Date: Sat, 24 Sep 1994 20:51:03 -0700 (PDT) From: Nick Szabo To: Extropians@extropy.org Subject: IDs and Ideals Here's a post I just made to the new Seattle Times 2020 column list (slightly edited): It's now technically possible to assign a single ID number to all citizen units: "the mark of the beast". It's just as technically possible to issue multiple IDs and pseudonyms for a person. More obscure, but just as feasible, is the work Chaum et. al. in cryptographic protocols: digital cash that works like real cash (eg it doesn't leave an identified transaction trail, like current credit/debit/ATM cards), and credentials that can be transferred between per-organization pseudonyms. So, for example, you could prove you have a valid driver's license, or are over age 18, or have a high school diploma, or prove any other valid credential you possess, without having to reveal your birth name if you don't want to. Another example of realizable "crypto magic" is a kind of virtual computer that can be set up on the net. These "multiparty secure computers" allow any number of parties to input and output programs and data to the virtual processor, without any party learning the input or output of any other party, except what they can infer from their own inputs and outputs. We will see developed a wide variety of these and other new forms of cryptographic "magic", unless the snoops they threaten succeed in stopping them. My use of the word "assign" already indicates that this is as much an economic and political issue, as a technological issue. The assumption that numbers must be assigned, rather than chosen, is a political bias with major consequences. Why can't I _choose_ ID numbers instead of having them assigned? With a reasonable variety of choices, and a political agreement on the desirability of choice, I can choose which of my purchases, medical activities and other activities I want to have recorded and tracked, and which I want to remain part of my private life. Here are a couple political steps we could take right now to increase our privacy: * Legalize the use of pseudonyms and fake social security numbers on most forms (instead of calling it "fraud"). This would take privacy issues out of the hands of obscure computer databases and even more obscure, loophole filled "privacy laws", and put the choice of privacy firmly back into the hands of the consumer * Fully legalize all forms of cryptography (eliminate obsolete Cold War export controls) Note that pseudonyms are nothing new in cyberspace. They've been a part of BBS culture since day one, and their use has recently grown quite substantially on the Internet. It's mostly the more "official" and backwards large institutions that are so paranoid about people filling out forms correctly. Of course, the first law means that these institutions will have to adjust the way they do business -- for example, many of the people who rely on social security numbers to build up databases will have to get their information some other way (perhaps by talking to people, gasp!) The second law (really removing existing obsolete laws) means, once today's crypto tech is put on the market, that spies and snoops won't be able to wiretap phones or read private e-mail unless the communicators choose to talk in the open. To a first approximation, just about any kind of social structure is technically possible in cyberspace. So rather than just asking what we fear most of it, we should be asking _what do we ideally want it to be_? In some ways we can be very idealistic and very practical at the same time. For example, in a purely virtual world, Gandhi's ideal of nonviolence, long considered impractical, is perfectly realized. Any kind of physical violence is impossible in cyberspace. Incredible! When it comes to creating the 21st century, we can often be very idealistic, and improve our lives by vast amounts in certain ways. Of course, we're going to still live much of our lives in the physical world, so this is hardly a complete solution to all our woes, including war and violent crime. But when it comes to creating new realities in cyberspace, we can quite often be very idealistic and very realistic at the same time: there we don't have to live by the same constraints we live by in the physical world. The great thing is that some of the most important issues of our modern world, such as privacy, are mostly issues of how we set up cyberspace. This means we can solve them in ways radically different than we would solve mostly physical world problems. Also, instead of asking what "we" as a society want, I think it's much better to promote diversity in cyberspace, and ask what you and your friends want. Let different cultures set up radically different new communities. Even if there is such a thing as a "best" kind of community, the only way we're going to find it is by trying out many different kinds and comparing them. Only God would have the omniscience needed to figure it out in advance. Besides nonviolence, my most important ideal is choice. I want, for example, the ability to choose which of my activities remain part of my private life, and which I make public. Of course, I must be willing to pay extra for choices that require expense of others (eg, trying to get credit anonymously might be quite expensive!). Even so, I want to be able to make the choice of whether or not to pay that price for myself, depending on my own unique circumstances. There are two basic arguments here: (1) a moral argument, that freedom is one of our most valuable inheritences, and (2) an information theoretic argument -- nobody knows as much about my unique circumstances and needs as myself and my loved ones. Removing choice from this intimate circle can impose gross ignorance and inefficiency on my life and the lives of my loved ones, as well as loss of freedom. In cyberspace, I think we can set up structures that will allow much more choice in many areas, as well as (and partly because of) our ability to conduct relations in cyberspace with perfect nonviolence. Nick Szabo szabo@netcom.com