Chapter Ten
Marx, Lenin, and Gates: Failed Counterrevolutions

Stopping an avalanche is a losing proposition, but that hasn’t stopped people, companies, and nations from trying. In this chapter, I’m going to discuss the strategy of resistance—the strategy of trying to stop or reverse disaggregation. There have been a few successes. Well… not really successes. I don’t think I’ve found any examples of permanent victories over disaggregation. What I’ve found is that suppressing disaggregation usually relies on force, and, although force is very effective in the short run, in the long run it’s almost impossible to permanently suppress a good idea. Let’s look at some of the examples, starting with an experiment in reaggregation that killed millions and millions of people.

Karl Marx, Vladimir Lenin, and the Nuts and Bolts of Socialism

One of the worst regimes of the twentieth century was that of the Soviet Union. The Soviet Union dominated an empire that stretched across Asia into Europe. The government deliberately murdered tens of millions of its own citizens and sent many others into a vast system of internal prison camps. Why? What drove the Soviet Union to such extreme behavior?

They weren’t simply sadists. The problem was that Soviet ideology, Marxism-Leninism, required that the Soviet Union overturn two fundamental disaggregations that are crucial to modern civilization: corporations and money. To reverse these disaggregations required massive force because doing away with the disaggregation meant doing away with the benefits. Just what are the benefits of corporations and of money?

Corporations, Trust, and Sharing

A modern semiconductor factory costs several billion dollars. How can any one person possibly afford to build or buy a semiconductor factory? The answer is that no one person can and no one person does. People form groups to raise the money, and the fundamental structure people use to manage and spend this kind of money is the corporation.

Corporations disaggregate both ownership and authority to solve the problems of large-scale enterprises. Disaggregation of ownership solves the problem of how to define ownership when there are millions of owners. Corporate rules govern how each shareholder exercises their property rights—for example, at the end of the year when it’s time to divide the profits, shareholders can’t just walk into the factory and make off with whatever they think is their fair portion. The disaggregation of ownership leads to sharing, and the corporation lets people who will never even meet each other pool their money, create enterprises, and generate more wealth.

Disaggregation of authority prevents chaos. When I buy stock in a corporation, I become a partial owner and I’m entitled to certain rights—but I can’t waltz into the factory and tell the workers what to do. Corporate rules about who has what authority, despite their ownership in the corporation, leads to trust that the corporation will be run properly.

The creation of the corporation resulted in thousands of innovative ideas in the world of finance—new ideas on how to share ownership and build trust. If you want to build, equip, and staff a factory, you have dozens if not hundreds of financial options to choose from.109

Karl Marx proposed to undo these disaggregations and Vladimir Lenin carried out the plan. Marxism-Leninism had no room for corporations; workers had to own the factories, or rather the state would, acting as proxy. Both ownership and authority were reaggregated: the State built, owned, and operated the factory. By disposing with corporations and all the creativity they allowed, Marxism-Leninism made the financial problems of building decent factories almost insoluble. Disaggregation of authority creates trust; it follows that reaggregation of authority dissolves trust, and the disastrous record of state-run factories completely justified that lack of trust.

Even more damaging was the Marx/Lenin doctrine that tried to undo money.

Money and Information

You know that phrase “reinventing the wheel”? Some innovations are so basic to human society and are so important that they’ve been invented and reinvented time and again throughout history. The wheel would seem to be one of them, but it isn’t—human cultures that were cut off from Europe and Asia, such as the civilizations in the Americas, never invented the wheel. What’s more important than the wheel? Money. Money has been reinvented, time and again, all across history.

Money disaggregates value from the thing itself, and this disaggregation has been applied to all sorts of things—from goods, services, and intangibles (e.g., fishing rights) to concepts and emotions (“pain and suffering”). How money works is pretty straightforward. People pick some scarce resource—salt; beads; cowrie shells; large rocks; brightly colored feathers; or precious metals, such as copper, silver, and gold—and trade those scarce items instead of the actual commodities. All that matters is that the “money” be scarce; what’s scarce is irrelevant. The United States dollar, for example, isn’t backed by silver or gold; objectively speaking, about 10% of the U.S. money supply consists of worthless pieces of scrap paper and the remaining 90% is even more worthless electronic bank records. What makes the U.S. dollar work is the U.S. government. The government keeps the supply of dollars sufficiently scarce, everyone accepts the fiction that U.S. dollars are valuable—and thus they are, because people value them.110

International commerce, banking, and the stock market all would be impossible without money. Money makes everyday life easier as well; it’s relatively simple to design a vending machine that accepts coins, but a vending machine that accepts live chickens would be much more difficult.

A lesser-known but absolutely vital function of money is to provide prices and thereby to quantify value. Prices are an interface, in our sense of the word—the economy consists of thousands of disaggregated businesses, and we know that disaggregated pieces require an interface to let them work together. Prices provide that interface; prices provide a way to convey extremely valuable information between businesses.111

Let’s say I own a tractor factory. Since I need nuts and bolts for the tractors, I call a few nut and bolt manufacturers and outline what I need: the sizes, schedule, and quality. The manufacturers will offer bids. If my requirements are too strict or if I’ve asked for something difficult or too unusual, the price will be higher and that will tell me that I’ve either got to adjust my requirements or give up some of my profits. But my nut and bolt problems are solved—I specialize in building and selling tractors, and someone else specializes in making and selling nuts and bolts. Money and prices provide the interface and let us work together.

What’s really amazing is that the process of supplying nuts and bolts for all of society’s needs is managed automatically through the price interface. Except in extraordinary circumstances, who has ever heard of a nut and bolt shortage in a modern capitalist society? If the demand for nuts and bolts goes up, the factory owners expand or build new factories. If the demand goes down, factories lower their prices, find new markets, or close. If a factory produces lousy products, fails to meet its delivery schedules, or provides poor service, customers leave and the problem corrects itself one way or another.

Marxism-Leninism and the Lack of Prices

The Soviet Union’s Marxist-Leninist government transformed prices into a meaningless exercise, a number set by governmental fiat. It’s not that they forced everyone to dispense with money; what they did was destroy the relationship between value, as measured by prices, and the underlying objects. With price’s information gone, the interface to allocate goods and services was gone. The Soviet government substituted central economic planning for prices—which failed, miserably. The Soviet Union had vast reaches of fertile farmland but couldn’t feed their own population. They had some of the greatest deposits of natural resources in the entire world—timber, oil, gold—but were incapable of exploiting them. Ask someone who lived there in the “good old days” of Communism about the shortages and shoddy quality of even the most basic consumer goods.

Let’s say I live in the Soviet Union after the Marxist-Leninist revolution. How would I manage that tractor factory? When I need nuts and bolts, I request them from the Glorious Heroes of the Revolution Nuts and Bolts Factory, which produces them according to quotas set by the State Planning Committee under the authority of the Council of Ministers. This system leads to two immediate problems. The first is that, with quotas set by government agencies, I can only hope that the agency anticipated all of society’s needs in advance when it laid out the current Five-Year Plan.

Any error could mean widespread disruption in the economy. It goes without saying that, even in an honest and open society, perfect information and perfect forecasting would be impossible; in a police state like the former Soviet Union, the idea would be laughable if the results weren’t so utterly tragic.

The other immediate problem is that the nut and bolt factory doesn’t earn a profit by supplying me with wonderful products and excellent customer service; their goal is to meet—or, better yet, evade—their sometimes wildly unrealistic quotas with the absolute minimum effort. (Under Stalin, another important job goal of managers was to avoid being shot, as an example to other managers, for poor performance.)

How did this play out in real Soviet life? The shoe industry provides a hint. The shoe manufacturers turned out plenty of shoes each year, three pairs for every person—but people still waited in line to buy shoes, mostly imported ones. The problem was that Soviet shoes were the wrong size or wrong style. The factories found it easier and more economical to set up their machines for a single size and style and run off as many pairs as they could get away with. If the quota was set by the number of shoes, smaller shoes were easier to produce; if the quota was set by tonnage, fewer but larger shoes were easier to produce. And as for what people actually wanted to wear in terms of style and color, that wasn’t relevant to the quota.112

Back to my hypothetical tractor factory. The nut and bolt industry also produced according to quotas—which meant that my factory couldn’t rely on their shipments. As manager, if I can’t rely on nuts and bolts from elsewhere and I have to deliver my quota of tractors, I’ll manufacture my own instead. That’s exactly what happened, in industry after industry. In the Soviet Union, no more than half of the nuts and bolts were made in the official factories. The rest were made in captive facilities—an industry, such as the tractor industry, would set up its own plant to make nuts and bolts. Hard as it is to believe, one of the earliest disaggregations of the modern era, that of nuts and bolts, became undone.

Every industry was cut off from every other industry, as though they were stranded alone on a desert island, and, like castaways, each industry built the necessities of life from scratch. The simple commodities that Western manufacturers take for granted—nuts and bolts, bricks, paint—were made in endlessly duplicated facilities. Aside from the horrible inefficiency, I believe that this lack of disag-gregation had another effect. No disaggregation means no benefits of disaggregation, and in particular the benefit of specialization—in this case, manufacturing expertise in how to make nuts and bolts, bricks, and paint to the highest quality standards. Back in my graduate student days, as we would build experimental equipment in the machine shop, we used to joke about the poor craftsmanship of Soviet products: the hammer and sickle on the Soviet Union’s flag stood for their low-tech production technique, “cut to fit and beat to match.”

We were wrong about the hammer and sickle. Marxism-Leninism’s inefficient system of central planning brought death, starvation, terror, deprivation, and poverty to the Soviet Union. Creativity, synergy, competitiveness—all of these were inevitably damaged or destroyed with the imposition of Marxist-Leninist doctrine. The unnatural act of rolling back basic innovations and their benefits led the Soviet government to oppress their own people because such a colossal failure could be imposed only by a police state. The hammer and sickle ultimately came to mean, “Hammer your population and cut down the ones who resist.”113

Bill Gates: Running the Wheels Backwards

I’ve seen Microsoft in action, and it’s really, really not very pretty. A few years ago I was at a conference for speech technologies. We were excited over a new standard for speech technology from the World Wide Web Consortium (W3C), one that would make it infinitely easier to write applications; easier applications would mean more applications, and that would be mean more business. To my great dismay, Microsoft came along with their own “standard” and muddied the water. Their marketing material—handed out during the presentations!—deliberately attacked the W3C standard, a breach of professional decorum that absolutely shocked me. The materials focused on fear, uncertainty, and doubt: which standard should I adopt; is it true the W3C standard is incomplete and unsound; which one will be adopted by customers, the one created by the speech industry, or Microsoft’s “standard”? A colleague who’d worked hard on the W3C industry standard was almost speechless with anger.

When I talk to people who work for Microsoft, they tell me that they make the best products in the world. In contrast, government prosecutors and private companies all over the world have taken Microsoft to court over their business practices and have complained about illegal behavior. If their stuff is so good, why does Microsoft act like a bully? I’ve heard the usual theory, that Microsoft has an aggressive internal culture that leads them to push the envelope and do things that other companies might avoid. I think it’s a more fundamental problem, the same problem that the Soviet Union had. I think Microsoft is attempting to run the wheels backwards and do away with innovations they don’t like, and the only way to accomplish that is to use force.

Here’s the irony: Microsoft was founded on a series of fundamental disaggregations, and now they’re trying to do away with some of them. One of the most fundamental disaggregations of computer science today is the disaggregation of operating systems and applications. Applications are the programs that do the useful things you bought the computer for in the first place: writing documents, browsing the Web, and playing video games. The “operating system” (e.g., Windows XP) takes care of the housekeeping of the computer: it figures out what you’re typing at the keyboard and is in charge of keeping track of files on the disk drives.114

The separation of operating system from applications offers all the usual benefits we expect from disaggregation. It makes for simplicity because each program can focus on doing just one task correctly; for more reliable programs because an application error shouldn’t crash the operating system and thus the entire computer; and for better security because the operating system can defend the computer against rogue applications.

And then there’s that problem benefit: competition—at least, it’s a problem from Microsoft’s perspective. When application programs are separate from the operating system, people are free to choose which programs they want to purchase—just because they purchase a Microsoft operating system doesn’t mean they absolutely must purchase Microsoft applications.

If you control the operating system, there are tactics you can use to reaggregate applications with the operating system and to nudge customers toward buying your applications. First, there’s outright sabotage: change something in the operating system to make applications from other companies fail. Microsoft was accused of this for years. Another tactic is to change the way the operating system works with applications—the so-called API, or application programming interface. If Microsoft makes a few changes in the API and simultaneously releases new versions of its software, Microsoft’s software will work—while other software vendors scramble to fix their software when they finally discover what’s happened to the API. Then there’s the “hidden API” tactic, where Microsoft keeps some of the API secret. Competitors worry that Microsoft developers can use the hidden API to build speedier or slicker applications.

And the most retrograde tactic of all is barefaced reaggregation. In the most infamous and blatant example, Microsoft integrated the Internet Explorer Web browser into the Windows operating system. This maneuver landed Microsoft in hot water; after a protracted court battle with the U.S. Department of Justice, Microsoft ended up with a slap on the wrist and the lion’s share of the Web browser market.115

Embrace, Extend, Extinguish

These tactics show how someone with monopoly or near-monopoly control of a crucial piece of technology can fight off disaggregation, but actually the lesson goes deeper than this. Microsoft gave Internet Explorer away for free even though Microsoft’s competitors were, for the most part, charging for their Web browsers. Why did Microsoft use reaggregation to try to lock other browsers out of the desktop? If Microsoft was just giving away the software anyway, why did they care about competing browsers?

Web browsers are a gateway to the World Wide Web and the rest of the Internet. If Microsoft were to gain complete control of the Web browsers, they could use that control as a wedge to capture the World Wide Web—to take it away from the international community and bring it under Microsoft’s control. This tactic, which perhaps you might recognize from your industry, is often called “embrace, extend, and extinguish.” If you’re the dominant player in the industry, here’s how you can use this tactic.

• Embrace an industry standard by incorporating it into your product.
• Extend the standard by adding nonstandard, proprietary extensions under your exclusive control.
• Extinguish the standard. Make the proprietary extensions important or vital in some way—for example, make them part of your other products, and use your market power to push them into the marketplace. Eventually, anyone who wants to be part of the marketplace has to adopt the extensions. The industry ends up with a “standard” that is completely controlled by you, the dominant player. Competitors struggle, but lack of access to the proprietary technology in the “standard” places them at a disadvantage.

In the case of Internet Explorer, Microsoft added nonstandard extensions—well, I should be careful. When I talk to folks from Microsoft, they call anything that Microsoft uses an “industry standard”; what I mean by “nonstandard” is that Microsoft added its proprietary extensions. Some businesses followed Microsoft’s lead and started to use the proprietary extensions, which had two implications. First, people had to use Internet Explorer to view those Web pages. Second, the Web pages couldn’t be viewed at all on computers that used the Unix operating system. I used to run across Web pages like that from time to time, and all I could do was throw my hands up in the air.116

That’s a crucial point—part of the Web became unusable if you didn’t use Microsoft products. The rest of the scenario is easy to project: at some point, Internet Explorer would start to use proprietary technology to interact with Web servers—Microsoft Web servers. If everyone uses Internet Explorer, and Internet Explorer uses proprietary technology that works only with Microsoft Web servers, then all businesses on the Web must use Microsoft’s Web server software, which does cost money and generates a nice profit for Microsoft. Fortunately, this never came to pass: the Internet standards remained intact. At the time of this writing, the vast majority of Web servers don’t use Microsoft Web server software, Internet Explorer uses standard technology to interact with Web servers, and most businesses don’t use the proprietary features of Internet Explorer—because their customers complained.

Another way to use the concept of embrace, extend, and extinguish is to sabotage competitors. Microsoft tried this against the Java programming language, which Microsoft quite correctly viewed as a threat. The Java programming language, from Sun Microsystems, lets you write a program once and have it run, without any changes, on any computer: Windows, Macintosh, or Linux. Java is even built into many cell phones. This undermines the Windows operating system: if the application is completely independent of the operating system, why bother with Windows? Microsoft embraced Java and then extended it with Microsoft-specific stuff, in violation of their Java licensing agreement. The extensions made Microsoft’s Java incompatible with other Java, with the result that programs lost the portability between operating systems that was the whole point of Java in the first place.

Microsoft had picked a fight with Sun Microsystems, the owner of Java and a company with deep pockets of its own that could afford to take Microsoft to court. After years of litigation, Microsoft and Sun eventually settled; Microsoft paid Sun about $2 billion to settle various antitrust and patent claims.117

The Open Source Challenge to Microsoft

The open source software movement has a unique answer to Microsoft’s counterrevolution: they’re fighting reaggregation with more disaggregation.

Microsoft is a classic software company, and classic companies all share certain features: in-house experts (“gurus”) decide what the software should do; legions of programmers write the software’s “source code”; the source code is a guarded, highly proprietary secret; in-house specialists test the software; marketing finds customers; and technical staff members write documentation and help customers. Managers exercise strict control over productivity and schedule.

The open source movement challenges classical software companies. The open source movement has a completely different philosophy, which in turn leads to a completely different way of writing software. The open source software movement believes that source code should be available for public scrutiny, and many adherents believe, in addition, that software ought to be freely available—literally without charge. These ideas lead directly to a new way to produce software, a kind of disaggregated software company: designing, writing, testing, documenting, marketing, selling, distributing, and maintaining software are all separate functions and can be performed by separate individuals who don’t necessarily coordinate their activities. The work isn’t necessarily performed by companies, even though it often is; fantastic amounts of work are done by volunteers who are simply interested in the software.

The most famous open source project is Linux, an operating system that can be used instead of Microsoft Windows. Linux dominates the Internet infrastructure. Linux continues to make inroads into the backoffice operations of businesses, a market that Microsoft lusts after. Linux even poses a terrific challenge to Microsoft Windows for the ordinary desktop computer because of all the advantages Linux offers.

• Lower Costs. Windows is a single-source product available only from Microsoft, which allows Microsoft to control the price of Windows. Linux, on the other hand, is available from many different places, and many versions of Linux are available for free. These free versions provide the baseline against which any for-profit Linux product or business must compete, and the result is that Linux software is less expensive than Windows software.118
• Better quality. Linux is different than traditional software projects. Most software companies have relatively few people to review the software or test it. Even if the company has thousands of programmers, the programmers all work on different projects or sections of the software, and any particular piece of code is seen only by a relative handful of programmers. Since the source code is a proprietary secret, only company insiders can debug the source code. In contrast, Linux doesn’t hide its source code; ownership is disaggregated from any individual. Every time software is released, thousands of knowledgeable individuals compile and run the latest version. With thousands of “eyeballs” looking for errors, it’s not surprising that Linux finds and fixes bugs rapidly; the fixes are transparent—no mysterious, unexplained changes.
• Superb security. Linux has major advantages over Windows in the area of security; here are three of them. The first is that Linux is designed in a disaggregated but orderly fashion, so security errors remain isolated, unlike Windows. Second, Linux has thousands of eyeballs—knowledgeable users who test Linux in a huge variety of settings and with many different tools. Security errors tend to be caught quickly and fixed quickly. Third, anyone who is concerned about potential security flaws can inspect Linux software for themselves—they don’t have to rely on assurances from others.

If you want to stop a project like Linux, which competes directly with Windows, there’s no single company to crush or buy out—thousands of experienced programmers around the world contribute time and energy to Linux, and they can’t be bribed or intimidated. How can Microsoft defend its monopoly against this disaggregated structure?

One Microsoft tactic has been a campaign to discredit Linux. The fun started with a list of weird accusations against open source: Open Source software is a “cancer”; Open Source software is un-American; Open Source endangers your company’s intellectual property. After that stage fizzled out, Microsoft started on the next round, which purported to show that open source software’s quality isn’t as high as Microsoft’s and that it costs more; critics and governmental authorities denounced Microsoft advertisements as misleading. Lately, Microsoft claimed that their Shared Source program is the same as open source, while casting aspersions on open source. This is a classic FUD campaign—casting fear, uncertainty, and doubt—and Microsoft is very good at it.119

Which will win, Microsoft or Linux? Linux continues to gain momentum; at the same time, Microsoft is fiercely defending its turf on the desktop. Rather than fight power with power, the open source movement in general and Linux in particular follow a different path. Open source disaggregates the entire model of how software is written, and has created a production method that can never be suppressed or dominated; they’ve shifted the battleground. They’ve left the monopoly frantically trying to stem the avalanche before the avalanche sweeps aside the monopoly’s entire business.