12 ms·

The High-Res Society

168 points by mqt 18y ago

izaidi 18y ago
I like this a lot, but I'm not sure I understand the title.
- pg 18y ago
  Future economies will be made out of more, smaller pixels.
  - Retric 18y ago
    Successful companies tend to grow. Could Google make just about as much money with 5000 people as they currently make with 20,000? Perhaps, but while adding people has a positive ROI they are going to keep doing that. There high value for large organisations to buy small ones and IP law tends to create cash cows which buy up all the interesting companies around them. PS: When companies that can kill Google, IBM, and Toyota while staying small show up I might agree with you until then I expect things to stay like they are.
    - pg 18y ago
      "This doesn't mean big companies will disappear. To say that startups will succeed implies that big companies will exist, because startups that succeed either become big companies or are acquired by them. But large organizations will probably never again play the leading role they did up till the last quarter of the twentieth century." (http://www.paulgraham.com/highres.html http://www.paulgraham.com/highres.html)
      - albertcardona 18y ago
        Successful large companies may become environment providers, like universities. That's their self-assigned role as long as starting up is expensive and hard (bureaucracy, health care, etc.) IBM, Microsoft, Google--they have one or a few core selling products, the rest of their activities could be explained as self-promotion to those that would otherwise invest their time and skills elsewhere. In short, survival. Stretching it, such large companies may become countries in the sense that they cover the needs of their citizens. Like a VM on an OS. The lucky ones will have wise benevolent dictators (the best political option in Plato's government evolution stages--not that I am endorsing it.)
      - netcan 18y ago
        That is actually the point I thought needed more exploration. For one thing it seems to accept that 'startups' are a product of an environment dominated by large companies. There are some interesting questions that arise: * A theme running through many of pg's essay is that startups are a far more productive use of resources. Pg sometimes uses acquisitions as an example taking the same resources & applying them to a large organisation. The result is reduced productivity. Doesn't that imply that acquisitions shouldn't take place in the first place? A transaction is supposed to take place when it produces some sort of a net surplus. * Can a startup complex survive without acquisitions? * If size is a disadvantage, isn't this a disincentive to grow? Wouldn't that put a company in danger of being out competed by smaller players? * If growth is capped by anti - economy of scale effects, is there still enough incentive for startups? * Today's startups often float around winner take most areas. Isn't this at odds with small organisations being dominant? The solution to this one seems alarming: Small organisations with huge revenues. If this scales we get a large number of Googles being run entirely by a few hundred employees & responsible for a level of wealth creation grossly disproportionate to their number of employees. My ideas are pretty uncooked. But what I'm saying is that this prediction might be more robust if it explored a separation startup & big business. IE what does a startup look like in an economy dominated by startups. Foxes can never outnumber rabbits.
  - agentbleu 18y ago
    I would agree with this completely but extend it to monolithic intuitions such as governments, legal systems, policing, armies, councils, schools, etc. I envision that the networking changes we are witnessing are much more profound than just for business, which will apply themselves just the same to the very structures that currently control nations.
    - kragen 18y ago
      "extend it to monolithic intuitions such as ... policing, armies, ..." I hear they tried that in Somalia. Do you read John Robb?
      - agentbleu 18y ago
        I had not heard of John Robb. Many thanks for the suggestion.
        kragen 18y ago
        No problem. (I'd reply privately if I could, so as not to clutter up the thread, sorry.)
  - bk 18y ago
    "Free Agent Nation" describes this (or at least very similar) process/pattern as well. A world of free agents is at the highest resolution possible in that sense.
swombat 18y ago
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....
- RobGR 18y ago
  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.
  - time_management 18y ago
    Regarding the pharmaceutical industry, there's a lot of redundancy. As soon as one company develops a blockbuster molecule, others begin testing chemically similar molecules, hoping to "get in" on the category (e.g. statins). This is because there's a higher rate of success in developing a profitable brand in an established category where one successful molecule is already known. It's not necessarily a bad thing-- for example, it's quite good that Lexapro is available for those who might otherwise resort to the older and more side-effect-ridden Prozac-- but this leads to a proliferation of very similar molecules and devices, instead of research into entirely new approaches. In other words, the market ends up favoring optimization and refinement over innovation.
cool-RR 18y ago
"But large organizations will probably never again play the leading role they did up till the last quarter of the twentieth century." I suggest retracting this statement. What idea do you have of how the world will look 50 years from now? This may be remembered as your "640kb will always be enough" quote.
- harpastum 18y ago
  pg isn't prescient, but based on his argument this isn't much of a stretch. It may not turn out to be true, but it would be an even worse idea to cripple the conclusion to your thought because you're worried about embarrassment. . "The best way to predict the future is to invent it." - Alan Kay YC is definitely working to prove pg right.
  - cool-RR 18y ago
    "this isn't much of a stretch"?! from a sample period of a few decades to "never again"?
    - andreyf 18y ago
      No. Read the rest of the essay - the factors that made large organization work are dissolving.
- tsetse-fly 18y ago
  You missed the point of the essay. It was written to support that claim. Also, the 640k quote attributed to Bill Gates is folklore. http://en.wikiquote.org/wiki/Bill_Gates#Misattributed http://en.wikiquote.org/wiki/Bill_Gates#Misattributed
  - cool-RR 18y ago
    It looks like the article is saying there's a trend towards small organizations. That's a different thing than saying that the era of large organizations will never return.
sdurkin 18y ago
"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.
- hs 18y ago
  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)
  - taw 18y ago
    There's little correlation between cruelty during wars and cruelty during normal operations. Most successful empires were borderline genocidal when fighting, and mostly benign and well organized at home, as this combination allows expansion and prevents internal failure. I cannot really think of a good counterexample, Rome and other ancient powers, British Empire and successful colonial powers, and USA all fit here. Medieval Catholicism (which wasn't really a country, but it can be considered an empire of sorts) also expanded by genocide, but was rather benign inside. Soviet Union was very brutal inside and fell quickly due to inside forces.
ph0rque 18y ago
"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.
old-gregg 18y ago
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 Has always been like that.
- pg 18y ago
  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.
  - unalone 18y ago
    VCs are an industry unto themselves: I agree that that implies a shift towards start-ups. My question would be: what's there to stop a shift in corporations' becoming merely comprised of much smaller modules? Google works by having many small clusters of teams working on big things, so they're able (in some cases, not all) to innovate as quickly as start-ups in their fields, but with the advantage of firm corporate backing. Not many large companies are doing the same, but what's stopping new corporations from forming specifically with modulation in mind? Or would you argue that a development like that is similarly higher-res?
steveplace 18y ago
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. My $.02
geebee 18y ago
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". Don't ask.
- herdrick 18y ago
  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.
  - geebee 18y ago
    Revenge of the nerds had programmers. God, I loved that movie. My younger brother liked it even more. Remarkably, my parents let him watch it over and over on VHS. Violent movies were more likely to get scrutiny: for an explanation, see "Hair" ;)
  - jgrahamc 18y ago
    David Lightman. He was smart, nerdy, witty and got the girl.
  - zandorg 18y ago
    Hollywood nerd rule #1 - First the nerd saves the day with a hack, then gets killed so the leads can kiss.
- jedberg 18y ago
  Melanie Griffith starred in Working Girl. The rest of your post was spot on though. :) http://www.imdb.com/title/tt0096463/ http://www.imdb.com/title/tt0096463/
RobGR 18y ago
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.
- Retric 18y ago
  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.
1gor 18y ago
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.
- kds 18y ago
  That's a very good insight. I couldn't say it better myself.
- marvin 18y ago
  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.
- pchristensen 18y ago
  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.
nebula 18y ago
PG, your use of the term "economies of scale" seems to be slightly misleading: My understanding is that to benefit from "economies of scale", you don't necessarily have to be big in size. You can be a two people startup serving a huge user base with a small margin. It's possible to make huge profits in this scenario because of economies of scale in terms of the size of the market served. Your use of the terms seems to indicate that you can use economies of scale to your advantage only if you are big in size: Those who bet on economies of scale generally won, which meant the largest organizations were the most successful ones. But in the late twentieth century something changed. It turned out that economies of scale were not the only force at work. Particularly in technology, the increase in speed one could get from smaller groups started to trump the advantages of size.
- bdr 18y ago
  I think you're confusing "economies of scale" with what Taleb calls "professions that scale".
  - nebula 18y ago
    May be I'm; but you are not helping with your one liner :( [for "professions that scale" google is pointing to this page] Btw, Wikipedia seems to be more or less in agreement with my stand here: [http://en.wikipedia.org/wiki/Economies_of_scale http://en.wikipedia.org/wiki/Economies_of_scale] "Economies of scale are the cost advantages that a firm obtains due to expansion. Diseconomies of scale are the opposite. Economies of scale may be utilized by any size firm expanding its scale of operation." OK! It seems I missed the cost advantage part; i.e., the term "economies of scale" is used only if there is a cost advantage to be gained from expansion. Is that right?
    - bdr 18y ago
      Sorry, I assumed that that was where you were coming from. Economies of scale usually refers to how the cost of producing a unit decreases as you are producing more units at a time. Taleb calls an endeavor scalable if your reward can increase practically unboundedly with little no extra effort. Examples include pop stars and websites.
Alex3917 18y ago
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.
biohacker42 18y ago
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.
- DaniFong 18y ago
  "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.
  - MaysonL 18y ago
    Also - Feynman's "What I cannot create, I do not understand."
Xichekolas 18y ago
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.
bdr 18y ago
I had two responses: - 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.
crabapple 18y ago
this is almost 100% wrong. the more frictionless a market is, the more it will consolidate to a winner-take-all scenario. in ten years, five companies will control 95% of web traffic. we're already on the way. you can see the same slow-motion activity in the PC market. anyone can build a PC, so we must end up with 10,000 little PC builders right? wrong. you get HP, dell and Apple. even IBM has left the PC game. pg you are a righteous coder but this is economics, not tech
uuilly 18y ago
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.
- hs 18y ago
  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
- bokonist 18y ago
  As a recent ivy grad, I've noticed the exact same thing.
hs 18y ago
"There's no evidence that famously successful organizations like the Roman army or the British East India Company were any less afflicted by protocol and politics than organizations of the same size today. But they were competing against opponents who couldn't change the rules on the fly by discovering new technology." the mongol empire was the game changer [adopting siege machine, biological weapon (throwing infected corpse thru enemy's wall -- the seed of black death), religion (islam, christian, budhism), trade (silk route, 'pony' express), and the rise of creative class] it conquered the world with bow and arrow, lasted 3.5 centuries (longer than the nuclear powered usa) barbar/dictator/despotic empire can't/won't last that long -- there's truly something about mongol
mixmax 18y ago
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. Link: http://en.wikipedia.org/wiki/Theory_of_the_firm http://en.wikipedia.org/wiki/Theory_of_the_firm
pjf 18y ago
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?
dpapathanasiou 18y ago
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.
cadalac 18y ago
Quick question: Is the RSS for PG's essays not working? I never got the update.
13ren 18y ago
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. summary: http://en.wikipedia.org/wiki/The_Nature_of_the_Firm http://en.wikipedia.org/wiki/The_Nature_of_the_Firm actual essay: http://web.cenet.org.cn/upfile/30998.pdf http://web.cenet.org.cn/upfile/30998.pdf
raffi 18y ago
PG should also read The Third Wave by Alvin Toeffler. This sounds very much like his thinking.
pwim 18y ago
"To say that startups will succeed implies that big companies will exist, because startups that succeed either become big companies or are acquired by them." A startup can have a goal besides becoming a big company or being acquired by one. If the company meets that goal, it is successful.
akkartik 18y ago
"..there were already a handful of countries past that stage when the Industrial Revolution happened. There do not seem to be that many ready this time." Speculation: This factor may be the difference between life and death for the trend towards high-res. There's today a good chance the US will weaken economically, energy will get catastrophically more expensive, a city somewhere will experience a suitcase nuke. 'should', not 'will'.
known 18y ago
People join big company if they are opportunistic and join a start-up if they are ambitious.
bokonist 18y ago
Behind the success of most corporate giants is usually the not-so-invisble-hand of government connections. The 1960's organization man economy was a result of the tremendous growth of government during the New Deal and the World War. IBM was a small company until it received a giant government contract to run the Social Security accounting system. GE grew enormously from government spending on space, nuclear energy, and advanced AT&T's dominance was a result of nationalizing the telephone industry in 1918. The economies of scale for Wall St. companies are solely the result of the Federal Reserve's "too big to fail" policy ( and also regulatory costs). I think that the early 1900's were a more economical natural time. There were some big corporations, but the economy was largely dominated by small businesses and startups. The 1960's were the anomaly, a direct result of the creation of the American mega-state. Today the government and large government connected corporations have calcified. Look at NASA. It's not sexy to work there anymore because everyone knows that it hasn't done a single interesting thing since it put a man on the moon. But because it is a government program, it cannot fail and be replaced with something more dynamic. It's a zombie organization, not really alive, but unable to die. But till recently this was an anomalous route that tended to be followed only by outsiders. It was no coincidence that the great industrialists of the nineteenth century had so little formal education. As huge as their companies eventually became, they were all essentially mechanics and shopkeepers at first. In 1900, getting an education was anomalous. Most smart people were mechanics and shopkeepers, and if they were ambitious and talented, they created successful startups. The hacker in 1900 viewed a college degree like a hacker today views an MBA - as an expensive way to rot your brain. And finally, for those who got engineering degrees, I doubt starting a company was that uncommon. Off the top of my head, the very first student at Stanford ( Herbert Hoover) did indeed start his own company. So the tradition has been there for a long time. Now I would guess that practically every Stanford or Berkeley undergrad who knows how to program has at least considered the idea of starting a startup. East Coast universities are not far behind, and British universities only a little behind them. In the 2000's your average ambitious Ivy league graduate wanted to work for Wall St. The rest would go into consulting, academia, or get a profession degree. Very few people think of doing startups. If the Wall St. bubble never comes back, and graduates are forced to make an honest living, perhaps there will be a trend towards doing more startups. 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. Startups are an anomaly in Europe, Japan, and Korea because of government regulations. China is filled with startups because it has the freest market in the world. Startups are common in Sillicon valley because it is the home of the tech industry, which is the most unregulated industry in the United States.
prakash 18y ago
[3] It's possible that companies will one day be able to grow big in revenues without growing big in people, but we are not very far along that trend yet. PG: Can you expand your thoughts on this? Why do you think this trend is not very far along? What would help in accelerating such a trend? After startups, this trend could have a significant impact -- meaning folks that start companies/early employees would like to stay longer at such companies, and still be nimble, maybe not to the extent of a startup but not a BigCo. either.
SwellJoe 18y ago
So there's at least one problem with this, I think (though in principle, I suppose I agree). This line bothers me: Even if Internet-related applications only become a tenth of the world's economy A tenth? Internet applications? A tenth of the world's economy could be Internet applications? I'm a bit flabbergasted by the thought. I mean, I guess...if entertainment spending comes to be merged into the Internet applications category, and media, and toys and games, and travel, and education, etc. I guess the argument can be made that a tenth of the investment an average business makes (regardless of what the business actually does) in the future will be on Internet-related technology. But, I think we're still a long way from that, aren't we? And it doesn't alter the basic realities of the economy, and where people spend most of their money. How much do grocery stores and food producers spend on Internet applications? Certainly not 10%. Real estate and other big ticket item sellers like cars and boats? Is it 10%? Seems very unlikely, though maybe with advertising becoming more Internet focused one could say that their advertising budget will eventually be going to the Internet. Food and houses and other necessities are where 1/3 to 2/3 of the average person spends their money, worldwide. So, what else is there? Energy. I suspect energy companies spend maybe a tenth of one percent of their annual budget on Internet-related expenses, if that much (though they spend a lot on technology, in general). Unless we imagine that the percentage of money people spend on these things will drop significantly in the near future relative to entertainment, toys, sports, and knowledge-related activities (and all of those things become dramatically more Internet focused--which is the believable aspect of this scenario), a 10% Internet application economy future is hard for me to visualize. Of course pg is theorizing that the wealth being created is independent of existing systems of wealth generation. Being a libertarian, I'm sympathetic to that notion. But, there needs to be a productivity boost created by new technologies for it to literally create wealth. It needs to save people time or make some sort of production or transit more efficient. Entertainment doesn't produce wealth, IMHO, it merely redirects it...and some percentage of modern Internet applications are entertainment. I'm not saying that's a bad thing...just that I believe when the wealth being "created" isn't actually in response to production or increased efficiency, I suspect it's squeezing a balloon rather than blowing more air into it. And, the natural fallout of that is that if the brightest young lads and lasses are going into Internet applications, instead of taking part in the physical world, we might end up slowing the growth of wealth generation in the world rather than accelerating it. Particularly if they're mostly doing pure entertainment applications. Luckily, as long as we have a free market, it will correct for those kinds of mistakes, even if it takes a while for us to recognize the corrections (and even if the mistakes get subsidized across several years because investors are slow to react to shifts in the landscape).
istara 18y ago
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.
ralph 18y ago
Pedantically, Uncle Sid, not uncle Sid? Sid is my uncle, but to me, he's Uncle Sid. Or is that a Britishism?
RyanElisei 18y ago
Startups don't go against the traditional grain of society. They are part of the fauna that resulted from an economy of extremely large scale. Things like anti-trust laws, and investment banking - the food for startups - only happen in instances of super-scale.
mmilenko 18y ago
n-th restatement of the same idea. jeez, what a bore already
rpgoldman 18y ago
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.
coliveira 18y ago
Paul mentions the decreasing costs of technology, but he doesn't mention the fundamental property of technology that allows this to happen. This economic trend is directly related to technology and software. One of the problems for big corporations is that more and more is done by machines, i.e., software. However, the singular phenomenon of software is that, the more people working on it, the harder it is to maintain and evolve. Small groups are much better prepared for success in our present world because they can leverage software. As a single organization grows, so grows the complexity of its internal systems.
pestwave 18y ago
本文中文版请参考：http://www.yeeyan.com/articles/view/pestwave/19240 http://www.yeeyan.com/articles/view/pestwave/19240 A Chinese version of this article is available at:http://www.yeeyan.com/articles/view/pestwave/19240 http://www.yeeyan.com/articles/view/pestwave/19240
yahc 18y ago
"Which means the ambitious can now do arbitrage on them. It will be very valuable to understand precisely which ideas to keep and which can now be discarded." Do you mean "arbitrage" or do you mean "triage"?
h3h 18y ago
A corollary to the move toward smallness, if I might: http://h3h.net/2008/12/economy-of-software-maintenance/ http://h3h.net/2008/12/economy-of-software-maintenance/
pramodas23 18y ago
You can equate the development of internet to the development of language in humans.Traditionally in mature industries there are specialized roles for each unit and thus it means the entry barrier for a startup in the industry rises as the industry matures.But that's different for web based startups.The challenge for new startups will not be specialization(in technology) but specialization in understanding user behaviour.
zdrum 18y ago
There's a lot of start-ups per capita in Israel. The big sharks buy them so fast, Israelis are constantly busy opening the new ones. What a whirlpool.
JamesMitchell 18y ago
In many industries, there is a non-stoppable trend towards larger and larger companies. In retailing, Wal-Mart is crushing their competitors, who are also extremely large but they do not have the economies of scale that Wal-Mart. (In addition, Wal-Mart has better management.) In some cases, large companies succeed because consumers want the perceived value of purchasing from a large company. Dell and HP basically own the server and desktop computer market, and there is no reason to assume this will change in the next ten years. People feel that Dell will not go out of business. One of the industries hurting most right now is the automobile industry. GM and Chrysler say they will run out of money unless the U.S. government provides several billions of dollars in aid soon. Ford says it does not need government financing but it has pledged all of its assets in a collateral-based loan. Yet almost no one is willing to buy a car from a small automobile company, they are viewed as risker than GM or Chrysler. In choosing software companies, one factor sophisticated purchasers look at is, "How likely is this company to be in business 5 and 10 years from now?" When you commit to a software package, you are making a huge investment in time and energy, far beyond the purchase price of the software. You want to know that in the future, the software company will continue to enhance the product and will be around to provide technical support. This is why companies, for example, purchase software from Oracle, even though Oracle is ridiculously. People expect Oracle to be in business ten years from now. Yes, open source to some extent mitigates this risk, but putting aside very successful open source projects (Linux, Apache, MySQL, how do I know that they will be around ten years from now. Do I really want to writing the code myself?) Law firms have consolidated tremendously in the past decade. Large clients want one firm that can provide experise in dozen of areas of the law and can handle complex transactions that span the globe. In Boston, ten years a 100 lawyer firm was considered to be a large sophisticated firm. Now such firms (at least on the corporate side) have merged with other firms to create 500 attorney firms. In public accounting, the Big Eight has consolidated into the Big Four. Clients want the brand name that only a Big Four firm can provide. There is no clear trend here. In some industries, consolidation is taking place and will never stop. In other industries, it is the opposite effect. James Mitchell www.bostonconvivium.com jmitchell@kensingtonllc.com
hualibaobao 18y ago
RPatershuk 18y ago
This trend was examined in Zuboff's 2004 book "the Support Economy". I've commented it on my (fledgling) blog. www.efficientfactory.com http://efficientfactory.com/2009/01/17/the-support-economy-book-review/ http://efficientfactory.com/2009/01/17/the-support-economy-b...
mxtbcca 18y ago
I hear echos of William H. Davidow's "The Virtual Corporation", isbn 0887306578, in this post. I thought those ideas were innovative back then - 1993 - and still waiting for then to arrive in mass. mxt
connellybarnes 18y ago
The "corporate ladder" analogy is broken, because the ladder is tree shaped. It's statistically impossible to climb tree shaped ladders, because they get exponentially smaller at the top, and there are too many levels. "Climb the corporate n-ary tree" would be a better analogy, for large values of n, and large tree height. It should be obvious to computer scientists, via pigeonhole arguments, why this isn't statistically feasible. In any case, it's probably easier to find loopholes, like startups, or being an early arrival to an industry, or starting at a smaller company that inflates later, or tunneling your way between points in the corporate topology, by switching from one company to another, if you're ambitious. But betting against statistics never pays.

168 points by mqt 18y ago

izaidi 18y ago

I like this a lot, but I'm not sure I understand the title.

pg 18y ago
Future economies will be made out of more, smaller pixels.
- Retric 18y ago
  Successful companies tend to grow. Could Google make just about as much money with 5000 people as they currently make with 20,000? Perhaps, but while adding people has a positive ROI they are going to keep doing that. There high value for large organisations to buy small ones and IP law tends to create cash cows which buy up all the interesting companies around them. PS: When companies that can kill Google, IBM, and Toyota while staying small show up I might agree with you until then I expect things to stay like they are.
  - pg 18y ago
    "This doesn't mean big companies will disappear. To say that startups will succeed implies that big companies will exist, because startups that succeed either become big companies or are acquired by them. But large organizations will probably never again play the leading role they did up till the last quarter of the twentieth century." (http://www.paulgraham.com/highres.html http://www.paulgraham.com/highres.html)
    - albertcardona 18y ago
      Successful large companies may become environment providers, like universities. That's their self-assigned role as long as starting up is expensive and hard (bureaucracy, health care, etc.) IBM, Microsoft, Google--they have one or a few core selling products, the rest of their activities could be explained as self-promotion to those that would otherwise invest their time and skills elsewhere. In short, survival. Stretching it, such large companies may become countries in the sense that they cover the needs of their citizens. Like a VM on an OS. The lucky ones will have wise benevolent dictators (the best political option in Plato's government evolution stages--not that I am endorsing it.)
    - netcan 18y ago
      That is actually the point I thought needed more exploration. For one thing it seems to accept that 'startups' are a product of an environment dominated by large companies. There are some interesting questions that arise: * A theme running through many of pg's essay is that startups are a far more productive use of resources. Pg sometimes uses acquisitions as an example taking the same resources & applying them to a large organisation. The result is reduced productivity. Doesn't that imply that acquisitions shouldn't take place in the first place? A transaction is supposed to take place when it produces some sort of a net surplus. * Can a startup complex survive without acquisitions? * If size is a disadvantage, isn't this a disincentive to grow? Wouldn't that put a company in danger of being out competed by smaller players? * If growth is capped by anti - economy of scale effects, is there still enough incentive for startups? * Today's startups often float around winner take most areas. Isn't this at odds with small organisations being dominant? The solution to this one seems alarming: Small organisations with huge revenues. If this scales we get a large number of Googles being run entirely by a few hundred employees & responsible for a level of wealth creation grossly disproportionate to their number of employees. My ideas are pretty uncooked. But what I'm saying is that this prediction might be more robust if it explored a separation startup & big business. IE what does a startup look like in an economy dominated by startups. Foxes can never outnumber rabbits.
- agentbleu 18y ago
  I would agree with this completely but extend it to monolithic intuitions such as governments, legal systems, policing, armies, councils, schools, etc. I envision that the networking changes we are witnessing are much more profound than just for business, which will apply themselves just the same to the very structures that currently control nations.
  - kragen 18y ago
    "extend it to monolithic intuitions such as ... policing, armies, ..." I hear they tried that in Somalia. Do you read John Robb?
    - agentbleu 18y ago
      I had not heard of John Robb. Many thanks for the suggestion.
      - kragen 18y ago
        No problem. (I'd reply privately if I could, so as not to clutter up the thread, sorry.)
- bk 18y ago
  "Free Agent Nation" describes this (or at least very similar) process/pattern as well. A world of free agents is at the highest resolution possible in that sense.

swombat 18y ago

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....

RobGR 18y ago
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.
- time_management 18y ago
  Regarding the pharmaceutical industry, there's a lot of redundancy. As soon as one company develops a blockbuster molecule, others begin testing chemically similar molecules, hoping to "get in" on the category (e.g. statins). This is because there's a higher rate of success in developing a profitable brand in an established category where one successful molecule is already known. It's not necessarily a bad thing-- for example, it's quite good that Lexapro is available for those who might otherwise resort to the older and more side-effect-ridden Prozac-- but this leads to a proliferation of very similar molecules and devices, instead of research into entirely new approaches. In other words, the market ends up favoring optimization and refinement over innovation.

cool-RR 18y ago

"But large organizations will probably never again play the leading role they did up till the last quarter of the twentieth century." I suggest retracting this statement. What idea do you have of how the world will look 50 years from now? This may be remembered as your "640kb will always be enough" quote.

harpastum 18y ago
pg isn't prescient, but based on his argument this isn't much of a stretch. It may not turn out to be true, but it would be an even worse idea to cripple the conclusion to your thought because you're worried about embarrassment. . "The best way to predict the future is to invent it." - Alan Kay YC is definitely working to prove pg right.
- cool-RR 18y ago
  "this isn't much of a stretch"?! from a sample period of a few decades to "never again"?
  - andreyf 18y ago
    No. Read the rest of the essay - the factors that made large organization work are dissolving.
tsetse-fly 18y ago
You missed the point of the essay. It was written to support that claim. Also, the 640k quote attributed to Bill Gates is folklore. http://en.wikiquote.org/wiki/Bill_Gates#Misattributed http://en.wikiquote.org/wiki/Bill_Gates#Misattributed
- cool-RR 18y ago
  It looks like the article is saying there's a trend towards small organizations. That's a different thing than saying that the era of large organizations will never return.

sdurkin 18y ago

"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.

hs 18y ago
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)
- taw 18y ago
  There's little correlation between cruelty during wars and cruelty during normal operations. Most successful empires were borderline genocidal when fighting, and mostly benign and well organized at home, as this combination allows expansion and prevents internal failure. I cannot really think of a good counterexample, Rome and other ancient powers, British Empire and successful colonial powers, and USA all fit here. Medieval Catholicism (which wasn't really a country, but it can be considered an empire of sorts) also expanded by genocide, but was rather benign inside. Soviet Union was very brutal inside and fell quickly due to inside forces.

ph0rque 18y ago

"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.

old-gregg 18y ago

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 Has always been like that.

pg 18y ago
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.
- unalone 18y ago
  VCs are an industry unto themselves: I agree that that implies a shift towards start-ups. My question would be: what's there to stop a shift in corporations' becoming merely comprised of much smaller modules? Google works by having many small clusters of teams working on big things, so they're able (in some cases, not all) to innovate as quickly as start-ups in their fields, but with the advantage of firm corporate backing. Not many large companies are doing the same, but what's stopping new corporations from forming specifically with modulation in mind? Or would you argue that a development like that is similarly higher-res?

steveplace 18y ago

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. My $.02

geebee 18y ago

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". Don't ask.

herdrick 18y ago
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.
- geebee 18y ago
  Revenge of the nerds had programmers. God, I loved that movie. My younger brother liked it even more. Remarkably, my parents let him watch it over and over on VHS. Violent movies were more likely to get scrutiny: for an explanation, see "Hair" ;)
- jgrahamc 18y ago
  David Lightman. He was smart, nerdy, witty and got the girl.
- zandorg 18y ago
  Hollywood nerd rule #1 - First the nerd saves the day with a hack, then gets killed so the leads can kiss.
jedberg 18y ago
Melanie Griffith starred in Working Girl. The rest of your post was spot on though. :) http://www.imdb.com/title/tt0096463/ http://www.imdb.com/title/tt0096463/

RobGR 18y ago

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.

Retric 18y ago
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.

1gor 18y ago

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.

kds 18y ago
That's a very good insight. I couldn't say it better myself.
marvin 18y ago
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.
pchristensen 18y ago
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.

nebula 18y ago

PG, your use of the term "economies of scale" seems to be slightly misleading: My understanding is that to benefit from "economies of scale", you don't necessarily have to be big in size. You can be a two people startup serving a huge user base with a small margin. It's possible to make huge profits in this scenario because of economies of scale in terms of the size of the market served. Your use of the terms seems to indicate that you can use economies of scale to your advantage only if you are big in size: Those who bet on economies of scale generally won, which meant the largest organizations were the most successful ones. But in the late twentieth century something changed. It turned out that economies of scale were not the only force at work. Particularly in technology, the increase in speed one could get from smaller groups started to trump the advantages of size.

bdr 18y ago
I think you're confusing "economies of scale" with what Taleb calls "professions that scale".
- nebula 18y ago
  May be I'm; but you are not helping with your one liner :( [for "professions that scale" google is pointing to this page] Btw, Wikipedia seems to be more or less in agreement with my stand here: [http://en.wikipedia.org/wiki/Economies_of_scale http://en.wikipedia.org/wiki/Economies_of_scale] "Economies of scale are the cost advantages that a firm obtains due to expansion. Diseconomies of scale are the opposite. Economies of scale may be utilized by any size firm expanding its scale of operation." OK! It seems I missed the cost advantage part; i.e., the term "economies of scale" is used only if there is a cost advantage to be gained from expansion. Is that right?
  - bdr 18y ago
    Sorry, I assumed that that was where you were coming from. Economies of scale usually refers to how the cost of producing a unit decreases as you are producing more units at a time. Taleb calls an endeavor scalable if your reward can increase practically unboundedly with little no extra effort. Examples include pop stars and websites.

Alex3917 18y ago

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.

biohacker42 18y ago

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.

DaniFong 18y ago
"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.
- MaysonL 18y ago
  Also - Feynman's "What I cannot create, I do not understand."

Xichekolas 18y ago

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.

bdr 18y ago

I had two responses: - 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.

crabapple 18y ago

this is almost 100% wrong. the more frictionless a market is, the more it will consolidate to a winner-take-all scenario. in ten years, five companies will control 95% of web traffic. we're already on the way. you can see the same slow-motion activity in the PC market. anyone can build a PC, so we must end up with 10,000 little PC builders right? wrong. you get HP, dell and Apple. even IBM has left the PC game. pg you are a righteous coder but this is economics, not tech

uuilly 18y ago

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.

hs 18y ago
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
bokonist 18y ago
As a recent ivy grad, I've noticed the exact same thing.

hs 18y ago

"There's no evidence that famously successful organizations like the Roman army or the British East India Company were any less afflicted by protocol and politics than organizations of the same size today. But they were competing against opponents who couldn't change the rules on the fly by discovering new technology." the mongol empire was the game changer [adopting siege machine, biological weapon (throwing infected corpse thru enemy's wall -- the seed of black death), religion (islam, christian, budhism), trade (silk route, 'pony' express), and the rise of creative class] it conquered the world with bow and arrow, lasted 3.5 centuries (longer than the nuclear powered usa) barbar/dictator/despotic empire can't/won't last that long -- there's truly something about mongol

mixmax 18y ago

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. Link: http://en.wikipedia.org/wiki/Theory_of_the_firm http://en.wikipedia.org/wiki/Theory_of_the_firm

pjf 18y ago

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?

dpapathanasiou 18y ago

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.

cadalac 18y ago

Quick question: Is the RSS for PG's essays not working? I never got the update.

13ren 18y ago

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. summary: http://en.wikipedia.org/wiki/The_Nature_of_the_Firm http://en.wikipedia.org/wiki/The_Nature_of_the_Firm actual essay: http://web.cenet.org.cn/upfile/30998.pdf http://web.cenet.org.cn/upfile/30998.pdf

raffi 18y ago

PG should also read The Third Wave by Alvin Toeffler. This sounds very much like his thinking.

pwim 18y ago

"To say that startups will succeed implies that big companies will exist, because startups that succeed either become big companies or are acquired by them." A startup can have a goal besides becoming a big company or being acquired by one. If the company meets that goal, it is successful.

akkartik 18y ago

"..there were already a handful of countries past that stage when the Industrial Revolution happened. There do not seem to be that many ready this time." Speculation: This factor may be the difference between life and death for the trend towards high-res. There's today a good chance the US will weaken economically, energy will get catastrophically more expensive, a city somewhere will experience a suitcase nuke. 'should', not 'will'.

known 18y ago

People join big company if they are opportunistic and join a start-up if they are ambitious.

bokonist 18y ago

Behind the success of most corporate giants is usually the not-so-invisble-hand of government connections. The 1960's organization man economy was a result of the tremendous growth of government during the New Deal and the World War. IBM was a small company until it received a giant government contract to run the Social Security accounting system. GE grew enormously from government spending on space, nuclear energy, and advanced AT&T's dominance was a result of nationalizing the telephone industry in 1918. The economies of scale for Wall St. companies are solely the result of the Federal Reserve's "too big to fail" policy ( and also regulatory costs). I think that the early 1900's were a more economical natural time. There were some big corporations, but the economy was largely dominated by small businesses and startups. The 1960's were the anomaly, a direct result of the creation of the American mega-state. Today the government and large government connected corporations have calcified. Look at NASA. It's not sexy to work there anymore because everyone knows that it hasn't done a single interesting thing since it put a man on the moon. But because it is a government program, it cannot fail and be replaced with something more dynamic. It's a zombie organization, not really alive, but unable to die. But till recently this was an anomalous route that tended to be followed only by outsiders. It was no coincidence that the great industrialists of the nineteenth century had so little formal education. As huge as their companies eventually became, they were all essentially mechanics and shopkeepers at first. In 1900, getting an education was anomalous. Most smart people were mechanics and shopkeepers, and if they were ambitious and talented, they created successful startups. The hacker in 1900 viewed a college degree like a hacker today views an MBA - as an expensive way to rot your brain. And finally, for those who got engineering degrees, I doubt starting a company was that uncommon. Off the top of my head, the very first student at Stanford ( Herbert Hoover) did indeed start his own company. So the tradition has been there for a long time. Now I would guess that practically every Stanford or Berkeley undergrad who knows how to program has at least considered the idea of starting a startup. East Coast universities are not far behind, and British universities only a little behind them. In the 2000's your average ambitious Ivy league graduate wanted to work for Wall St. The rest would go into consulting, academia, or get a profession degree. Very few people think of doing startups. If the Wall St. bubble never comes back, and graduates are forced to make an honest living, perhaps there will be a trend towards doing more startups. 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. Startups are an anomaly in Europe, Japan, and Korea because of government regulations. China is filled with startups because it has the freest market in the world. Startups are common in Sillicon valley because it is the home of the tech industry, which is the most unregulated industry in the United States.

prakash 18y ago

[3] It's possible that companies will one day be able to grow big in revenues without growing big in people, but we are not very far along that trend yet. PG: Can you expand your thoughts on this? Why do you think this trend is not very far along? What would help in accelerating such a trend? After startups, this trend could have a significant impact -- meaning folks that start companies/early employees would like to stay longer at such companies, and still be nimble, maybe not to the extent of a startup but not a BigCo. either.

SwellJoe 18y ago

So there's at least one problem with this, I think (though in principle, I suppose I agree). This line bothers me: Even if Internet-related applications only become a tenth of the world's economy A tenth? Internet applications? A tenth of the world's economy could be Internet applications? I'm a bit flabbergasted by the thought. I mean, I guess...if entertainment spending comes to be merged into the Internet applications category, and media, and toys and games, and travel, and education, etc. I guess the argument can be made that a tenth of the investment an average business makes (regardless of what the business actually does) in the future will be on Internet-related technology. But, I think we're still a long way from that, aren't we? And it doesn't alter the basic realities of the economy, and where people spend most of their money. How much do grocery stores and food producers spend on Internet applications? Certainly not 10%. Real estate and other big ticket item sellers like cars and boats? Is it 10%? Seems very unlikely, though maybe with advertising becoming more Internet focused one could say that their advertising budget will eventually be going to the Internet. Food and houses and other necessities are where 1/3 to 2/3 of the average person spends their money, worldwide. So, what else is there? Energy. I suspect energy companies spend maybe a tenth of one percent of their annual budget on Internet-related expenses, if that much (though they spend a lot on technology, in general). Unless we imagine that the percentage of money people spend on these things will drop significantly in the near future relative to entertainment, toys, sports, and knowledge-related activities (and all of those things become dramatically more Internet focused--which is the believable aspect of this scenario), a 10% Internet application economy future is hard for me to visualize. Of course pg is theorizing that the wealth being created is independent of existing systems of wealth generation. Being a libertarian, I'm sympathetic to that notion. But, there needs to be a productivity boost created by new technologies for it to literally create wealth. It needs to save people time or make some sort of production or transit more efficient. Entertainment doesn't produce wealth, IMHO, it merely redirects it...and some percentage of modern Internet applications are entertainment. I'm not saying that's a bad thing...just that I believe when the wealth being "created" isn't actually in response to production or increased efficiency, I suspect it's squeezing a balloon rather than blowing more air into it. And, the natural fallout of that is that if the brightest young lads and lasses are going into Internet applications, instead of taking part in the physical world, we might end up slowing the growth of wealth generation in the world rather than accelerating it. Particularly if they're mostly doing pure entertainment applications. Luckily, as long as we have a free market, it will correct for those kinds of mistakes, even if it takes a while for us to recognize the corrections (and even if the mistakes get subsidized across several years because investors are slow to react to shifts in the landscape).

istara 18y ago

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.

ralph 18y ago

Pedantically, Uncle Sid, not uncle Sid? Sid is my uncle, but to me, he's Uncle Sid. Or is that a Britishism?

RyanElisei 18y ago

Startups don't go against the traditional grain of society. They are part of the fauna that resulted from an economy of extremely large scale. Things like anti-trust laws, and investment banking - the food for startups - only happen in instances of super-scale.

mmilenko 18y ago

n-th restatement of the same idea. jeez, what a bore already

rpgoldman 18y ago

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.

coliveira 18y ago

Paul mentions the decreasing costs of technology, but he doesn't mention the fundamental property of technology that allows this to happen. This economic trend is directly related to technology and software. One of the problems for big corporations is that more and more is done by machines, i.e., software. However, the singular phenomenon of software is that, the more people working on it, the harder it is to maintain and evolve. Small groups are much better prepared for success in our present world because they can leverage software. As a single organization grows, so grows the complexity of its internal systems.

pestwave 18y ago

本文中文版请参考：http://www.yeeyan.com/articles/view/pestwave/19240 http://www.yeeyan.com/articles/view/pestwave/19240 A Chinese version of this article is available at:http://www.yeeyan.com/articles/view/pestwave/19240 http://www.yeeyan.com/articles/view/pestwave/19240

yahc 18y ago

"Which means the ambitious can now do arbitrage on them. It will be very valuable to understand precisely which ideas to keep and which can now be discarded." Do you mean "arbitrage" or do you mean "triage"?

h3h 18y ago

A corollary to the move toward smallness, if I might: http://h3h.net/2008/12/economy-of-software-maintenance/ http://h3h.net/2008/12/economy-of-software-maintenance/

pramodas23 18y ago

You can equate the development of internet to the development of language in humans.Traditionally in mature industries there are specialized roles for each unit and thus it means the entry barrier for a startup in the industry rises as the industry matures.But that's different for web based startups.The challenge for new startups will not be specialization(in technology) but specialization in understanding user behaviour.

zdrum 18y ago

There's a lot of start-ups per capita in Israel. The big sharks buy them so fast, Israelis are constantly busy opening the new ones. What a whirlpool.

JamesMitchell 18y ago

In many industries, there is a non-stoppable trend towards larger and larger companies. In retailing, Wal-Mart is crushing their competitors, who are also extremely large but they do not have the economies of scale that Wal-Mart. (In addition, Wal-Mart has better management.) In some cases, large companies succeed because consumers want the perceived value of purchasing from a large company. Dell and HP basically own the server and desktop computer market, and there is no reason to assume this will change in the next ten years. People feel that Dell will not go out of business. One of the industries hurting most right now is the automobile industry. GM and Chrysler say they will run out of money unless the U.S. government provides several billions of dollars in aid soon. Ford says it does not need government financing but it has pledged all of its assets in a collateral-based loan. Yet almost no one is willing to buy a car from a small automobile company, they are viewed as risker than GM or Chrysler. In choosing software companies, one factor sophisticated purchasers look at is, "How likely is this company to be in business 5 and 10 years from now?" When you commit to a software package, you are making a huge investment in time and energy, far beyond the purchase price of the software. You want to know that in the future, the software company will continue to enhance the product and will be around to provide technical support. This is why companies, for example, purchase software from Oracle, even though Oracle is ridiculously. People expect Oracle to be in business ten years from now. Yes, open source to some extent mitigates this risk, but putting aside very successful open source projects (Linux, Apache, MySQL, how do I know that they will be around ten years from now. Do I really want to writing the code myself?) Law firms have consolidated tremendously in the past decade. Large clients want one firm that can provide experise in dozen of areas of the law and can handle complex transactions that span the globe. In Boston, ten years a 100 lawyer firm was considered to be a large sophisticated firm. Now such firms (at least on the corporate side) have merged with other firms to create 500 attorney firms. In public accounting, the Big Eight has consolidated into the Big Four. Clients want the brand name that only a Big Four firm can provide. There is no clear trend here. In some industries, consolidation is taking place and will never stop. In other industries, it is the opposite effect. James Mitchell www.bostonconvivium.com jmitchell@kensingtonllc.com

hualibaobao 18y ago

RPatershuk 18y ago

This trend was examined in Zuboff's 2004 book "the Support Economy". I've commented it on my (fledgling) blog. www.efficientfactory.com http://efficientfactory.com/2009/01/17/the-support-economy-book-review/ http://efficientfactory.com/2009/01/17/the-support-economy-b...

mxtbcca 18y ago

I hear echos of William H. Davidow's "The Virtual Corporation", isbn 0887306578, in this post. I thought those ideas were innovative back then - 1993 - and still waiting for then to arrive in mass. mxt

connellybarnes 18y ago

The "corporate ladder" analogy is broken, because the ladder is tree shaped. It's statistically impossible to climb tree shaped ladders, because they get exponentially smaller at the top, and there are too many levels. "Climb the corporate n-ary tree" would be a better analogy, for large values of n, and large tree height. It should be obvious to computer scientists, via pigeonhole arguments, why this isn't statistically feasible. In any case, it's probably easier to find loopholes, like startups, or being an early arrival to an industry, or starting at a smaller company that inflates later, or tunneling your way between points in the corporate topology, by switching from one company to another, if you're ambitious. But betting against statistics never pays.