August 05, 2005

tracking tokens

Wired and Boston report on a new mechanism to read a fingerprint from paper. Not the fingerprint of a person touching it, but the fingerprint of the paper itself. Scanning the micro-bumps with a laser is a robust way to create this index, and it even works on plastic cards. (Note, don't be distracted about the marketing bumph about passports, it is way too early to see where this will be used as yet.)

Our recent pressed flowers adventure resulted in a new discovery along similar lines - we can now make a token like a banknote with cheap available tools that is unforgeable and uncopyable. It does this by means of the unique identifier of the flower itself; we can couple this digitally by simply scanning and hashing (a routine act in the ongoing adventure of FC). What's more, it integrates well with the pre-monetary economics that is built into our very genes, if you subscribe to the startling new theory presented in "Shelling Out," a working paper by Nick Szabo.

In other tracking information, the EFF has started tracking printers that track pages. Some manufacturers print a tiny fingerprint of the printer onto every page that gets produced. Originally "suggested" by monetary authorities so as to trace forgeries and copies of paper money, it will of course make its way into evidence in general court proceedings. Predictably the EFF finds that there is no protection whatsoever for this.

(older DRM techniques by the ECB.)

From the benches of the Montana Supreme Court, a judgement that instantiates the Orwell Society. Dramatically written by Judge Nelson and brave for its refreshing honesty, it recalls that famous line from Scott McNeally, "you have no privacy left, get used to it." It's worth reading for its clear exposition of how your garbage is abandoned and therefore open to collection by ... anyone. But it should also serve as a wakeup call to the limits of privacy. Judge Nelson writes (from Wired):

In short, I know that my personal information is recorded in databases, servers, hard drives and file cabinets all over the world. I know that these portals to the most intimate details of my life are restricted only by the degree of sophistication and goodwill or malevolence of the person, institution, corporation or government that wants access to my data.

I also know that much of my life can be reconstructed from the contents of my garbage can.

I don't like living in Orwell's 1984; but I do. And, absent the next extinction event or civil libertarians taking charge of the government (the former being more likely than the latter), the best we can do is try to keep Sam and the sub-Sams on a short leash.

In such a world, we should be delivering privacy, as we cannot rely on anyone else to do it. In this sense, the recent (popularish) argument between Phil Zimmerman's new VoIP product and some PKI apologists is easily defended by Phil as such:

My secure VoIP protocol also requires almost no activation energy, so I expect it to do well.

No more need be said. Go Phil. (See the recent article on Security Usability (PDF only, sorry) for IEEE's Security & Privacy mag as well.)

Posted by iang at August 5, 2005 07:00 AM | TrackBack

Reliably trackable, cheap, unique and unforgeable physical items have been research topics for decades, initially to help enforcing the nuclear disarmament regime (the "potential adversary" would put such a mark on each missile and warhead of the counterpart to be able to count them reliably). They typically spray-painted the items with epoxy-glue and used the arrangement of the drops as an identifier (and then voiced suspicions that the other side made impression-prints of the pattern and forged it).

There are other, better techniques, too, which never made it into arms-control applications, but most published ones are quarded with patents. My personal opinion is that devising such techiques is easy enough to come up with a new one when it's needed. One that I liked particularly (and which is protected by microsoft a patent, AFAIR) is throwing short (as in the 0.5-3 mm range) glass fiber pieces in some transparent glue (with a different refraction coefficient), and cover the whole thing with a protective sheet of glass or transparent plastic.

It is very easy to scan, as the fiber pieces, when illuminated on one end, shine on the other. The piece is either present or not (no problems like those with small, ambiguous drops of glue), and the counterfeiter has to accurately copy the arrangement and the length of each piece of fiber-glass. It is extremely inexpensive to produce , and hideously expensive to accurately reproduce. Scanning the fibers in a piece of paper is even cheaper to produce, and more expensive to forge, but identification is far less reliable, IMHO. It is also more vulnerable to wear-and-tear.

There are other methods as well (pressed flowers are one, though reliable automated recognition is no trivial matter in their case), and one can come up with new ones, in order to avoid the patent minefield.

As for tracking printed documents by marks left by the printer: I don't believe that printer manufacturers are easily convinced not to put a trivial way of disabling this "feature" into their products. Remember the DVD region codes and the players that could play anything after removing a jumper? I would expect the same with printers. There is a clear incentive issue working against Big Brother, as usual. Maybe because of having lived under a government that was so desperate to track and control every activity going on in the country, I am less afraid of Big Brother than many people I meet in the West; he (BB) has to swim very hard against the current of various economic incentives of the subjects and even his own bureucrats and enforcers. There are always huge cracks in these systems and they get wider and wider as time goes on.

Devices with crippled and/or easily disabled tracking mechanisms tend to sell better and cost less to produce than the carefully designed, "perfect" ones, so go figure. Some EFF-style paranoia is good, but in the end it is the police state or the control-freak corporation, who fights the up-hill battle against almost everybody including many insiders. If you really need privacy (as in being ready to pay for it), you can almost always get it. Of course, you will have to excercise a little more care than usual and pay the right people along the way. Free privacy may be in danger, but there's no such thing as a free lunch anyway (actually I like the Russian version of this proverb better: "Free cheese can be found only in the mouse-trap").

Posted by: Daniel A. Nagy at August 8, 2005 12:49 PM

One more thought for developing the above argument: hoping that people actually pay for their own tracking is even more absurd than what the Soviet government did: they actually paid for all the nasty stuff they wanted us to have. And even that (subsidising inferior products that are less attractive and more expensive than the alternatives) is no more than a recipe for imminent bankruptcy.

Although I am often accused of assuming people to be more rational than they actually are.

However, my experience is that when it really matters, most are pretty rational. And this very much includes criminals and terrorists, too. Moreover, it includes government and corporate officials, whose goal in pushing for hopeless objectives is usually not achieveing them, but getting paid for trying or handing out pork (which are two sides of the same coin). And this they do achieve with enviable efficiency.

Posted by: Daniel A. Nagy at August 8, 2005 01:20 PM

IF you want to read about the right and desire to transact as fundamental human right, a core driver of history, and would like to be genuinely informed in the process, read Development As Freedom. In this work, Sen defines the freedom to transact as a fundamental freedom and illustrates how it is intergrated with other freedom which tend to be classified as political (e.g., speech) or cultural rights (e.g., the right to love across various socially constructed barriers). His aside about the rationality of asking those in power if change is in order as much for questions on economic change to multinationals as it is for questions of cultural changes in isolated third work environs.

As I recall, Nick Szabo's paper is deeply flawed. For example, tobacco was used for decades in VA as a currency. It was Gresham’s law that defeated its' use (as well as storage limits). That is, people began to grow lousy tobacco for domestic currency use rather than using the substitute of export-quality tobacco as currency. He is the only person I have read who seriously argues that it was the international balance of payments and the interaction of the manufactured beads & the pre-industrial Indian currencies that were the drivers for ending this currency. (He was serious, right?) You do not serve your readers well by continuing to point to such an article which is at best bad scholarship and at worst ideological misinformation.

As for having no privacy - it is the marketplace at its finest. Personally identifiable data (pid) are property, and as such not inalienable. You have NO INTEREST in your proprety after it is sold. You can no more tell ChoicePoint to forget your information than you can go back and tell the new owner of your home to paint the living room sage.

I found the recent disclosure that 100% of ChoicePoint records had errors to be of interest. To the extent that pid are money, then good money will be driven out. That is, when there is no control over your information other than lying people will lie. The lies will filter into the databases, and viola we are all CEOs of companies we have never heard of according to data brokers. Individuals may be rational but companies are schizophrenic almost by definition - they behave like machines following processes regardless of how appropriate the responses may be. And their data-compilation process are rarely argued to be optimal.


Posted by: L Jean Camp at August 9, 2005 12:15 PM


Your criticisms fall only loosely towards the target. Szabo doesn't discuss tobacco in that article, and it's pretty clear that it's not part of his thrust because he's talking about collectibles, not smokables. Your comment about balance of payments is equally odd, about all Szabo claims is that the English starved their colonies of coin, until about 1661. That hardly qualifies as a balance of payments question, unless I'm misunderstanding something. Unless he got those facts that he asserted wrong I can't see the problem.

It may be that future researchers show his claims and conclusions ludicrous. This is the risk of the polymath; one has difficulty in mastering all the fields, and this particular article crosses monetary economics, psychology and pre-society history.

Also, I'm not sure what to make of a post that criticises something as ideological and then goes on to fall to the same assumptions. E.g., use of "freedom" is a particularly american ideology, and is hard to translate out to the rest of the world as nobody knows whether it means "rights we agree that humans hold" or "bases for manipulation." Americans use the two interchangeably, others find this as doublespeak, contradictory and intractable. A book that is entitled "Development as Freedom" is immediately marked as a text marketing to the American audience, although hopefully this is not too much of a drag on its contents. (Before I read that book, I'd like to track down deSoto who more addresses property rights.)

Likewise, your claim that "personal data is property" is an American viewpoint that is only agreed upon inside America and those places that benefit from that viewpoint. Outside America, that claim is wide of the mark.

Europe Zips Lips; U.S. Sells ZIPs

What is an open question for future is whether the American view will triumph, or the "European" view will survive. Economics or Rights? The stage is now set for that battle. To predict mildly, the foreigners don't have an issue to deal with at the moment, while the Americans have a huge problem ("identity theft") so I'd say the betting today is on the Americans changing their view. But who knows?

Posted by: Iang at August 11, 2005 07:02 AM

Of course data are not property only in places where there is no data protection. Data privacy can be about seclusion, harm, autonomy and boundaries. Data are propriety in America does not imply that they are property in Canada or the EU or any other place. There are some special protected arenas in the US where data are not property (the most entertaining being video rental records). And we both know that I know that. ;->

I have long hoped that the US would change its collective mind. I believe a data protection regime in the US would vastly increase Internet participation and thus efficiency and thus wealth. But obviously the rulers of the merchant class do not agree.

NS' fundamental conclusion appears to be that economic transactions are a core element of the human experience. Sen does a much better job making that argument, and making in a context of development, food security, human rights, and economics. So, given Sen has a Nobel Prize for his observations (in 1998) I would not call that conclusion Nick's. Rather Nick has picked up an old heme and wrapped half of it in new ideological clothing.

FYI, the author of Development as Freedom is from India, and his expertise is international economics. So his view is global.

Sen's core argument is economics are rights. That no freedom to transact means a limit to autonomy. Freedom consists of a set of capacities; economic transactions are one part of this. In many places in the book I thought you could put the word "identity" in place of "transaction" and have an interesting claim. Sen's brilliance is to integrate economics and rights, and to bring those together with a graceful, coherent theoretical line of reasoning.

My core problem is that I learned nothing from Nick's article, and I found unsupported historical assertions (the fall of the Roman empire, pls). In contrast, I learned a great deal from Sen's book. The book predates the article and goes into greater depth that Economics as fundamental elements to the human experience – economics as rights. Of course the implications for currency are not his focus, but in many paragraphs I could pencil in “identity” and see a strong implication.

Posted by: L Jean Camp at August 11, 2005 02:58 PM

I agree with Jean that people lie to Big Brother all the time and that a lot of this purportedly "extorted" identity is not worth stealing. For example, have you entered anything remotely true in New York Times' soul-sucking registration form? I sure haven't.
This newfangled Homeland Security office asks a written statement from everyone entering the US about where they want to go. Guess what, people make up all sorts of destinations having very little to do with where they actually want to go.
Ian has once posted here an interesting entry about lies in the job market. I think, it would be interesting for economists to study economically motivated mis-information and its effects on the society. So far, I know only one such study, done by John Kornai about lies in the planned economy. I think, most of his results apply directly to capitalism as well. But there's a lot more to mis-information than mis-representing one's resource demands in a planned economy, resulting in shortages and waste (btw, have you noticed the incessant shortage -- and waste -- of paper, paperclips and staplers in corporate offices? It is for the exact same reason -- planning.).

False signals, fake identities and others are also worth studiing.

Posted by: Daniel A. Nagy at August 11, 2005 05:24 PM
Post a comment

Remember personal info?

Hit preview to see your comment as it would be displayed.