August 2005(This essay is derived from a talk at Oscon 2005.)Lately companies have been paying more attention to open source.Ten years ago there seemed a real danger Microsoft would extend itsmonopoly to servers. It seems safe to say now that open source hasprevented that. A recent survey found 52% of companies are replacingWindows servers with Linux servers.[1]More significant, I think, is which 52% they are. At this point,anyone proposing to run Windows on servers should be prepared toexplain what they know about servers that Google, Yahoo, and Amazondon't.But the biggest thing business has to learn from open source is notabout Linux or Firefox, but about the forces that produced them.Ultimately these will affect a lot more than what software you use.We may be able to get a fix on these underlying forces by triangulatingfrom open source and blogging. As you've probably noticed, theyhave a lot in common.Like open source, blogging is something people do themselves, forfree, because they enjoy it. Like open source hackers, bloggerscompete with people working for money, and often win. The methodof ensuring quality is also the same: Darwinian. Companies ensurequality through rules to prevent employees from screwing up. Butyou don't need that when the audience can communicate with oneanother. People just produce whatever they want; the good stuffspreads, and the bad gets ignored. And in both cases, feedbackfrom the audience improves the best work.Another thing blogging and open source have in common is the Web.People have always been willing to do great workfor free, but before the Web it was harder to reach an audienceor collaborate on projects.AmateursI think the most important of the new principles business has to learn isthat people work a lot harder on stuff they like. Well, that'snews to no one. So how can I claim business has to learn it? WhenI say business doesn't know this, I mean the structure of businessdoesn't reflect it.Business still reflects an older model, exemplified by the Frenchword for working: travailler. It has an English cousin, travail,and what it means is torture.[2]This turns out not to be the last word on work, however.As societies get richer, they learn something aboutwork that's a lot like what they learn about diet. We know now that thehealthiest diet is the one our peasant ancestors were forced toeat because they were poor. Like rich food, idlenessonly seems desirable when you don't get enough of it. I think we weredesigned to work, just as we were designed to eat a certain amountof fiber, and we feel bad if we don't.There's a name for people who work for the love of it: amateurs.The word now has such bad connotations that we forget its etymology,though it's staring us in the face. "Amateur" was originally rathera complimentary word. But the thing to be in the twentieth centurywas professional, which amateurs, by definition, are not.That's why the business world was so surprised by one lesson fromopen source: that people working for love often surpass those workingfor money. Users don't switch from Explorer to Firefox becausethey want to hack the source. They switch because it's a betterbrowser.It's not that Microsoft isn't trying. They know controlling thebrowser is one of the keys to retaining their monopoly. The problemis the same they face in operating systems: they can't pay peopleenough to build something better than a group of inspired hackerswill build for free.I suspect professionalism was always overrated-- not just in theliteral sense of working for money, but also connotations likeformality and detachment. Inconceivable as it would have seemedin, say, 1970, I think professionalism was largely a fashion,driven by conditions that happened to exist in the twentieth century.One of the most powerful of those was the existence of "channels." Revealingly,the same term was used for both products and information: therewere distribution channels, and TV and radio channels.It was the narrowness of such channels that made professionalsseem so superior to amateurs. There were only a few jobs asprofessional journalists, for example, so competition ensured theaverage journalist was fairly good. Whereas anyone can expressopinions about current events in a bar. And so the average personexpressing his opinions in a bar sounds like an idiot compared toa journalist writing about the subject.On the Web, the barrier for publishing your ideas is even lower.You don't have to buy a drink, and they even let kids in.Millions of people are publishing online, and the averagelevel of what they're writing, as you might expect, is not verygood. This has led some in the media to conclude that blogs don'tpresent much of a threat-- that blogs are just a fad.Actually, the fad is the word "blog," at least the way the printmedia now use it. What they mean by "blogger" is not someone whopublishes in a weblog format, but anyone who publishes online.That's going to become a problem as the Web becomes the defaultmedium for publication. So I'dlike to suggest an alternative word for someone who publishes online.How about "writer?"Those in the print media who dismiss the writing online because ofits low average quality are missing an important point: no one readsthe average blog. In the old world of channels, it meant somethingto talk about average quality, because that's what you were gettingwhether you liked it or not.But now you can read any writer you want. So the averagequality of writing online isn't what the print media are competingagainst. They're competing against the best writing online. And, like Microsoft, they're losing.I know that from my own experience as a reader. Though most printpublications are online, I probablyread two or three articles on individual people's sites for everyone I read on the site of a newspaper or magazine.And when I read, say, New York Times stories, I never reachthem through the Times front page. Most I find through aggregatorslike Google News or Slashdot or Delicious. Aggregators show howmuch better you can do than the channel. The New York Times front page isa list of articles written by people who work for the New York Times. Deliciousis a list of articles that are interesting. And it's only now thatyou can see the two side by side that you notice how little overlap there is.Most articles in the print media are boring. For example, thepresident notices that a majority of voters now think invading Iraqwas a mistake, so he makes an address to the nation to drum upsupport. Where is the man bites dog in that? I didn't hear thespeech, but I could probably tell you exactly what he said. Aspeech like that is, in the most literal sense, not news: there isnothing new in it.[3]Nor is there anything new, except the names and places, in most"news" about things going wrong. A child is abducted; there's atornado; a ferry sinks; someone gets bitten by a shark; a smallplane crashes. And what do you learn about the world from thesestories? Absolutely nothing. They're outlying data points; whatmakes them gripping also makes them irrelevant.As in software, when professionals produce such crap, it's notsurprising if amateurs can do better. Live by the channel, die bythe channel: if you depend on an oligopoly, you sink into bad habitsthat are hard to overcome when you suddenly get competition.[4]WorkplacesAnother thing blogs and open source software have in common is thatthey're often made by people working at home. That may not seemsurprising. But it should be. It's the architectural equivalentof a home-made aircraft shooting down an F-18. Companies spendmillions to build office buildings for a single purpose: to be aplace to work. And yet people working in their own homes,which aren't even designed to be workplaces, end upbeing more productive.This proves something a lot of us have suspected. The averageoffice is a miserable place to get work done. And a lot of whatmakes offices bad are the very qualities we associate withprofessionalism. The sterilityof offices is supposed to suggest efficiency. But suggestingefficiency is a different thing from actually being efficient.The atmosphere of the average workplace is to productivity whatflames painted on the side of a car are to speed. And it's notjust the way offices look that's bleak. The way people act is justas bad.Things are different in a startup. Often as not a startup beginsin an apartment. Instead of matching beige cubiclesthey have an assortment of furniture they bought used. They workodd hours, wearing the most casual of clothing. They look atwhatever they want online without worrying whether it's "work safe."The cheery, bland language of the office is replaced by wicked humor. Andyou know what? The company at this stage is probably the mostproductive it's ever going to be.Maybe it's not a coincidence. Maybe some aspects of professionalismare actually a net lose.To me the most demoralizing aspect of the traditional office isthat you're supposed to be there at certain times. There are usuallya few people in a company who really have to, but the reason mostemployees work fixed hours is that the company can't measure theirproductivity.The basic idea behind office hours is that if you can't make peoplework, you can at least prevent them from having fun. If employeeshave to be in the building a certain number of hours a day, and areforbidden to do non-work things while there, then they must beworking. In theory. In practice they spend a lot of their timein a no-man's land, where they're neither working nor having fun.If you could measure how much work people did, many companieswouldn't need any fixed workday. You could just say: this is whatyou have to do. Do it whenever you like, wherever you like. Ifyour work requires you to talk to other people in the company, thenyou may need to be here a certain amount. Otherwise we don't care.That may seem utopian, but it's what we told people who came towork for our company. There were no fixed office hours. I nevershowed up before 11 in the morning. But we weren't saying this tobe benevolent. We were saying: if you work here we expect you toget a lot done. Don't try to fool us just by being here a lot.The problem with the facetime model is not just that it's demoralizing, butthat the people pretending to work interruptthe ones actually working. I'm convinced the facetime modelis the main reason large organizations have so many meetings.Per capita, large organizations accomplish very little.And yet all those people have to be on site at least eight hours aday. When so much time goes in one end and so little achievementcomes out the other, something has to give. And meetings are themain mechanism for taking up the slack.For one year I worked at a regular nine to five job, and I rememberwell the strange, cozy feeling that comes over one during meetings.I was very aware, because of the novelty, that I was being paid forprogramming. It seemed just amazing, as if there was a machine onmy desk that spat out a dollar bill every two minutes no matterwhat I did. Even while I was in the bathroom! But because theimaginary machine was always running, I felt I always ought to beworking. And so meetings felt wonderfully relaxing. Theycounted as work, just like programming, but they were so much easier.All you had to do was sit and look attentive.Meetings are like an opiate with a network effect. So is email,on a smaller scale. And in addition to the direct cost in time,there's the cost in fragmentation-- breaking people's day up intobits too small to be useful.You can see how dependent you've become on something by removingit suddenly. So for big companies I propose the following experiment.Set aside one day where meetings are forbidden-- where everyone has tosit at their desk all day and work without interruption onthings they can do without talking to anyone else.Some amount of communication is necessary in most jobs, but I'msure many employees could find eight hours worth of stuff they coulddo by themselves. You could call it "Work Day."The other problem with pretend workis that it often looks better than real work. When I'mwriting or hacking I spend as much time just thinking as I doactually typing. Half the time I'm sitting drinking a cup of tea,or walking around the neighborhood. This is a critical phase--this is where ideas come from-- and yet I'd feel guilty doing thisin most offices, with everyone else looking busy.It's hard to see how bad some practice is till you have somethingto compare it to. And that's one reason open source, and even bloggingin some cases, are so important. They show us what real work looks like.We're funding eight new startups at the moment. A friend askedwhat they were doing for office space, and seemed surprised when Isaid we expected them to work out of whatever apartments they foundto live in. But we didn't propose that to save money. We did itbecause we want their software to be good. Working in crappyinformal spaces is one of the things startups do right withoutrealizing it. As soon as you get into an office, work and lifestart to drift apart.That is one of the key tenets of professionalism. Work and lifeare supposed to be separate. But that part, I'm convinced, is a mistake.Bottom-UpThe third big lesson we can learn from open source andblogging is that ideas can bubble up from the bottom, instead offlowing down from the top. Open source and blogging both workbottom-up: people make what they want, and the best stuffprevails.Does this sound familiar? It's the principle of a market economy.Ironically, though open source and blogs are done for free, thoseworlds resemble market economies, while most companies, for alltheir talk about the value of free markets, are run internally likecommunist states.There are two forces that together steer design: ideas aboutwhat to do next, and the enforcement of quality. In the channelera, both flowed down from the top. For example, newspaper editorsassigned stories to reporters, then edited what they wrote.Open source and blogging show us things don't have to work thatway. Ideas and even the enforcement of quality can flow bottom-up.And in both cases the results are not merely acceptable, but better.For example, open source software is more reliable precisely becauseit's open source; anyone can find mistakes.The same happens with writing. As we got close to publication, Ifound I was very worried about the essays in Hackers& Paintersthat hadn't been online. Once an essay has had a couple thousandpage views I feel reasonably confident about it. But these had had literally orders of magnitude less scrutiny. It felt likereleasing software without testing it.That's what all publishing used to be like. Ifyou got ten people to read a manuscript, you were lucky. But I'dbecome so used to publishing online that the old method now seemedalarmingly unreliable, like navigating by dead reckoning once you'dgotten used to a GPS.The other thing I like about publishing online is that you can writewhat you want and publish when you want. Earlier this year I wrotesomething that seemed suitable for a magazine, soI sent it to an editor I know.As I was waiting to hear back, I found to my surprise that I washoping they'd reject it. Then I could put it online right away.If they accepted it, it wouldn't be read by anyone for months, andin the meantime I'd have to fight word-by-word to save it from beingmangled by some twenty five year old copy editor.[5]Many employees would like to build great things for the companiesthey work for, but more often than not management won't let them.How many of us have heard stories of employees going to managementand saying, please let us build this thing to make money for you--and the company saying no? The most famous example is probably Steve Wozniak,who originally wanted to build microcomputers for his then-employer, HP.And they turned him down. On the blunderometer, this episode rankswith IBM accepting a non-exclusive license for DOS. But I think thishappens all the time. We just don't hear about it usually,because to prove yourself right you have to quitand start your own company, like Wozniak did.StartupsSo these, I think, are the three big lessons open source and blogginghave to teach business: (1) that people work harder on stuff theylike, (2) that the standard office environment is very unproductive,and (3) that bottom-up often works better than top-down.I can imagine managers at this point saying: what is this guy talkingabout? What good does it do me to know that my programmerswould be more productiveworking at home on their own projects? I need their asses in hereworking on version 3.2 of our software, or we're never going tomake the release date.And it's true, the benefit that specific manager could derive fromthe forces I've described is near zero. When I say business canlearn from open source, I don't mean any specific business can. Imean business can learn about new conditions the same way a genepool does. I'm not claiming companies can get smarter, just thatdumb ones will die.So what will business look like when it has assimilated the lessonsof open source and blogging? I think the big obstacle preventingus from seeing the future of business is the assumption that peopleworking for you have to be employees. But think about what's goingon underneath: the company has some money, and they pay it to theemployee in the hope that he'll make something worth more than theypaid him. Well, there are other ways to arrange that relationship.Instead of paying the guy money as a salary, why not give it to himas investment? Then instead of coming to your office to work onyour projects, he can work wherever he wants on projects of his own.Because few of us know any alternative, we have no idea how muchbetter we could do than the traditional employer-employee relationship.Such customs evolve with glacial slowness. Our employer-employee relationship still retains a big chunk ofmaster-servant DNA.[6]I dislike being on either end of it.I'll work my ass off for a customer, but I resent being told whatto do by a boss. And being a boss is also horribly frustrating; half the time it's easier just to do stuff yourself than to getsomeone else to do it for you.I'd rather do almost anything than give or receive aperformance review.On top of its unpromising origins, employmenthas accumulated a lot of cruft over the years. The list of whatyou can't ask in job interviews is now so long that for convenienceI assume it's infinite. Within theoffice you now have to walk on eggshells lest anyone say or dosomething that makes the company prey to a lawsuit. And God helpyou if you fire anyone.Nothing shows more clearly that employment is not an ordinary economicrelationship than companies being sued for firing people. In anypurely economic relationship you're free to do what you want. Ifyou want to stop buying steel pipe from one supplier and startbuying it from another, you don't have to explain why. No one canaccuse you of unjustly switching pipe suppliers. Justice impliessome kind of paternal obligation that isn't there intransactions between equals.Most of the legal restrictions on employers are intended to protectemployees. But you can't have action without an equal and oppositereaction. You can't expect employers to have some kind of paternalresponsibility toward employees without putting employees in theposition of children. And that seems a bad road to go down.Next time you're in a moderately large city, drop by the main postoffice and watch the body language of the people working there.They have the same sullen resentment as children made to dosomething they don't want to. Their union has exacted payincreases and work restrictions that would have been the envy ofprevious generations of postal workers, and yet they don't seem anyhappier for it. It's demoralizingto be on the receiving end of a paternalistic relationship, nomatter how cozy the terms. Just ask any teenager.I see the disadvantages of the employer-employee relationship becauseI've been on both sides of a better one: the investor-founder relationship.I wouldn't claim it's painless. When I was running astartup, the thought of our investors used to keep me up at night.And now that I'm an investor,the thought of our startups keeps meup at night. All the pain of whatever problem you're trying tosolve is still there.But the pain hurts less when it isn'tmixed with resentment.I had the misfortune to participate in what amounted to a controlledexperiment to prove that. After Yahoo bought our startup I wentto work for them. I was doing exactly the same work, except withbosses. And to my horror I started acting like a child. The situation pushed buttons I'd forgottenI had.The big advantage of investment over employment, as the examples of opensource and blogging suggest, is that people working on projects oftheir own are enormously more productive. And astartup is a projectof one's own in two senses, both of them important: it's creativelyone's own, and also economically ones's own.Google is a rare example of a big company in tune with the forcesI've described. They've tried hard to make their offices less sterilethan the usual cube farm. They give employees who do great worklarge grants of stock to simulate the rewards of a startup. Theyeven let hackers spend 20% of their time on their own projects.Why not let people spend 100% of their time on their own projects,and instead of trying to approximate the value of what they create,give them the actual market value? Impossible? That is in factwhat venture capitalists do.So am I claiming that no one is going to be an employee anymore--that everyone should go and start a startup? Of course not.But more people could do it than do it now.At the moment, even the smartest students leave school thinkingthey have to get a job. Actually what they need to do is makesomething valuable. A job is one way to do that, but the moreambitious ones will ordinarily be better off taking money from aninvestor than an employer.Hackers tend to think business is for MBAs. But businessadministration is not what you're doing in a startup. What you'redoing is business creation. And the first phase of thatis mostly product creation-- that is, hacking. That's thehard part. It's a lot harder to create something people love thanto take something people love and figure out how to make money fromit.Another thing that keeps people away from starting startups is therisk. Someone with kids and a mortgage should think twice beforedoing it. But most young hackers have neither.And as the example of open source and blogging suggests, you'llenjoy it more, even if you fail. You'll be working on your ownthing, instead of going to some office and doing what you're told.There may be more pain in your own company, but it won't hurt asmuch.That may be the greatest effect, in the long run, of the forces underlying open source and blogging: finally ditching the oldpaternalistic employer-employee relationship, and replacing it witha purely economic one, between equals.Notes[1]Survey by Forrester Research reported in the cover story ofBusiness Week, 31 Jan 2005. Apparently someone believed you have toreplace the actual server in order to switch the operating system.[2]It derives from the late Latin tripalium,a torture device so called because it consisted of three stakes.I don't know how the stakes were used. "Travel" has the same root.[3]It would be much bigger news, in that sense, if the presidentfaced unscripted questions by giving a press conference.[4]One measure of the incompetence of newspapers is that so manystill make you register to read stories. I have yet to find a blogthat tried that.[5]They accepted the article, but I took so long tosend them the final version that by the time I did the section ofthe magazine they'd accepted it for had disappeared in a reorganization.[6]The word "boss" is derived from the Dutch baas, meaning"master."Thanks to Sarah Harlin, Jessica Livingston, and Robert Morris for reading drafts of this.
