Notes on network computers, blind copies, democratic culture, monopolies, and the idea of an Internet establishment... As a periodic reminder, a Web archive of nearly all the RRE messages ever sent can be found through http://communication.ucsd.edu/pagre/rre.html It turns out that the article on cryptography by Bruce Schneier that I forwarded from the Risks Digest the other day was actually an unfinished draft that Peter Neumann sent out by mistake. Bruce asks that everyone refrain from propagating that version around the net. Steve Lohr's article on the "network computer" in Monday's New York Times includes the following priceless quotes: The line between these network computers being used as really effective corporate tools and being used as mind-control tiger cages is a fine one. There will be a real temptation for corporate managers to go too far in the direction of control. -- Paul Saffo The paradise of shared knowledge and a more egalitarian working environment just isn't happening. Knowledge isn't really shared because management doesn't want to lose authority. -- Shoshana Zuboff What's a floppy disk? It's a way to steal company secrets. -- Scott McNealy We think we can go after the dumb-terminal market too. -- Bill Gates I wish we could rewind this whole computer revolution thing and start over. After my last couple of "notes", some RRE readers wrote to complain about my long paragraphs. I'm afraid I think in long paragraphs, but in the future I'll try to chop them up. Non-Americans keep remarking on the predominance of American items on RRE. That's simply a function of the items that RRE subscribers send me. If you come across interesting items from other countries, I hope you'll consider sending them along. You've heard me complain about Eudora's unfortunate "redirect" feature, and the unfortunate custom of using it to forward messages with misleading headers. I've recently put my finger on another unfortunate e-mail custom -- abuse of the Bcc: or "blind copy" header field. Recently, for example, I have received several messages that were addressed To: some mailing list and Bcc: some other individuals and/or mailing lists, so that many people receiving the message thought that they had been added to the mailing list in the To: field without their permission. Asked what motivates this kind of behavior, the senders usually say that they want to spare their readers from seeing all of the addressees. The truth, though, is that you are not doing anybody a favor by creating a misleading header. The Bcc: field has few legitimate uses. One is the custom of creating a one-shot mailing list by sending a message To: yourself and Bcc: many other people, then explaining at the top of the message what you have done. Let's not make the Internet any more confusing than it has to be. A while back, I came across an op-ed column addressed to teen-agers which purported to explain certain points about life. The first of these points was, and I quote, Life's not fair. Get used to it. All of the points were like this: each of them presupposed that their reader held a putatively naive or self-serving opinion about life, and they proposed to set the reader straight in a remarkably nasty and disrespectful way. I view this article as part of a larger and very depressing trend: the return of authoritarian culture. The purpose of authoritarian culture is to instil a mindless obedience to authority. It employs two basic methods. The first of these methods is stereotype: one's normal human tendencies to think critically and oppression resist are caricatured and ridiculed; endless stories are adduced to portray people who employ these innate faculties in a bad light; and labels such as "whining" and "complaining" and "victim" are liberally applied. The second method of authoritarian culture is the attempt to naturalize authority by hiding it behind large abstractions. In this case, the abstraction in question is "life". Having established that "life" is unfair, it becomes possible to label any protest against unfairness as a demand that the whole world conform to one's own immature whims. These teenagers are counseled to "get used to it" and to reconcile themselves to a life of being treated unfairly. No liberal nostrums about self-esteem here: this columnist's message was that nobody deserves to be treated with respect, and that it's arrogant to think otherwise. Another example of authoritarian culture is the contemporary American use of the word "accountability". Accountability, we are told, means "accepting the consequences of your actions". Everyone is supposed to "be accountable", and to embrace this condition as a concomitant of responsible adulthood. Authority here is hidden through a grammatical device. In normal usage, the word "accountable" takes a complement, as in "accountable to ...". When the complement is omitted, the human authority is displaced into the woodwork, and obedience to that authority is conflated with a variety of quite different conditions: responsibility, honesty, etc. No reasonable person has a problem with the idea of being responsible for one's actions. But nobody who believes in a democratic society can accept the idea that any person has absolute authority to judge any other person's actions, and hand down "consequences", without likewise being constrained by norms of responsibility, which in the old days were called "justice" and -- yes, that's right -- "fairness". It is not surprising when people in authority employ such language. But authoritarian culture requires more: it requires that people internalize this language, applying it to themselves and dissociating all desire for justice and fair treatment. We all have our failings, but we we cannot have a decent society unless everyone is treated with respect and judged with a reasonable regard for proportionality and due process. Healthy people don't just "get used to" injustice. Quite the contrary, genuine maturity begins with the skill and discipline of helping people organize to identify and overcome injustice. Shame and ridicule are the least violent of the tools that have historically been employed to condemn people to passivity. But they are also the most basic and, in the end, the most destructive. The basic method for promoting an extreme position is to harp on the evils of the opposite extreme. Authoritarian culture thus lives in a symbiotic union with its evil twin, libertarian culture, whose sole value is freedom from constraint. The symbiosis between authoritarian and libertarian culture has many facets: * Authoritarian culture holds that people are essentially bad and that nothing can be done about this; libertarian culture holds that people are essentially good and that nothing needs to be done to encourage this. * Authoritarian culture imposes constraint without respect for individual dignity; libertarian culture holds that individual dignity consists in the absence of constraint. * Authoritarian culture holds that people are innately irresponsible; libertarian culture denounces responsibility as an authoritarian myth. * Authoritarian culture and libertarian culture both conflate feelings with action, authoritarian culture to repress them both and libertarian culture to license them both. * Authoritarian culture crushes the spirit and eventually gives rise to an immature impulse toward libertarian culture; libertarian culture stands indifferent as great industries arise to support an epidemic of addiction, which then gives rise to a fearful impulse toward authoritarian culture. What authoritarian and libertarian have in common is their claim to follow a simple, objective rule that lies beyond human interpretation: the rule of order or the rule of freedom. The terra incognita that lies beyond the dysfunctionality of both authoritarian and libertarian culture is democratic culture: the form of culture within which everyone takes responsibility for living together constructively. Democratic culture is not just a matter of voting. It is a set of values, and it is a set of skills. Some of these skills are organizational: you can't have democratic culture unless people know, deep down in their bones, how to hold a productive consensus-based meeting. Other skills are emotional: you can't have democratic culture unless people can tell the difference between resisting oppression and acting out resentment, between organizing and polarizing, between freedom and irresponsibility, between pleasure and addiction, between discipline and shame, between personal boundaries and passive aggression. The sixties have left an ambiguous legacy because they blurred together two very distinct impulses: countercultural libertarianism, for which I don't have an awful lot of respect, and democratic experimentation, for which I have a great deal of respect. Recreational drug use, for example, is stupid and boring. But the democratic organizing traditions that arose and flourished in the sixties were an important cultural contribution, and it's sad to see them forgotten. The rise of authoritarian culture depends on this forgetting, and on crushing at an early age the hopes for human dignity with which all of us are born. Why isn't this obvious? One reason is that the American conservative movement originated as a marriage of convenience between authoritarians and libertarians, both of whom portrayed themselves as opponents of something called "government". But opposition to government tout court is opposition to democracy. We have been inundated in recent years by rhetoric that seeks to make democracy literally unthinkable by conflating all types of government, whether democratic or totalitarian, into a single stereotype of oppression. This stereotype requires its proponents to construct themselves as powerless victims, and it licenses all sorts of whining and complaint by the very people who make a big point of censuring whining and complaint by others. By treating the institutions of a democratic society as inherently beyond control, it also licenses an abdication of personal responsibility -- the responsibility to learn, practice, and teach the values and skills of a democratic society. Back in the Later Middle Ages, around 1994, I often found myself lectured by experts who asserted that great bureaucratic institutions, government and corporate alike, would necessarily and inevitably disintegrate in the world of the Internet. It's almost 1997 now, and as the months go by I find it ever harder to remember why this great disintegration was supposed to take place. If evidence counts for anything, we are actually living in an unprecedented era of concentration and centralization. ABC has now become the Disney Infomercial Channel, and British Telecom is buying MCI. And the more I learn about Internet economics, the more the Internet seems like a veritable engine of monopoly-creation. The reasons are numerous. Most of them are explained in the recommendations and bibliographies that I have provided in past issues of The Network Observer, and any one of them would suffice: * The Internet backbone business, like any other utility, involves high fixed costs and low marginal costs of serving additional customers, so that bigness is highly rewarded. And every provider has an incentive to give priority to packets that remain within its network, thus giving a quality-of-service advantage in the market to whichever provider first establishes a dominant position in a given segment of the market for network services. * The pattern of high fixed costs and low marginal costs is even stronger for software: if two software products compete then the price of each is determined by its development costs divided by the number of customers. As a result, other things being equal, the company with the larger market share completely destroys its competition. If other things aren't equal, a company with an existing flow of monopoly rents can simply give away its products until its competitors' capital is depleted. * And then there are de facto standards: the dynamics of standards are fantastically complicated, but the Internet still needs several additional service layers, and even a single proprietary standard could give one firm immense power to extract rents from Internet users and dictate future directions for Internet architecture. Since most of the forthcoming standards will be implemented in software, the monopoly tendencies of the software market will promote the rise of proprietary networking standards. Past experience with the Internet is misleading: small players and ARPA philosopher-kings can establish a new standard in the world if the big players are asleep, but the big players are awake now. In the future it will take capital and speed to impose a standard, and winner takes all. * The economics of information -- "content" -- are no more reassuring. Since information can be replicated cheaply, an overwhelming, life-or- death incentive exists to leverage creative effort across as many channels as possible. Of course, every industry has an incentive to increase its revenue. But as the marginal cost of selling additional units approaches zero, prices are directly determined by market share, thus giving rise to a positive feedback loop and eventual monopoly. This suggests that, to the extent that the Internet is a vehicle for the delivery of information commodities, it will necessarily be absorbed into the institutions of the media system. Nobody will be able to afford to produce a news service that only operates on the Internet, for example, given that existing news services can adapt their work to the Internet for much less than it costs to produce a competing service from scratch. And content providers who already operate in other media will have a powerful incentive to adapt their services to the Internet, provided only that additional money can be made by providing those services in the Internet medium. It is commonly argued that the Internet, by reducing transaction costs, will cause organizations to break apart and industries to reorganize on a more entrepreneurial basis. I have argued this myself. But the argument doesn't work. The Internet (and not only the Internet, but a world of other technologies) reduces transaction costs and coordination costs alike, making it less efficient to operate large organizations that cross many disparate functions but more efficient to operate large organizations that do one standardized thing. Thus we have waves of both outsourcing (due to decreased transaction costs) and concentration (due to decreased coordination costs). And high-tech industries increasingly favor those firms that can see standards battles coming and prepare to fight them in a coordinated way. Other arguments against the monopoly scenarios have more force. Internet service provision is not yet a monopoly, and it is not obviously headed in that direction right now. Nobody understands the ISP business very well, and it would appear that the smaller entrepreneurial firms have an advantage in adaptability in an environment of rapid technological and market change. AT&T, moreover, would seem to be falling apart at the seams in very much the fashion that Schumpeter, that patron saint of entrepreneurs, would have predicted. Even MCI has been hitting itself on the head pretty hard during the last few years. So my point is not that the Internet is definitely going to be locked up by cigar-smoking corporate titans. My point is simply this. It's almost 1997, and yet the airwaves and magazine pages are still full of ideologues, still reciting old-wave technological determinist mantras about how the Internet will inevitably bring us a decentralized world of freedom. Technological determinism, however, is a kind of cargo cult: it asserts that we can obtain a happy society if we simply work hard enough at idolizing the technology. The world doesn't work like that. Significant forces are operating in several directions, some of those directions are more pleasant to contemplate than others, and the outcome is not preordained. It seems to me that the proponents of a totally unregulated economy should learn a little more economics. It also seems to me that it's time for a revival of the democratic values that make it possible to imagine a conscious choice about our technological future. When I discussed the October issue of Wired the other day, I should have mentioned that their "net surf" department included a nice advertisement for RRE. Everyone assumes that this resulted in thousands of new RRE subscribers, but long experience shows that it doesn't work like that. Everyone gets too much e-mail, and so it takes an awful lot to get them to subscribe to a new mailing list, especially one so untargeted as RRE. I would say that the post-Wired surge in subscriptions was 300 at most. The forwardable advertisement that I subsequently sent to the list myself produced at least as many. I think it's true what the advertisers used to say: people need to hear a message 17 times before they buy. The list currently has about 4300 direct subscribers, creeping upward by slow increments. After my comment about the Journal of Internet Banking and Commerce the other day, I got an irate message from the Journal's editor. Among his less extravagant suggestions was that I am a member of the Internet establishment who has the power to destroy people with a word. Gosh. My first response was to wonder if Hallmark has a card for this occasion: To a childhood friend, after all these years I can still remember how, it seems You never picked me for your teams I have no idea now where you've went But I'm a member of the Internet establishment So you better watch out what you say Or with my keyboard I'll blow you away Then I thought about it more seriously. I'm not here to take shots at struggling Internet publications, unless you count Slate. My only goal had been to plug the Scout Report, and the truth is that I was too lazy to dig up an issue of TSR that I was 100% comfortable with. I won't go into the reasons for the opinion I formed of JIBC back in the spring, since you can easily judge for yourselves. Of course I should take ordinary care not to slam people at random, and by trying to run this list in odd moments of the day I'm all too likely to get careless. The harder question is, is this guy right? Never mind that he had a grossly exaggerated estimate of RRE's readership -- the idea that I am a member of any kind of establishment is completely foreign to me. Of course, this could simply be a self-serving delusion. After all, as a white guy who was born in the United States, I enjoy numerous social privileges that I have never earned. And as a graduate student at MIT, my cohort and I were located so close to the center of the post-WWII military-academic complex that we had the resources to pretend that we were antiestablishment nonconformists. Still, if the Internet establishment exists and holds councils, I have never been invited to them. I have met some of the publicly prominent Internet people once or twice and have found them to be remarkably decent, but I hope you will agree that my message on RRE is not particularly congruent with theirs. Yes, I organized a CPSR conference a few years back, but look what's happening to CPSR. So, at the end of the day, am I basically just another guy with a mailing list? When I started RRE, I assumed that surely by now the Internet would be crawling with filter lists such as my own. But even though some other excellent filter lists do exist, including some (David Farber's, for example) with many more subscribers than my own, my assumption hasn't come close to being true. Why is this? I can think of a few reasons: * Limited access to the tools. The hardware keeps getting faster, but the software for maintaining mailing lists isn't much better than it was in 1993. It's still impractical for 95%+ of Internet users to maintain large mailing lists. This is a scandal. * Flexibility of the tools. The assumption that lots of people would start filter lists depended on another, implicit assumption: that the Internet only supports a limited number of interesting mechanisms. Lots of people are doing great things on the Internet, and they have shaped tools to fit their particular visions, which simply differ from mine. Even the other filter lists vary widely in their traffic, contents, respect for copyrights, and amplitude of editorial voice. * Network externalities. As I mentioned the other day, it's possible that once a list like RRE establishes a large subscriber base, other lists will have a hard time getting enough traction to compete. This is because the success of such lists depends in part on the number of subscribers they have, and I got the subscribers first. On the other hand, 4300 is a small fraction of the total world of Internet users. * Critical mass. RRE is mostly about the social and political aspects of networking and computing. This is a fashionable topic, and it is a topic likely to appeal to a large proportion of Internet users. Someone who wanted to run a filter list devoted to medical issues, for example, would be limited by the number of Internet users in the medical field. On the other hand, one major early source of RRE messages was Gleason Sackman's high-volume filter list on education. I wish I could say that the Internet affords freedom of the press to those who don't own one. But it's not true, not yet. So maybe, on some very tiny scale, I'm an Internet analog of Rush Limbaugh or Scott Adams. Those guys could hardly be more different on one level, but they also reflect an important convergence. Rush didn't invent talk radio with its structured listener involvement; the innovation of his show was that it's basically about his opinions. Dilbert didn't invent the comic strip; his innovation was to use the Internet to involve his readers in thinking up the ideas. In each case, the result is a voice that seems on the surface like a synthesis of the star's voice with that of his audience, and the fear is that this result is a sham, giving just enough of an appearance of audience involvement to authenticate the star as the Voice of the People. Likewise, by forever soliciting subscribers' submissions while exercising absolute editorial control, I can imagine that this list conveys a sense of omniscience that has little basis in reality. As a recovering know-it- all, I can testify how addictive this position can be. I can imagine how this list can be perceived as more authoritative than it really is. And since perceptions of authority are largely self-fulfilling, I can appreciate how an RRE message can do more damage than it has any right to. I don't really know for sure, nor is it mine to judge. But I suppose I should err on the safe side.