On Mon, 10 Jul 2000 [EMAIL PROTECTED] wrote:
> both companies seem to make the point that you don't have to trust
> them as a third party, and yet the first thing they ask for is my credit
> card number -
This *is* a problem. Not least because I don't have a credit card. :)
If I recall correctly, the official reason why ZKS hired Brands was
to fix this problem by launching some kind of e-cash. (of course,
you'll probably have to buy the e-cash with a credit card, won't you,
and for a while ZKS may be the only vendor using it...).
>
> an economy based on "nyms" is a pipe dream. No human has ever purchased
> a car, or purchased a home, or taken out a loan, or started a business,
> or gotten a job by using an anonymous "nym". Any significant economic
> transaction, in both the real world and the virtual, requires accountability,
>
What do you consider a "moral person" - i.e. a corporation?
It may be helpful to introduce a distinction between anonymity and
pseudonymity. In anonymous communication, the message comes from 'no one',
and messages are not linked together. In pseudonymous communication,
the messages can be linked together by some "nym." That nym may
resolve to a 'real' person or a True Name, but it doesn't have to.
This does blur a bit in some cases -- a concrete example would be
the mysterious anonymous posters to coderpunks and cypherpunks who
like citing academic papers in cryptology. From the writing style
I would guess the same person is writing all these messages, but
there's no way to be sure; he she it or they have declined to sign
them. Nevertheless, the distinction is useful in that it allows us
to see how we can still have -- and better yet, manage -- identity.
> and accountability is only engendered through identity, not anonymity.
> (the anarchists, I'm sure, will cringe, but it IS a sobering fact
> of this reality that we all live in)
You get accountability with pseudonymity. Especially if there's a high
barrier in terms of reputation to entering a market. IBM is very much
accountable for the "actions of IBM," despite the fact that there
isn't any one person you can point to and say "that is IBM." Sometimes
you can trace particular decisions to people within IBM if you're
persistent, but usually the best you can do is point at the chairman
and CEO, watch them get axed, and replaced with someone just the same.
Despite the fact that "IBM" doesn't resolve to anyone in particular, it's
one of the best-known nyms in computing.
ZKS is selling nyms and pseudonymity. You get 5 per pack. Yours to build
into IBMs or turn into net.kooks as you like.
>
> it seems a more appropriate use of cryptography to enhance privacy
> would be to make sure that any transaction I partake in is REALLY
> done by me.. identity theft is the worst privacy violation in this world,
> not IP logging...
Agreed. This is why I think that authentication and signatures are more
interesting than encryption. For most people, there is more to lose if
someone steals your name and sells your house out from underneath you than
if they read your HD. of course, reading a HD can help with selling a
house out from underneath someone...
In addition, the issues surrounding signatures seem to be thornier than
those surrounding encryption. For instance, losing a decryption key can be
very bad, but losing a signing key means you've lost control of your
identity. Even *prior* to this, we have the well known problem of
someone putting out another key with your name on it. How do we interpret
digital signature laws to account for this? what kind of assurance are
"we" going to require that a key belongs to the "right" person?
>
> privacy IS NOT synonymous w/ anonymity (again, the anarchists will
> cringe, but it's another sobering fact of reality), but rather with
> controlling the personal information that the world DOES come to know about
> you..
That is why pseudonyms and credentials are important.
If I use the nym "David Molnar" to interact with you, you can go look up
all the data associated with it. There's a rather large amount. I have
only a fuzzy idea of how much; I try to do a web search on myself now and
then to keep up, but that's just the tip of the iceberg.
Sometimes suprises pop up; for instance, some guy in Germany once
included a link to an article I wrote in his project on the CDA and called
me an "Uber radikal-konservative stromungen in der Amerikanschen Politik."
You might see that and decide that I am completely, unreservedly, totally
whacked. (you wouldn't see that I wrote the article at 16, though).
I might have no idea that you saw that, and thus no way to reply when
you start calling me a heartless creep and break off contact.
This cuts both ways, though. You might also see my posts to sci.crypt
and come to believe I know something about cryptography. This might
encourage you to solicit me for a job, despite having never met me, or
even knowing if I "exist." I don't mind *that* particular consequence of
having a trail.
If I had and used the nym "[EMAIL PROTECTED]" to interact with you via
ZKS Freedom, then you would get much, much less from a web search. Plus,
since I know that "[EMAIL PROTECTED]" is going to be tracked and profiled
on the web, I can carefully control what I use it for, and be much better
informed about what you might see. It gives me a measure of control
over what personal information you see, and the ZKS system is supposed
to prevent you from breaking that control by linking "[EMAIL PROTECTED]"
and "David Molnar."
That's what ZKS can give you right now.
The next step would then be to work out a way to transfer trust from
"David Molnar" to "[EMAIL PROTECTED]" without linking the two nyms.
This is where credentials come in handy; this is perhaps what ZKS is
working on next (I don't know). For instance, maybe "[EMAIL PROTECTED]"
wants to take out a loan from you. "David Molnar" has a good enough credit
rating, but I don't want to show you that credit report.
We could work out a system by which the credit bureau issues "David
Molnar" a signed credit report of a certain rating R, with these
properties :
* Neither "David Molnar" nor "[EMAIL PROTECTED]" can
determine what exactly rating R is or how it was calculated.
* The nym "[EMAIL PROTECTED]" can execute a special protocol
by which it 'shows' you the credit report in such a way that
- you learn whether R is higher than your required
threshold
- you don't learn what R actually is
(although if you're allowed to play bar-kochba/20
questions, you can get it easily of course)
- you can tell that the credit report is really
from a reputable credit bureau
- you can tell that the credit report really
does "belong" to "[EMAIL PROTECTED]"
- but there's no way for you to link this credit
report to "David Molnar"
- and you can't prove anything about
"[EMAIL PROTECTED]"'s credit to anyone else.
(although you can always tell people about it,
without proof. which may be just as bad)
In this manner, trust in the form of a credit rating is transferred
from "David Molnar" to "[EMAIL PROTECTED]" without linking the two.
I haven't explicitly worked out how such a protocol would work, but
it's the kind of thing which is plausible in cryptography.
(tangent -- is anyone interested in the actual details of such a thing?
can we do this with brands or chaum style credentials trivially?)
ANYway, this is the kind of thing which ZKS is or should be aiming for -
the ability to create and manage multiple identities as you see fit,
without losing the benefits of having a nice monolithic, stable,
creditworthy, big puppy dog identity.
does that help at all? I know I've omitted any discussion of Privada,
but I know about 0 about them.
-David
