product would be prohibitive. By creating firms the transaction costs are lowered - it is cheaper to create and maintain the structure needed to uphold the firm than it is to locate and purchase services in the marketplace when you need them .
The transaction costs are coming down with the advent of the global information flow - it is much easier for me to locate and purchase the services of an individual coder than it would be to locate and purchase the services of a welder 20 years ago. The transaction costs are especially low in industries that sell immaterial services and goods that can easily be moved around the globe, such as web businesses.
This makes firms increasingly less competitive with individuals, it also explains why some sectors such as the auto and aero industries are dominated by big companies and probably will be in the future. The transaction costs of individuals building a Boeing 747 are just too prohibitive.
Link: http://en.wikipedia.org/wiki/Theory_of_the_firm
[+] [-] izaidi|17 years ago|reply
[+] [-] pg|17 years ago|reply
[+] [-] rpgoldman|17 years ago|reply
However, I think there is another trend that reinforces what you say here. Previously there was a strong incentive for risk-averse people to work in large corporations. I'm a risk-averse person and I once worked at a large corporation. With the change in organizational behavior so that layoffs have become a first resort action, instead of an action only appropriate in desperate circumstances, there is little risk-aversion benefit to large organizations. So other than health insurance, large corporations have become a generally inferior class of employers. This is true even if you are not entrepreneurial.
[+] [-] RobGR|17 years ago|reply
Without writing a thesis on it, I offer the following disconnected observations:
* The US Military, one of the largest human organizations, was successfully attacked at it's headquarters for a total expenditure of less than $400,000 and a few lives, and has not and may never catch the attacker
* The Roman Catholic Church's sex scandals
* To be fair to all religious factions, punch in "baptist financial scandal" into google, or any other large religous denomination
* The disappearence of any sort of financial security in corporate employment or retirement
* The fact that there is no "corporate ladder" that PG refers to, and it has likely existed only in people's hopes and fantasies for two decades or so. The way to advance in most corporations is to have the corporation grow beneath you, or to leave for another corporation and come back to a higher position. Promotions and raises inside one corporation do not meet inflation when averaged over all employees. Loyalty is punished, almost as if in some sense the organization knows it is a bad thing and is trying to kill itself.
* The collapse of joint financial organizations that were once emmensely powerful and efficient: * We used to all put premiums into large insurance corporations, and those insurance corporation used their economy of scale to invest those premiums, such that they paid out something like 109% of premiums as claims and still profited. Today most big insurance corporations pay out less in claims than what they take in via premiums. * We used to all put money into banks, which then made loans and paid interest to the depositors. Now we don't save, and as of a few months ago, banks don't lend.
* The US Federal Government, arguably the largest human corporation, is bankrupt and disfunctional -- it was never a model of efficiency, but in times past it got some things done, built dams and won the wars it fought and etc. Now it does nothing productive at all.
* IBM, Microsoft, GE, GM, Exxon, etc -- the readers here don't need to be told of their failures, but I would additionally point out the trend that they don't want new additions to their organization, and attempt to hire mainly contract, outsourced, non-employee employees these days.
* The United Nations
One thing to note, is that I think if the trend is away from big organizations, the trend of integrating Europe into one big EU is a bad one. The trend of increasing Federalization in the United States is probably also generally the wrong direction.
[+] [-] Retric|17 years ago|reply
PS: The Pentagon is just an office building NORAD is far more important.
[+] [-] jgrahamc|17 years ago|reply
[+] [-] DenisM|17 years ago|reply
It seems you are mistaking your wishes for reality...
[+] [-] steveplace|17 years ago|reply
It's decentralization, and it's been occuring throughout history. For example, equality is a form of decentralization that has taken quite a while to manifest. You've also got decentralization in information, labor (specialization), and housing.
The next one that we desperately need is energy decentralization: we need to have smaller, agile, and more incremental methods of generating energy. The tech is coming along so we'll be able to do this, and some of the tech will only work well if it's not scalable.
My $.02
[+] [-] bdr|17 years ago|reply
- The Internet is different from other fields because of low barriers to entry and the ease of collaboration between groups (low transaction costs). This means that even if our field grows very large (and I think it will), the principles underlying its success might not successfully spread to other industries.
- Another thing that might happen is that large organizations gain enough power that they can control the Internet. If they're able to do so and prevent the rise of alternative, more free networks, then hegemony could come to dominate this industry too, as it did all the others.
[+] [-] Xichekolas|17 years ago|reply
So I would agree insofar as to say that startups will become the de facto standard method of commercializing new technology, but I don't think they will ever represent a significant portion of the economy as far as the number of people employed. Having such a small percentage of the population involved in startups doesn't afford them much opportunity to affect social norms on a large scale.
Short of complete automation, big organizations will always be best at doing commodity things like picking up your trash and making your shirts. Even once those particular tasks are automated, big organizations will grow (often from startups) elsewhere to tackle new commodity goods and services. Once something is no longer cutting edge, it'll be provided by a big organization.
Startups will always disrupt the way big organizations currently do things, but by being successful they will just become big organizations themselves.
[+] [-] ph0rque|17 years ago|reply
With the advent of open hardware and 3D printers, I think you can take out the word Internet and have the statement still be true.
[+] [-] biohacker42|17 years ago|reply
In my opinion there is not such a huge cultural gap that's responsible for the different startup rates.
Few (if any) cultures are so deeply authoritative as to be anti-success.
The valley's lead is much more the result of rather simpler causes. People - a whole lot of smart ones, two recession proof large smart people producers (the universities), and funding lots and lots of funding.
That's where the lead comes from and it will be difficult to close it. Culture doesn't have anything to do with it.
[+] [-] DaniFong|17 years ago|reply
I would argue that most cultures, even in California, are too authoritarian and inflexible to really cultivate productive, disruptive change. For all the talk about innovation, it is genuinely difficult to argue against the status quo. It is very, very hard to change one's frame of mind, and people tend elsewhere to defer to expertise and experience over ideas with are promising, make sense, but are not so fleshed out. In the Bay Area you will at least find an audience, if not easily convince them. Being anti-authority is obviously insufficient for startup success, though.
A good quote to keep in mind:
"...I would design my own, fresh, without knowing how other people do it. That was another thing that made me very good. All the best things that I did at Apple came from (a) not having money and (b) not having done it before, ever. Every single thing that we came out with that was really great, I'd never once done that thing in my life." -- Steve Wozniak
We are currently encountering much existing knowledge in engine design. Some of it is undoubtedly wise, but a lot of it is probably out of date. We have the chance to experiment. But it would be genuinely impossible to do such a thing in most auto companies today.
[+] [-] 13ren|17 years ago|reply
summary: http://en.wikipedia.org/wiki/The_Nature_of_the_Firm
actual essay: http://web.cenet.org.cn/upfile/30998.pdf
[+] [-] pjf|17 years ago|reply
"An ambitious kid graduating from college now doesn't want to work for a big company. They want to work for the hot startup that's rapidly growing into one."
Thats a generalization and I believe its untrue and biased towards talented, ambitious young people Paul meets. Sorry if I'm wrong, but if that was to convince me, I'd need a proof. From my point of view, the rest of the world is still at sort of peak of belief in the "corporate world" (but I dont have any research doc to reference as well :)).
"Fifty years later, startups are ubiquitous in Silicon Valley and common in a handful of other US cities, but they're still an anomaly in most of the world."
Well thats what I dont understand the most. If the whole article talks about the huge positive impact that startups are going to have on economy, let me ask Paul: what has changed since then ("fifty years [ago]") that makes you believe that NOW is the moment startups are going to change the whole world so much? What was the reason you wrote (published) the article NOW?
[+] [-] istara|17 years ago|reply
Another point to bear in mind: more small companies means more outsourcing, particularly for legal, HR, IT support, and so on. We should also see masses more companies using shared business premises, with shared receptions and so on.
Small companies should also mean more teleworking, since (1) people in small companies are generally more open to the idea of it, and (2) it's easier to keep track of what everyone in a smaller company is doing, even if they are working remotely, and (3) it means small companies can have international representation which may be one guy in a specific Asian country working from a bedroom-office.
[+] [-] dpapathanasiou|17 years ago|reply
"... your strategy is to be as hyperconservative and hyperaggressive as you can be instead of being mildly aggressive or conservative. Instead of putting your money in 'medium risk' investments, you need to put a portion, say 85 to 90 percent, in extremely safe instruments, like Treasury bills—as safe a class of instruments as you can manage to find on this planet. The remaining 10 to 15 percent you put in extremely speculative bets, as leveraged as possible (like options), preferably venture capital-style portfolios."
It seems that going to work for a startup is the exact opposite of this.
I.e., it would be better to be an investor in many startups rather than working solely on/for one startup.