The story of technology is largely the story of people who guess wrong about which problems will be easy to solve and which will be hard. For example, less than a decade before the Wright Brothers’ flight, Lord Kelvin announced that “heavier than air flying machines are impossible.” The Scientific American of 1909 concluded that “the automobile has practically reached the limit of its development,” on evidence “that during the past year no improvements of a radical nature have been introduced.” Thomas Watson of IBM famously said in the 1940s that “there is a world market for maybe five computers.”1 In The Road Ahead, published in 1996, Bill Gates of Microsoft said that his—or anyone’s—predictive writings about technology were bound to be judged harshly ten years after publication. “What I’ve said that turned out to be right will be considered obvious, and what was wrong will be humorous.”

Cass Sunstein
Cass Sunstein; drawing by David Levine

People in the computer industry have already criticized Gates for one such “humorous” error. In the early 1980s, Microsoft made its historic deal to be the sole supplier of operating-system software for the first IBM Personal Computer. This was like being the sole engine supplier for the first mass-produced automobiles, and it was the foundation of Microsoft’s later dominance. But the software that Microsoft provided, known as MS-DOS, had various limitations that frustrated its early users. One quote from Gates became infamous as a symbol of the company’s arrogant attitude about such limits. It concerned how much memory, measured in kilobytes or “K,” should be built into a personal computer. Gates is supposed to have said, “640K should be enough for anyone.” The remark became the industry’s equivalent of “Let them eat cake” because it seemed to combine lordly condescension with a lack of interest in operational details. After all, today’s ordinary home computers have one hundred times as much memory as the industry’s leader was calling “enough.”

It appears that it was Marie Thérèse, not Marie Antoinette, who greeted news that the people lacked bread with qu’ils mangent de la brioche. (The phrase was cited in Rousseau’s Confessions, published when Marie Antoinette was thirteen years old and still living in Austria.) And it now appears that Bill Gates never said anything about getting along with 640K. One Sunday afternoon I asked a friend in Seattle who knows Gates whether the quote was accurate or apocryphal. Late that night, to my amazement, I found a long e-mail from Gates in my inbox, laying out painstakingly the reasons why he had always believed the opposite of what the notorious quote implied. His main point was that the 640K limit in early PCs was imposed by the design of processing chips, not Gates’s software, and he’d been pushing to raise the limit as hard and as often as he could. Yet despite Gates’s convincing denial, the quote is unlikely to die. It’s too convenient an expression of the computer industry’s sense that no one can be sure what will happen next.

There are many examples showing how hard it is to predict the speed of technological advance or its effect on social or commercial life. Space travel. Cloning. Cures for cancer. The search for clean or renewable energy sources. The sense of apprehension in the last days of 1999 arose from the fact that while most experts believed the “Y2K bug” would not shut down computer systems, no one really could be sure.

The most dramatic recent demonstration of this problem involves the Internet. As a matter of pure technology, the Internet has worked far better than almost anyone dared hope. A respected engineer and entrepreneur, Robert Metcalfe, who invented the networking standard called Ethernet, predicted in 1995 that as more and more people tried to connect, the Internet would “go spectacularly supernova and in 1996 catastrophically collapse.” (Metcalfe later had the grace to literally eat his words, pureeing a copy of the column containing his prediction and choking it down before an amused audience.) In fact, the Internet has become both faster and less crash-prone the larger it has grown. This is partly because of improved hardware but mainly because of the brilliance of the “distributed processing” model by which it operates, which automatically steers traffic away from any broken or congested part of the network.

The financial assumptions surrounding the Internet have of course changed radically in a very short time. It was only three years ago that Lawrence Summers, then secretary of the treasury, joked that Brazil could solve its debt problems by changing its name to Brazil.com, since venture capital would then be sure to flow in. Until the summer of that year, any dot-com was assumed (by financiers) to be a winner, whether or not it had a plan for making profits. Since the spring of last year, any dot-com is assumed to be a loser, even if it in fact is becoming profitable.


On-line sales continue to grow, despite the general recession and the depression among dot-com companies. In 2001, sales at “normal” retail stores were flat, but sales by on-line merchants rose by 20 percent. The two most celebrated dot-com bankruptcies in 2001 were DrKoop.com, a medical advice site and on-line drugstore, and WebVan, a grocery-delivery service that in 2000 had a market value of $1.2 billion. There was nothing preposterous about either of their business concepts, only about the lavishness with which they were carried out. Sooner or later some company will make money letting customers fill prescriptions or order staple groceries on line. The successful companies will probably be branches of established drugstore or grocery chains. The hit film Startup.com, released a year ago, told the story of the rise and embarrassing fall of GovWorks.com, which was intended to let state and local governments do part of their business on line. This, too, is a sound concept, one that seems destined to spread. Few people who have the choice to register a car on line will want to trudge downtown to do it the normal way. But for now, the overcorrection and rush from dot-com investments leaves good ideas as underfunded as bad ideas were overfunded before.

While assessments of the Internet’s economic prospects have gone through manic swings, interpretations of its political and social effects have displayed a kind of stable schizophrenia. That is, through the last half-dozen years of the Internet’s explosive rise, observers have agreed that it would do something significant to society, but have disagreed about whether the effect would be good or bad.

The utopian view, strongest in the Internet’s early days but still heard now, boils down to the idea that the truth will make people free. Elections will become more about “issues,” as voters can more easily investigate each candidate’s position. Government will become more honest, as the role of money is exposed. People in different cultures will become more tolerant, as they build electronic contacts across traditional borders. Tyrants will lose their grip, as the people they oppress gain allies in the outside world and use the Internet to circumvent censorship. Liberal democracies will govern with a lighter hand, as information technology makes them more efficient and responsive. The recent struggles of dot-com companies have dampened the enthusiasm of some of the strongest exponents of these views, and have postponed estimates of when the desired effects might occur. But the concepts have not gone away.

The opposing, dystopian view shares the assumption that the Internet will weaken traditional power structures, but it emphasizes all the ways in which that will be bad. Families and communities will be weakened, as each member spends more time with “virtual” friends and associates. Childhood may be destroyed, because of the lure of pornography—of which a huge supply is available on the Internet—and of addictive on-line games and the threat of sexual predators. If the Internet ultimately erodes the barriers among people and parts of the world, then any culture, community, or institution that requires a sense of separate identity to survive is threatened. A variant of this concern has been the strongest international complaint about the Internet: that it would be another means of promoting American values and the English language.

The Internet’s effect on language already seems to be evolving in an unexpected way. Initially nearly all the Internet’s users were native English speakers, and nearly all Web pages were in English. Now well under half of all users are native English speakers, and the proportion can only fall. But the proportion of English-language pages is falling more slowly. It was 85 percent five years ago and about 60 percent now, as pages in Japanese, German, Chinese, Spanish, French, and other languages have been added rapidly. But in many parts of the non-Anglophone world, users are posting pages in English along with their national language, or instead of their own language, to make the sites comprehensible to the broadest-possible group of readers. One recent academic study of this trend, called Internet—Flagship of Global English?, concludes that the Internet will cement the role of English as universal lingua franca. The study was carried out at the University of Lecce, in Italy, and the results were posted in English.2

Previous big, modern innovations have made a significant difference in family, community, and national life. Antibiotics and immunization dramatically cut childhood mortality in the developed countries, which in turn altered family patterns and the place of women. So too for electricity, the telephone, automobiles, air travel, radio and television, and modern techniques of farming. The Internet could have effects as profound as any of these—but they won’t be clear all at once.



The discussion that has surrounded Cass Sunstein’s republic.com is a useful way to begin considering these long-term effects. I emphasize the discussion as much as the book, for it highlights some of the issues in the book in an unusual way. republic.com was published last spring. In it Sunstein, a highly regarded law professor and First Amendment specialist at the University of Chicago, addressed two different subjects with different degrees of authority and success.

One of his subjects was the connection between information flow and democratic government. His argument was that in a democracy “free expression” must mean something more than mere absence of censorship. Instead, a “well-functioning system of free expression,” one adequate to equip citizens for self-government, had additional tests to meet:

First, people should be exposed to materials that they would not have chosen in advance. Unplanned, unanticipated encounters are central to democracy itself…. I do not suggest that government should force people to see things that they wish to avoid. But I do contend that in a democracy deserving the name, people often come across views and topics that they have not specifically selected.

Second, many or most citizens should have a range of common experiences. Without shared experiences, a heterogeneous society will have a much more difficult time in addressing social problems…. Common experiences, emphatically including the common experiences made possible by the media, provide a form of social glue.

Roughly half the book is a constitutional and political analysis supporting this view of free expression, and it is convincing and clear. The first kind of exposure, to information one has not chosen in advance, is important because otherwise one’s views would never change. The second kind of exposure, to common experiences, is especially important in a big, racially and economically diverse nation in which citizens may have very few assumptions in common. America’s “shared experiences” of the last generation have largely been spectacles: entertainment (the Oscars, Survivor); sports (the Super Bowl); “public” events that attracted more attention for their melodramatic rather than their political meaning: the O.J. Simpson trial, the death of Princess Diana, the Gary Condit affair, even the sex drama that became the occasion for a presidential impeachment. The reaction to the September 11 attacks was the most truly consequential shared experience in at least a generation.

There was another half to republic .com, and a less convincing one. Having defined the kind of free expression that was necessary for democracy, Sunstein went on to identify a major threat to it: namely, the Internet. In particular, he was concerned about the “filtering” and the personalizing technology of the Internet, which would in principle allow people to define in advance exactly the information they did—and did not—want to see. The more efficient the filter, the less chance that a citizen would be exposed to healthy surprise—or share experiences with the rest of society. As technology evolved, democracy would deteriorate. The “Daily Me” was Sunstein’s name for the news publication of the future. It would destroy the underpinnings of the “us” that is democracy.

His book began with a “thought experiment” about the nature of this new world:

It is some time in the future. Technology has greatly increased people’s ability to “filter” what they want to read, see, and hear…. With the aid of a television or computer screen, and the Internet, you are able to design your own newspapers and magazines….

You need not come across topics and views that you have not sought out. Without any difficulty, you are able to see exactly what you want to see, no more and no less.

“In reality, we are not so very far from complete personalization of the system of communications,” he concludes. “The changes now being produced by new communications technologies are understated, not overstated, by the thought experiment with which I began.”

This line of reasoning, which warns against the Internet as an impediment to democracy, has two problems: one involves the Internet, and the other involves the nonelectronic ways in which citizens interact.

Sunstein’s warnings last spring about the ominously perfect info-filtering technology did not, to put it mildly, have the easy authority shown in his discussions of the First Amendment. They were more like suburban fretting about the bad things that must be happening on the other side of town. After his opening “thought experiment,” Sunstein proceeded with a list of Internet companies whose advanced filtering technologies were leading to the “Daily Me.” Several of these companies had gone out of business by the time Sunstein’s book appeared, and several more have since. Sunstein can’t have spent much time using any of these sites if he thinks their filtering is effective enough to pose a threat. To see for yourself, go to one of the main news sites that offers a personalized compendium of information, such as CNN.com, MSNBC.com, or Go.com, and see how “me”-like you can make it. You can set it to display your city’s weather, and the stock quotes you care about, and the movie listings in your neighborhood, and the scores for the local teams. But the rest of the information you see has a high chance of being “unexpected.”

The filtering available on Internet sites is primitive compared to the filters, cushions, and blinders that surround us the rest of the time. The patterns Sunstein warns about—a lack of shared experience and the balkanization of Americans according to class, region, religion, and ethnicity—are real and worrisome enough. But the Internet is a trivial source of the problem—let’s say one thousandth as important as the educational system, from school districts with their unequal funding to the faulty system of college admissions. Or residential patterns. Or who marries whom. Or tax policy. Or the existing broadcast media, which let you drive coast to coast listening to nothing but right-wing talk radio or NPR. Or cable TV, with one channel showing only bass fishermen and another showing only success-motivation seminars. Or patterns of commuting, which have evolved from buses to cars, and remove people from accidental contact with others. You could un-invent the Internet and still have every problem Sunstein fears.

The discussion of Sunstein’s book since it was published has itself been telling. In its summer issue last year, the Boston Review printed comments on republic.com from seven scholars and writers. All sympathized with Sunstein’s concern about ensuring healthy, democratic discussion. Most were skeptical about the Internet as a source of the problem.3

Michael Schudson, of the University of California–San Diego, said, “The Internet may very well reduce our common media fare, as Sunstein fears, but even in our mass-mediated era we do not live very much of our lives through the media.” He mentioned the evidence that the Internet may actually encourage more civic engagement—through means as simple as e-mailed community newsletters—rather than less.

Ronald Jacobs, of SUNY–Albany, argued persuasively that major portals like yahoo.com and aol.com, with their search engines, links to news stories, chat and message board services, and advertising, “function precisely like the general interest intermediaries that Sunstein thinks are so important. That is, they provide unanticipated encounters as well as common experiences.” Shanto Iyengar, of Stanford, said there was no “serious ground for concern that online sources will only attract users who already share their points of view. The available evidence suggests the contrary.” Henry Jenkins, of MIT, offered a sarcastic “thought experiment” of his own:

Some years ago, a local bank announced plans to discontinue its “Time and Temperature” service, prompting me to whimsical speculation about how this decision could lead to total anarchy. Without a means of synchronization, our clocks would gain or lose time until we drifted out of sync with each other. Workers would arrive late or leave early; teachers wouldn’t know when to end classes; participants in social and professional gatherings would stomp off impatiently when the expected party failed to arrive. Some groups of friends might create their own time zones and ignore everyone else.

Running through all these comments is an appreciation of something that Sunstein may not have wasted enough time in front of a computer to share. Compared with most other indoor activities, time with the Internet is less filtered, more open-ended, more likely to lead to surprises. If you read a book or magazine, you usually keep reading. If you watch a video, you watch. But if you start looking up information on Web sites, you almost never end up where you expected. There’s a link to something you’d never heard of before, some news you hadn’t known was interesting. It’s not the same as walking to a new part of town, but it’s a lot more surprising than listening to the radio. The feeling is similar to that of going through library stacks—if there were no dust and you could instantly zoom from floor to floor.

In a forthcoming book, Small Pieces, Loosely Joined,4 David Weinberger, who runs the Web site “Journal of the Hyperelinked Organization” at www .hyperorg.com, elaborates this theme of the chaotic, always surprising nature of the Internet. Without mentioning Sunstein’s book, he refutes its central claim that the Internet has a narrowing effect on people’s minds. With the World Wide Web’s ceaseless growth, he says,

there is more and more to distract us—more sites to visit, more arguments to jump into, more dirty pictures to download, more pure wastes of time. The fact that the Web is distracting is not an accident. It is the Web’s hyperlinked nature to pull our attention here and there. But it is not clear that this represents a weakening of our culture’s intellectual powers, a lack of focus…. Maybe set free in a field of abundance, our hunger moves us from three meals a day to day-long grazing…. Perhaps the Web isn’t shortening our attention span. Perhaps the world is just getting more interesting.

I suppose I shouldn’t say that the Boston Review exchange illustrates the give-and-take of the Internet, since it originally appeared in the printed magazine. But the seven responses, and Sunstein’s reply, are now available on line along with two dozen other Boston Review exchanges.5 The very ease of reading through them that way, and comparing them with comments in other exchanges on other topics posted months or years earlier, creates a kind of discovery and linkage that would be harder to equal in any other way. What it could lead to politically is as unpredictable as ever.

Sunstein responded to the Boston Review replies with an intriguing shift of position. He began his response, which also serves as an afterword to the forthcoming paperback edition of his book, with this restatement of his two main contentions:

  1. A democracy requires both a range of common experiences and unanticipated, unchosen exposures to diverse topics and ideas. For those who accept this claim, democracy might well be jeopardized by a system in which each person decides, in advance, what to see and what not to see….
  2. The Internet is bad for democracy, because it is reducing common experiences and producing a situation in which people live in echo chambers of their own design. For those who accept this second claim, the current communications system is inferior to one in which general interest intermediaries dominated the scene.

I endorse the first claim…. But I do not endorse the second claim. I believe that the second claim is basically wrong, because the Internet is allowing millions of people to expand their horizons and to encounter new worlds of topics and ideas.

You have to admire the panache of this statement. Apparently the Boston Review panel was lulled, as I was, into misunderstanding Sunstein’s true intent, by sentences in the book such as “For countless people, the Internet is producing a substantial decrease in unanticipated, unchosen encounters,” or “There can be no assurance of freedom in a system committed to the ‘Daily Me.'” But I also take his revised view as a sign of the collective effort underway to improve and revise our understanding of this new technology—and the likelihood that whatever we think now may soon prove wrong.

An E-mail from Bill Gates

Denying he ever said “640K should be enough for anyone,” Bill Gates wrote me recently as follows. Some technical expressions are explained in brackets:

This is one of those “quotes” that won’t seem to go away.

I’ve explained that it’s wrong when it’s come up every few years, including in a newspaper column and in interviews.

There is a lot of irony to this one. Lou Eggebrecht (who really designed the IBM PC original hardware) and I wanted to convince IBM to have a 32-bit address space, but the 68000 [a Motorola-designed processing chip, eventually used in the Apple Macintosh] just wasn’t ready. Lew had an early prototype but it would have delayed things at least a year.

The 8086/8088 [the Intel-designed chip used in early personal computers] architecture has a 20-bit address bus [the mechanism used by the microprocessor to access memory; each additional “bit” in the address bus doubles the amount of memory that can be used], and the instruction set [the basic set of commands that the microprocessor understands] only generates 20-bit addresses.

I and many others have said the industry “uses” an extra address bit every two years, as hardware and software become more powerful, so going from 16-bit to 20-bit was clearly not going to last us very long. The extra silicon to do 32-bit addressing is trivial, but it wasn’t there. The VAX was around and all the 68000 people did was look at the VAX! 2 to the 20th is 1 megabyte (1024K), so you might ask why the difference between 640K and 1024K—where did the last 384K go?

The answer is that in that 1M of address space we had to accommodate RAM [random access memory], ROM [read-only memory], and I/O addresses [Input/Output addresses used for “peripherals” like keyboards, disk drives, and hard drives], and IBM laid it out so those other things started at 640K and used all the memory space up to 1M. If they had been a bit more careful we could have had 800K instead of 640K available.

In fact, we had 800K on the Sirius machine, which I got to have a lot of input on (designed by Chuck Peddle, who did the Commodore Pet and the 6502, too). The key problem though is not getting to use only 640K of the 1M of address space that was available. It’s the 1M limit, which comes from having only 20 bits of address space, which is all that chip can handle!

So, this limit has nothing to do with any Microsoft software.

Although people talk about previous computing as 8-bit, it was 16-bit addressing in the 8080/Z80/6800/6502 [all early processing chips]. So we had only 64K of addressability.

Amazingly people like Bob Harp (Vector Graphics—remember them?) went around the industry saying we should stick with that and just use bank switching techniques. Bank switching comes up whenever an address space is at the end of its life. It’s a hack where you have more physical memory than logical memory. Fortunately we got enough applications moved to the 8086/8 machines to get the industry off of 16-bit addressing, but it was clear from the start the extra 4 bits wouldn’t be sufficient for long.

Now you MIGHT think that the next time around the chip guys would get it right.

But NO, instead of going from 20 bits to 32 bits, we got the 286 chip next. Intel had its A team working on the 432 (remember that? Fortune had a silly article about how it was so far ahead of everyone, but it was a dead end even though its address space was fine). The 286’s address space wasn’t fine. It only had 24 bits. It used segments instead of pages and the segments were limited to 24 bits.

When Intel produced the 32-bit 386 chip, IBM delayed doing a 386 machine because they had a special version of the 286 that only they could get, and they ordered way too many of them.

It’s hard to remember, but companies were chicken to do a 386 machine before IBM. I went down to Compaq five times and they decided to be brave and do it. They came out with a 386 machine! So finally the PC industry had a 32-bit address space.

We have just recently passed through the 32-bit limit and are going to 64-bit. This is another complex story. Itanium is 64-bit. Meanwhile, AMD on its own has extended the x86 to 64-bit.

Even 64-bit architecture won’t last forever, but it will last for quite a while since only servers and scientific stuff have run out of 32-bit space right now. In three or four years the industry will have moved over to 64-bit architecture, and it looks like it will suffice for more than a decade.

Apollo actually did 128-bit architecture really early, as did some IBM architectures. But there are tradeoffs that made those not ever become mainstream.

A long answer to just say “no.” I don’t want anyone thinking that the address limits of the PC had something to do with software or me or a lack of understanding of the history of address spaces.

My first address space was the PDP-8. That was a 12-bit address space!

Even the 8008, at 14 bits, was a step up from that.

This Issue

March 14, 2002