The World is Flat

So a kid in India with a cheap PC can learn the inner workings of the same operating system that is running in some of the largest data centers of corporate America. Linux has an army of developers across the globe working to make it better. As I was working on this segment of the book, I went to a picnic one afternoon at the Virginia country home of Pamela and Malcolm Baldwin, whom my wife came to know through her membership on the board of World Learning, an educational NGO. I mentioned in the course of lunch that I was thinking of going to Mali to see just how flat the world looked from its outermost edge-the town of Timbuktu. The Baldwins' son Peter happened to be working in Mali as part of something called the GeekCorps, which helps to bring technology to developing countries. A few days after the lunch, I received an e-mail from Pamela telling me that she had consulted with Peter about accompanying me to Timbuktu, and then she added the following, which told me everything I needed to know and saved me the whole trip: “Peter says that his project is creating wireless networks via satellite, making antennas out of plastic soda bottles and mesh from window screens! Apparently everyone in Mali uses Linux...”

“Everyone in Mali uses Linux.” That is no doubt a bit of an exaggeration, but it's a phrase that you'd hear only in a flat world.

The free software movement has become a serious challenge to Microsoft and some other big global software players. As Fortune magazine reported on February 23, 2004, “The availability of this basic, powerful software, which works on Intel's ubiquitous microprocessors, coincided with the explosive growth of the Internet. Linux soon began to gain a global following among programmers and business users... The revolution goes far beyond little Linux... Just about any kind of software [now] can be found in open-source form. The SourceForge.net website, a meeting place for programmers, lists an astounding 86,000 programs in progress. Most are minor projects by and for geeks, but hundreds pack real value... If you hate shelling out $350 for Microsoft Office or $600 for Adobe Photoshop, OpenOffice.org and the Gimp are surprisingly high-quality free alternatives.” Big companies like Google, E*Trade, and Amazon, by combining Intel-based commodity server components and the Linux operating system, have been able dramatically to cut their technology spending-and get more control over their software.

Why would so many people be ready to write software that would be given away for free? Partly it is out of the pure scientific challenge, which should never be underestimated. Partly it is because they all hate Microsoft for the way it has so dominated the market and, in the view of many techies, bullied everyone else. Partly it is because they believe that open-source software can be kept more fresh and bugfree than any commercial software, because of the way it is constantly updated by an army of unpaid programmers. And partly it is because some big tech companies are paying engineers to work on Linux and other software, hoping it will cut into Microsoft's market share and make it a weaker competitor all around. There are a lot of motives at work here, and not all of them altruistic. When you put them all together, though, they make for a very powerful movement that will continue to present a major challenge to the whole commercial software model of buying a program and then downloading its fixes and buying its updates.

Until now, the Linux operating system was the best-known success among open-source free software projects challenging Microsoft. But Linux is largely used by big corporate data centers, not individuals. However, in November 2004, the Mozilla Foundation, a nonprofit group supporting open-source software, released Firefox, a free Web browser that New York Times technology writer Randall Stross (December 19, 2004) described as very fast and filled with features that Microsoft's Internet Explorer lacks. Firefox 1.0, which is easily installed, was released on November 9. “Just over a month later,” Stross reported, “the foundation celebrated a remarkable milestone: 10 million downloads.” Donations from Firefox's appreciative fans paid for a two-page advertisement in The New York Times. “With Firefox,” Stross added, “open-source software moves from back-office obscurity to your home, and to your parents', too. (Your children in college are already using it.) It is polished, as easy to use as Internet Explorer and, most compelling, much better defended against viruses, worms and snoops. Microsoft has always viewed Internet Explorer's tight integration with Windows to be an attractive feature. That, however, was before security became the unmet need of the day. Firefox sits lightly on top of Windows, in a separation from the underlying operating system that the Mozilla Foundation's president, Mitchell Baker, calls a 'natural defense.' For the first time, Internet Explorer has been losing market share. According to a worldwide survey conducted in late November by OneStat.com, a company in Amsterdam that analyzes the Web, Internet Explorer's share dropped to less than 89 percent, 5 percentage points less than in May. Firefox now has almost 5 percent of the market, and it is growing.”

It will come as no surprise that Microsoft officials are not believers in the viability or virtues of the free software form of open-source. Of all the issues I dealt with in this book, none evoked more passion from proponents and opponents than open-source. After spending time with the open-source community, I wanted to hear what Microsoft had to say, since this is going to be an important debate that will determine just how much of a flattener open-source becomes.

Microsoft's first point is, How do you push innovation forward if everyone is working for free and giving away their work? Yes, says Microsoft, it all sounds nice and chummy that we all just get together online and write free software by the people and for the people. But if innovators are not going to be rewarded for their innovations, the incentive for path-breaking innovation will dry up and so will the money for the really deep R & D that is required to drive progress in this increasingly complex field. The fact that Microsoft created the standard PC operating system that won out in the marketplace, it argues, produced the bankroll that allowed Microsoft to spend billions of dollars on R & D to develop Microsoft Office, a whole suite of applications that it can now sell for a little over $100.

“Microsoft would admit that there are number of aspects of the open-source movement that are intriguing, particularly around the scale, community collaboration, and communication aspects,” said Craig Mundie, the Microsoft chief technology officer. “But we fundamentally believe in a commercial software industry, and some variants of the open-source model attack the economic model that allows companies to build businesses in software. The virtuous cycle of innovation, reward, reinvestment, and more innovation is what has driven all big breakthroughs in our industry. The software business as we have known it is a scale economic business. You spend a ton of money up front to develop a software product, and then the marginal cost of producing each one is very small, but if you sell a lot of them, you make back your investment and then plow profits back into developing the next generation. But when you insist that you cannot charge for software, you can only give it away, you take the software business away from being a scale economic business.”

Added Bill Gates, “You need capitalism [to drive innovation.] To have [a movement] that says innovation does not deserve an economic reward is contrary to where the world is going. When I talk to the Chinese, they dream of starting a company. They are not thinking, 'I will be a barber during the day and do free software at night.'... When you have a security crisis in your [software] system, you don't want to say, 'Where is the guy at the barbershop?'”

As we move into this flat world, and you have this massive Web-enabled global workforce, with all these collaborative tools, there will be no project too small for some members of this workforce to take on, or copy, or modify-for free. Someone out there will be trying to produce the free versions of every kind of software or drug or music. “So how will products retain their value?” asked Mundie. “And if companies cannot derive fair value from their products, will innovation move forward in this area, or others, at the speed that it could or should?” Can we always count on a self-organizing open-source movement to come together to drive things forward for free?

It seems to me that we are too early in the history of the flattening of the world to answer these questions. But they will need answers, and not just for Microsoft. So far-and maybe this is part of the long-term answer-Microsoft has been able to count on the fact that the only thing more expensive than commercial software is free software. Few big companies can simply download Linux off the Web and expect it to work for all their tasks. A lot of design and systems engineering needs to go around it and on top of it to tailor it to a company's specific needs, especially for sophisticated, large-scale, mission-critical operations. So when you add up all the costs of adapting the Linux operating system to the needs of your company and its specific hardware platform and applications, Microsoft argues, it can end up costing as much as or more than Windows.

The second issue Microsoft raises about this whole open-source movement has to do with how we keep track of who owns which piece of any innovation in a flat world, where some is generated for free and others build on it for profit. Will Chinese programmers really respect the rules of the Free Software Foundation? Who will govern all this?

“Once you start to socialize the global population on the idea that software or any other innovation is supposed to be free, a lot of people will not distinguish between free software, free pharmaceuticals, free music, or free patents on car designs,” argued Mundie. There is some truth to this. I work for a newspaper, that is where my paycheck comes from. But I believe that all online newspapers should be free, and on principle I refuse to pay for an online subscription to The Wall Street Journal. I have not read the paper copy of The New York Times regularly for two years. I read it only online. But what if my daughters' generation, which is being raised to think that newspapers are something to be accessed online for free, grows up and refuses to pay for the paper editions? Hmmm. I loved Amazon.com until it started providing a global platform that wasn't selling only my new books but also used versions. And I am still not sure how I feel about Amazon offering sections of this book to be browsed online for free.

Mundie noted that a major American auto company recently discovered that some Chinese firms were using new digital-scanning technology to scan an entire car and churn out computer-aided design models of every part within a very short period of time. They can then feed those designs to industrial robots and in short order produce a perfect copy of a GM car-without having to spend any money on R & D. American automakers never thought they had anything to worry about from wholesale cloning of their cars, but in the flat world, given the technologies that are out there, that is no longer the case.

My bottom line is this: Open-source is an important flattener because it makes available for free many tools, from software to encyclopedias, that millions of people around the world would have had to buy in order to use, and because open-source network associations-with their open borders and come-one-come-all approach-can challenge hierarchical structures with a horizontal model of innovation that is clearly working in a growing number of areas. Apache and Linux have each helped to drive down costs of computing and Internet usage in ways that are profoundly flattening. This movement is not going away. Indeed, it may just be getting started-with a huge, growing appetite that could apply to many industries. As The Economist mused (June 10, 2004), “some zealots even argue that the open-source approach represents a new, post-capitalist model of production.”

That may prove true. But if it does, then we have some huge global governance issues to sort out over who owns what and how individuals and companies will profit from their creations.

Flattener #5: Outsourcing, Y2K

India has had its ups and down since it achieved independence on August 15, 1947, but in some ways it might be remembered as the luckiest country in the history of the late twentieth century.

Until recently, India was what is known in the banking world as “the second buyer.” You always want to be the second buyer in business-the person who buys the hotel or the golf course or the shopping mall after the first owner has gone bankrupt and its assets are being sold by the bank at ten cents on the dollar. Well, the first buyers of all the cable laid by all those fiber-optic cable companies-which thought they were going to get endlessly rich in an endlessly expanding digital universe-were their American shareholders. When the bubble burst, they were left holding either worthless or much diminished stock. The Indians, in effect, got to be the second buyers of the fiber-optics companies.

They didn't actually purchase the shares, they just benefited from the overcapacity in fiber optics, which meant that they and their American clients got to use all that cable practically for free. This was a huge stroke of luck for India (and to a lesser degree for China, the former Soviet Union, and Eastern Europe), because what is the history of modern India? In short, India is a country with virtually no natural resources that got very good at doing one thing-mining the brains of its own people by educating a relatively large slice of its elites in the sciences, engineering, and medicine. In 1951, to his enduring credit, Jawaharlal Nehru, India's first prime minister, set up the first of India's seven Indian Institutes of Technology (IIT) in the eastern city of Kharagpur. In the fifty years since then, hundreds of thousands of Indians have competed to gain entry and then graduate from these IITs and their private-sector equivalents (as well as the six Indian Institutes of Management, which teach business administration). Given India's 1 billion-plus population, this competition produces a phenomenal knowledge meritocracy. It's like a factory, churning out and exporting some of the most gifted engineering, computer science, and software talent on the globe.

This, alas, was one of the few things India did right. Because its often dysfunctional political system, coupled with Nehru's preference for pro-Soviet, Socialist economics, ensured that up until the mid-1990s India could not provide good jobs for most of those talented engineers. So America got to be the second buyer of India's brainpower! If you were a smart, educated Indian, the only way you could fulfill your potential was by leaving the country and, ideally, going to America, where some twenty-five thousand graduates of India's top engineering schools have settled since 1953, greatly enriching America's knowledge pool thanks to their education, which was subsidized by Indian taxpayers.

“The IITs became islands of excellence by not allowing the general debasement of the Indian system to lower their exacting standards,” noted The Wall Street Journal (April 16, 2003). “You couldn't bribe your way to get into an IIT... Candidates are accepted only if they pass a grueling entrance exam. The government does not interfere with the curriculum, and the workload is demanding... Arguably, it is harder to get into an IIT than into Harvard or the Massachusetts Institute of Technology... IIT alumnus Vinod Khosla, who co-founded Sun Microsystems, said: 'When I finished IIT Delhi and went to Carnegie Mellon for my Masters, I thought I was cruising all the way because it was so easy relative to the education I got at IIT.'”

For most of their first fifty years, these IITs were one of the greatest bargains America ever had. It was as if someone installed a brain drain that filled up in New Delhi and emptied in Palo Alto.

And then along came Netscape, the 1996 telecom deregulation, and Global Crossing and its fiber-optic friends. The world got flattened and that whole deal got turned on its head. “India had no resources and no infrastructure,” said Dinakar Singh, one of the most respected young hedge fund managers on Wall Street, whose parents graduated from an IIT and then immigrated to America, where he was born. “It produced people with quality and by quantity. But many of them rotted on the docks of India like vegetables. Only a relative few could get on ships and get out. Not anymore, because we built this ocean crosser, called fiberoptic cable... For decades you had to leave India to be a professional... Now you can plug into the world from India. You don't have to go to Yale and go to work for Goldman Sachs [as I did.]”

India could never have afforded to pay for the bandwidth to connect brainy India with high-tech America, so American shareholders paid for it. Sure, overinvestment can be good. The overinvestment in railroads turned out to be a great boon for the American economy. “But the railroad overinvestment was confined to your own country and so too were the benefits,” said Singh. In the case of the digital railroads, “it was the foreigners who benefited.” India got to ride for free.

It is fun to talk to Indians who were around at precisely the moment when American companies started to discover they could draw on India's brainpower in India. One of them is Vivek Paul, now the president of Wipro, the Indian software giant. “In many ways the Indian information technology [outsourcing] revolution began with General Electric coming over. We're talking the late 1980s and early '90s. At the time, Texas Instruments was doing some chip design in India. Some of their key designers [in America] were Indians, and they basically let them go back home and work from there [using the rather crude communications networks that existed then to stay in touch.] At that time, I was heading up the operations for GE Medical Systems in Bangalore. [GE's chairman] Jack Welch came to India in 1989 and was completely taken by India as a source of intellectual advantage for GE. Jack would say, 'India is a developing country with a developed intellectual capability.' He saw a talent pool that could be leveraged. So he said, 'We spend a lot of money doing software. Couldn't we do some work for our IT department here?'” Because India had closed its market to foreign technology companies, like IBM, Indian companies had started their own factories to make PCs and servers, and Welch felt that if they could do it for themselves, they could do it for GE.

To pursue the project, Welch sent a team headed by GE's chief information officer over to India to check out the possibilities. Paul was also filling in as GE's business development manager for India at the time. “So it was my job to escort the corporate CIO, in early 1990, on his first trip,” he recalled. “They had come with some pilot projects to get the ball rolling. I remember in the middle of the night going to pick them up at the Delhi airport with a caravan of Indian cars, Ambassadors, based on a very dated 1950s Morris Minor design. Everyone in the government drove one. So we had a five-car caravan and we were driving back from the airport to town. I was in the back car, and at one point we heard this loud bang, and I thought, What happened? I shot to the front, and the lead car's hood had flown off and smashed the windshield-with these GE people inside! So this whole caravan of GE execs pulls over to the side of the road, and I could just hear them saying to themselves, 'This is the place we're going to get software from?'”

Fortunately for India, the GE team was not discouraged by the poor quality of Indian cars. GE decided to sink roots, starting a joint development project with Wipro. Other companies were trying different models. But this was still pre-fiber-optic days. Simon & Schuster, the book publisher, for instance, would ship its books over to India and pay Indians $50 a month (compared to $1,000 a month in the United States) to type them by hand into computers, converting the books into digitized electronic files that could be edited or amended easily in the future—particularly dictionaries, which constantly need updating. In 1991, Manmohan Singh, then India's finance minister, began opening the Indian economy for foreign investment and introducing competition into the Indian telecom industry to bring down prices. To attract more foreign investment, Singh made it much easier for companies to set up satellite downlink stations in Bangalore, so they could skip over the Indian phone system and connect with their home bases in America, Europe, or Asia. Before then, only Texas Instruments had been willing to brave the Indian bureaucracy, becoming the first multinational to establish a circuit design and development center in India in 1985. TI's center in Bangalore had its own satellite downlink but had to suffer through having an Indian government official to oversee it-with the right to examine any piece of data going in or out. Singh loosened all those reins post-1991. A short time later, in 1994, HealthScribe India, a company originally funded in part by Indian-American doctors, was set up in Bangalore to do outsourced medical transcription for American doctors and hospitals. Those doctors at the time were taking handwritten notes and then dictating them into a Dictaphone for a secretary or someone else to transcribe, which would usually take days or weeks. HealthScribe set up a system that turned a doctor's touch-tone phone into a dictation machine. The doctor would punch in a number and simply dictate his notes to a PC with a voice card in it, which would digitize his voice. He could be sitting anywhere when he did it. Thanks to the satellite, a housewife or student in Bangalore could go into a computer and download that doctor's digitized voice and transcribe it-not in two weeks but in two hours. Then this person would zip it right back by satellite as a text file that could be put into the hospital's computer system and become part of the billing file. Because of the twelve-hour time difference with India, Indians could do the transcription while the American doctors were sleeping, and the file would be ready and waiting the next morning. This was an important breakthrough for companies, because if you could safely, legally, and securely transcribe from Bangalore medical records, lab reports, and doctors' diagnoses-in one of the most litigious industries in the world-a lot of other industries could think about sending some of their backroom work to be done in India as well. And they did. But it remained limited by what could be handled by satellite, where there was a voice delay. (Ironically, said Gurujot Singh Khalsa, one of the founders of HealthScribe, they initially explored having Indians in Maine-that is, American Indians-do this work, using some of the federal money earmarked for the tribes to get started, but they could never get them interested enough to put the deal together.) The cost of doing the transcription in India was about one-fifth the cost per line of doing it the United States, a difference that got a lot of people's attention.

By the late 1990s, though, Lady Luck was starting to shine on India from two directions: The fiber-optic bubble was starting to inflate, linking India with the United States, and the Y2K computer crisis-the so-called millennium bug-started gathering on the horizon. As you'll remember, the Y2K bug was a result of the fact that when computers were built, they came with internal clocks. In order to save memory space, these clocks rendered dates with just six digits-two for the day, two for the month, and, you guessed it, two for the year. That meant they could go up to only 12/31/99. So when the calendar hit January 1, 2000, many older computers were poised to register that not as 01/01/2000 but as 01/01/00, and they would think it was 1900 all over again. It meant that a huge number of existing computers (newer ones were being made with better clocks) needed to have their internal clocks and related systems adjusted; otherwise, it was feared, they would shut down, creating a global crisis, given how many different management systems-from water to air traffic control-were computerized.

This computer remediation work was a huge, tedious job. Who in the world had enough software engineers to do it all? Answer: India, with all the techies from all those IITs and private technical colleges and computer schools.

And so with Y2K bearing down on us, America and India started dating, and that relationship became a huge flattener, because it demonstrated to so many different businesses that the combination of the PC, the Internet, and fiber-optic cable had created the possibility of a whole new form of collaboration and horizontal value creation: outsourcing.

Any service, call center, business support operation, or knowledge work that could be digitized could be sourced globally to the cheapest, smartest, or most efficient provider. Using fiber-optic cable-connected workstations, Indian techies could get under the hood of your company's computers and do all the adjustments, even though they were located halfway around the world.

“[Y2K upgrading] was tedious work that was not going to give them an enormous competitive advantage,” said Vivek Paul, the Wipro executive whose company did some outsourced Y2K drudge work. “So all these Western companies were incredibly challenged to find someone else who would do it and do it for as little money as possible. They said, 'We just want to get past the damn year 2000!' So they started to work with Indian [technology] companies who they might not have worked with otherwise.”

To use my parlance, they were ready to go on a blind date with India. They were ready to get “fixed up.” Added Jerry Rao, 'Y2K means different things to different people. For Indian industry, it represented the biggest opportunity. India was considered as a place of backward people. Y2K suddenly required that every single computer in the world needed to be reviewed. And the sheer number of people needed to review line-by-line code existed in India. The Indian IT industry got its footprint across the globe because of Y2K. Y2K became our engine of growth, our engine of being known around the world. We never looked back after Y2K.“

By early 2000, the Y2K work started to wind down, but then a whole new driver of business emerged-e-commerce. The dot-com bubble had not yet burst, engineering talent was scarce, and demand from dotcoms was enormous. Said Paul, “People wanted what they felt were mission-critical applications, key to their very existence, to be done and they could go nowhere else. So they turned to the Indian companies, and as they turned to the Indian companies they found that they were getting delivery of complex systems, with great quality, sometimes better than what they were getting from others. That created an enormous respect for Indian IT providersf.] And if [Y2K work] was the acquaintanceship process, this was the falling-in-love process.”

Outsourcing from America to India, as a new form of collaboration, exploded. By just stringing a fiber-optic line from a workstation in Bangalore to my company's mainframe, I could have Indian IT firms like Wipro, Infosys, and Tata Consulting Services managing my e-commerce and mainframe applications.

“Once we're in the mainframe business and once we're in e-commerce—now we're married,” said Paul. But again, India was lucky that it could exploit all that undersea fiber-optic cable. “I had an office very close to the Leela Palace hotel in Bangalore,” Paul added. “I was working with a factory located in the information technology park in Whitefield, a suburb of Bangalore, and I could not get a local telephone line between our office and the factory. Unless you paid a bribe, you could not get a line, and we wouldn't pay. So my phone call to Whitefield would go from my office in Bangalore to Kentucky, where there was a GE mainframe computer we were working with, and then from Kentucky to Whitefield. We used our own fiber-optic lease line that ran across the ocean-but the one across town required a bribe.”

India didn't benefit only from the dot-com boom; it benefited even more from the dot-com bust! That is the real irony. The boom laid the cable that connected India to the world, and the bust made the cost of using it virtually free and also vastly increased the number of American companies that would want to use that fiber-optic cable to outsource knowledge work to India.

Y2K led to this mad rush for Indian brainpower to get the programming work done. The Indian companies were good and cheap, but price wasn't first on customers' minds-getting the work done was, and India was the only place with the volume of workers to do it. Then the dot-com boom comes along right in the wake of Y2K, and India is one of the few places where you can find surplus English-speaking engineers, at any price, because all of those in America have been scooped up by e-commerce companies. Then the dot-com bubble bursts, the stock market tanks, and the pool of investment capital dries up. American IT companies that survived the boom and venture capital firms that still wanted to fund start-ups had much less cash to spend. Now they needed those Indian engineers not just because there were a lot of them, but precisely because they were low-cost. So the relationship between India and the American business community intensified another notch.

One of the great mistakes made by many analysts in the early 2000s was conflating the dot-com boom with globalization, suggesting that both were just fads and hot air. When the dot-com bust came along, these same wrongheaded analysts assumed that globalization was over as well. Exactly the opposite was true. The dot-com bubble was only one aspect of globalization, and when it imploded, rather than imploding globalization, it actually turbocharged it.