Questions

Agre:A couple of weeks ago, I was in Brazil at the University of Sao Paolo, and I spoke to a physicist who runs the server on which Linux and a wide variety of other free software is provided to people in Brazil. They have been experiencing exponential growth over the last several months, and so they have set up a whole series of institutions to support free software in Brazil. Brazil is a country that cannot afford to be buying an awful lot of expensive software so the attraction of free software is natural. He said that the main obstacle that free software faced in Brazil was that the kind of mutual assistance that it presupposes and requires. That was not congruent with Brazilian culture. It was unfamiliar to the culture. This is in the context of a society that was ruled by military dictatorship for some time. You don't start being ruled by a military dictatorship unless something is wrong with the mechanisms of civil society in the first place. So I was talking to him and to reporters in the b usiness press there about the institutions that they were setting up around the software both within and beyond Brazil. There is not a lot of English spoken, so in Portuguese principally, there were largely domestic institutions that didn't reach over borders. We talked about civil society, and we talked specifically about global civil society. This is an example of the infrastructure having a reciprocal relation to institutions of civil society. It seemed a very hopeful thing, and I tried to sell reporters on the idea of the larger significance of this being a very important story to write about, and I couldn't. I utterly failed to communicate that this was a significant story.

Now there is a major problem with the economics of software that we are up against which is that software exhibits vast economies of scale. It is very expensive to write software. But once you produce the first copy, it is very cheap to produce all subsequent copies. This rewards whichever organization has the best distribution system because if it can distribute twice as many copies of its software than its competition then it can afford to charge half as much. And unless the competitor is twice as good, it will lose out completely and a monopoly will arise. This is an acute, serious problem in all mature software markets. And so we really face this fork in the road: one direction leads to monopoly as the basic model of the workings of information in society, from straightforward economic arguments that explain the fine detail of a large number of cases; and in the other direction lies mutual assistance -- the rise of a vigorous civil society as the nature and the basis of an information society, if such a thing can be said to exist. I think the reality will be some of one and some of the other in the interaction between the two models. We are seeing some of the nature that this interaction might have. But I think we are very far from having the answer.

Code determines how the Internet runs and how much of modern life is constrained. When people talk about an alternative to open source, that means closed and proprietary source. What they are saying is that the part of the regulatory structure -- code that is just as powerful as law in the context of the internet -- will not be open to public scrutiny. Now think about that for a minute. Think about people arguing that you should do so much of your work on closed proprietary software that cannot be inspected by a populace, and then imagine the laws of the United States not being open to public inspection. I think that raising the issue of open source in the context of government is an absolutely crucial question for thinking about how our society is going to evolve.

Nick Sammond: Phil, I'd like you to address the conflation of democracy with equality, particularly around the notion of negotiation between supposedly equal partners. By way of an example, I live in San Francisco which is ground zero of the Internet Goldrush. In San Francisco, if you are a renter it's becoming almost impossible to live there because so much capital is flowing in the Bay area and so many people are going to work with that capital that they are driving the housing market crazy. The city government is responding by catering to them and so those of us who would like to stay in the city, who have been long-time residents of the city, now find ourselves unable to negotiate any kind of terms to remain there. So the Internet is having a tangential effect in which we have no place to negotiate, we have no access to democracy because the government is not interested in us as democratic players anymore.

With regard to the particular case: the level of rent in San Francisco is a measure of how much people in the industry are willing to talk to each other face to face. If anyone should use teleconferencing -- it is these people, they have the technical know-how, they would not be intimidated by it, they already have the money for it and they would use it enough to be willing to spend the money to install it. Well, but teleconferencing just isn't happening -- it is actually a mystery why. And the rents in San Francisco are a precise measure of how deeply teleconferencing is not happening. It is also a phenomenon of positive feedback. I think that something that has become clear in the economics of the "information society" is the large number of positive feedback mechanisms of which that is one. Silicon Valley has become Silicon Valley-er and Route 128 has become Route 128-ier at the other extreme. And there are quite a few of those mechanisms.

In Brazil we actually have long conversations about this. The network effects that promote English as a global language is another example. There are what are called aggregation economies in spatial economics whereby there is a lot of good stuff in one place and more good stuff goes to that place to be close to the other good stuff. And positive feedback loops coming from the increasing integration of the American economy. Silicon Valley is not as typical of the future wave of things as it pretends to be because the kind of software that they write there really is quite a distinctive commodity and calls for a distinctive kind of industrial organization. But the kind of hyper-integration whereby Silicon Valley itself is a unit and the social network of people there are the unit, rather than thinking of the corporation as a unit. That is certainly the way of things and that phenomenon exhibits positive feedback effects as well.

The simplistic ideas of the Internet as equalizer, or the kind of moronic argument like George Gilder's Life After Television -- the fact that people have lots of computer power on their desks means that political power will likewise be decentralized -- are just completely unrelated to real evidence or to serious theory of any sort, including theories produced by people who share George Gilder's politics, but have their brains turned on. So the effects cut multiple ways. I was in a good mood last night so I was emphasizing the happy effects, but you are right to open up some of the others.