Why Digital Cash is Not Being Used

Hal Finney asks us to think about and comment on the important issue of why
digital cash, in its myriad forms, is not in wider use. Especially on this
list, where the Magic Money/Tacky Tokens experiment has not (yet at least)
produced widespread use.

This question also goes to the heart of several related questions:

1. Why aren't crypto protocols other than simple encryption, digital
signatures (both implemented in PGP as the de facto standard in our
community), and remailings (implemented in Julf's anon.penet.fi remailer
and in the various Cypherpunks remailers) being *used*? Why no DC-Nets, no
data havens, no digital timestamping, etc.?

2. What *incentives* are there for creative programmers to devise and/or
implement new crypto protocols if essentially everything for the past year
and a half (since the fall of 1992, which is when PGP 2.0 and remailers
became widely available) has languished?

3. What are the "killer apps" of crypto?

4. What platforms and user environments should would-be developers target?
What machines? What networks? What languages? (An ongoing interest of mine.
Objects, scripts, Visual Basic (!) VBX tools, TCL, perl, many platforms,
etc. A tower of Babel of confusion is upon us.)


Here is my first-cut analysis of the digital cash situation.

I. Why is Magic Money/Tack Tokens, in particular, not being more widely used?

- Nothing of significance on the List to buy, hence no incentive to learn
how MM works. (Just because someone announces that their new article is
available for 10 Tacky Tokens doesn't a demand make!)

- Semantic gap. I confess to not having the foggiest ideas of how to go
about acquiring Tacky Tokens, how to send them to other people, how to
redeem them (and for what), etc. Having nothing to buy (no need), and
plenty of things to occupy my time, I've had no interest in looking at MM.

When I buy items like t-shirts from people on this list, I simply write
them a check and send it. Very simple. The banks handle the complexities.
And writing a check is a "prototype" (or script) that is learned early by
most of us. Not so with any of the various digital cash schemes. In 10 or
20 years, sure, but not now.

This is not to take away from the excellent work--I gather from comments by
others--that ProductCypher put into MM. His greatest achievement may turn
out to bring this issue to the fore, to wit, what will cause people to
bridge this semantic gap (understanding) and actually begin to *use* these
new constructs?

- as others have noted recently (and this is a well-known issue),
alternative currencies must offer some advantage over existing currencies,
or at least be roughly on a par with them.

For example, the airlines have their own currency, "frequent flier miles,"
which they pass out as an inducement for customer loyalty (repeat
business)....it is generally not advantageous for them to allow exchange.
(And really it's a kind of bribe, a transfer from the corporations which
pay for the plane tickets, with the frequent flier miles accruing--despite
futile attempts to halt this--to the individual passengers....this gives
"ffm"s a built-in advantage.)

(The proposal recently that vendors of products, like t-shirts, give a
discount for MM payments is of course unworkable. This is asking real
people to give up real dollars for an ideological cause of marginally
little significance to them. The advantages of MM must be real, not phony.)

II. Other Experiences with Digital Cash in Some Form

- On the Extropians list a while back (I've since left that list), there
was an interesting experiment involving reputations of posters and "shares"
in their reputations. Brian Hawthorne introduced is "Hawthorne Exchange,"
HeX, with eventually a few hundred or so reputations trading. The unit of
exchange was the "Thorne," with each new list member given 10,000 Thornes
to trade with.

Trading was very sparse, with most people apparently never bothering to
learn to trade (a la my own experiences with Magic Money). I downloaded the
docs one night, tried a few trial trades, and then proceeded to make dozens
of trades, trying to buy cheap and sell dear. Between my trades, the
reputation attached to my posts (and to my "nom du humor," Klaus! von
Future Prime) I amassed a sizable fortune in Thornes. I even offered to
exchange real dollars (checks) for Thornes, the better to amass a fortune
(for reasons I won't go into here). Edgar Swank offered to sell me his
Thornes for $20, I think it was, and I sent him a check immediately. (No
one else did.)

But I think the system was ultimately a failure. Nothing interesting was
for sale, and Thornes had a ridiculously low value (reflecting of course
their "toy" nature...my $20 bought 20,000 Thornes, as I recall). By "low
value" I mean that the number of Thornes given to each participant (Hint:
"given" is the important word) was worth nominally $100 (by Brian's sales
price--probably none were ever sold at this price), worth $10 to me and
others (by my offer of $1 per 1000 Thornes), and probably worth much _less_
as the HeX market languished and, probably, ultimately folded. (Does
anybody on the Extropians list know if it is still operating? And what
happened to by shares when I left the list?)

- Similar barter schemes have been described elsewhere. "Mother Jones" had
an interesting article last summer about a barter scheme in New England,
and other folks have mentioned here the articles in "Utne Reader" and so
forth.


III. What Markets Might Make Use of Digital Cash

- phone cards, subway cards, parking garage cards...all are examples. But
these are mainly to reduce the need for customers to carry coins and bills,
to reduce the dangers of theft of coins and bills (and the need to collect
them frequently from payment points), and to speed up processing by not
having customers fumble for change, etc.

- toll roads...this is a market that Chaum's DigiCash company has been
targeting for several years now. Privacy is a concern (don't want Big
Brother tracking your movements), and the infrastructure may allow
considerable investments in remote sensing of IDs and pseudonymous IDs,
online clearing, etc. Read the Chaum stuff for details on this.

- illegal markets, for transferring wealth in fairly large amounts. Not at
all clear how this will happen, and it sure won't happen with some
fly-by-night hackers and/or students offering a new service.

(I didn't mention that one of the persistent concerns about learning new
crypto protocols here on this list is the epiphenomenality (transience) of
it all...remailers appear and then vanish when the students go away or lose
their accounts, features added make past learning useless, and so on. Life
is too short to spend it learning crufty details that will go away in a
matter of months. I'd hate to buy $300 worth of TackyTokens and then find
that their value went away when J.Random User graduated!)

- betting markets, the "Internet Casino in Cyberspace," etc. Nick Szabo was
once championing this, and I think it could be an interesting, and very
real, market. Lots of issues here.

- Digital Postage. This remains my favorite. There's a _need_ for
untraceable payments (else why use a remailer?). I've written about this
extensively, as have others.

If remailers offered robust (see above point about crufty, flaky, hobby
remailers) services that they operated as _businesses_, with reasonable
attention to reliability, interconnectivity to other remailers, overall
robustness, and carefully articulated policies about logging, privacy,
etc., then MM or something similar could have a real value.

IV. Is there Any Hope for Cypherpunks Software Use?

The remailers (of Hughes and Finney, with other contributions) came in the
first few _weeks_ of existence of the Cypherpunks group. Julf's system
already existed.

Remailers were the "low-hanging fruit" that got plucked fairly easily (not
taking anything away from Eric, but he himself says he learned enough Perl
in one day to write the first, crude remailer the _next_ day!).

Later protocols have not fared as well. Why this is so is of great importance.

That's a topic unto itself, and one which I hope to write about soon. Lots
of important questions and interesting issues.

--Tim May


..........................................................................
Timothy C. May         | Crypto Anarchy: encryption, digital money,
tcmay@netcom.com       | anonymous networks, digital pseudonyms, zero
408-688-5409           | knowledge, reputations, information markets,
W.A.S.T.E.: Aptos, CA  | black markets, collapse of governments.
Higher Power: 2^859433 | Public Key: PGP and MailSafe available.
"National borders are just speed bumps on the information superhighway."

Date: Tue, 3 May 1994 11:48:18 -0800
   From: tcmay@netcom.com (Timothy C. May)

   - Digital Postage. This remains my favorite. There's a _need_ for
   untraceable payments (else why use a remailer?). I've written about this
   extensively, as have others.

   If remailers offered robust (see above point about crufty, flaky, hobby
   remailers) services that they operated as _businesses_, with reasonable
   attention to reliability, interconnectivity to other remailers, overall
   robustness, and carefully articulated policies about logging, privacy,
   etc., then MM or something similar could have a real value.

But there's a conflict here.  You'd like to be able to use the same
postage on multiple remailers.  But if the remailers know each other
well enough to agree on a common currency, then they know each other
well enough to remove the reason for using multiple remailers.

[tim writes:]
> I. Why is Magic Money/Tack Tokens, in particular, not being more widely used?

> - Nothing of significance on the List to buy, hence no incentive to learn
> how MM works. (Just because someone announces that their new article is
> available for 10 Tacky Tokens doesn't a demand make!)

> - Semantic gap. I confess to not having the foggiest ideas of how to go
> about acquiring Tacky Tokens, how to send them to other people, how to
> redeem them (and for what), etc. Having nothing to buy (no need), and
> plenty of things to occupy my time, I've had no interest in looking at MM.

> - as others have noted recently (and this is a well-known issue),
> alternative currencies must offer some advantage over existing currencies,
> or at least be roughly on a par with them.


So what is the natural currency to trade in on the Internet?  What is the
medium that is most widely spread across the myriad nodes and networks that
crisscross the globe?  What would someone like to be able to buy, that is easy
to acquire, and offers an advantage over real money?

The answer is quite simple:  information.  We need to find a way to trade in,
and subsequently value, information.  At first blush, this seems an impossible
task, and one that is highly subjective and prone to failure on an individual
level... but in a large enough group of people, there has to be a consensus on
the average value of a 'ware' of information.

Say I have a piece of code that you do not have, that you would like to get
from me - maybe it's something that I've written, or isn't publicly acessible
everywhere on the net.  I tell you that my code is worth 50 wares of digital
money; I have my own signature on the code that signifies that it's mine.
We agree, and exchange currency - Bob gives me 50 wares (with his signature
stripped from them), and I give him my code, with my signature removed.

So what's to stop Bob from replicating it and giving it to all his friends?
Well, bascially, that would devalue the 'ware cost' of the code.  If everyone
has it, it's hardly scarce, and therefore, not economically viable.

Any thoughts?  This is a pretty hefty topic, and I don't have the time to go
into it more just yet... I hope I've given some food for thought.


Zen, philosopher-at-large

"Michael V. Caprio Jr." says:
> So what is the natural currency to trade in on the Internet?

Dollars.

> What is the medium that is most widely spread across the myriad
> nodes and networks that crisscross the globe?  What would someone
> like to be able to buy, that is easy to acquire, and offers an
> advantage over real money?
> 
> The answer is quite simple:  information.

Information is useless as a currency, for five reasons.

1) It is not fungible.
2) In order to demonstrate that you have it you generally speaking
   have to have already given it away.
3) It can decay in value, unpredictably. My inside information that
   Joe Blow is a communist spy is valuable today and might become
   worthless tomorrow.
4) It cannot be effectively loaned or borrowed.
5) It has highly unpredictable value. Two pieces of information might
   be worth the same number of pieces of gold from me, but you may
   find one of them worthless and the other very worthwhile.

Dollars are a natural currency for use in internet trade. So are gold,
D-Marks, Yen, etc. There is nothing wrong with these things.

I'll agree that I don't like government sponsored currencies, but
since everything is denominated in them right now I'd say that they
are perfectly fine.

Perry

Perry writes:
> Information is useless as a currency, for five reasons.
> 
> 1) It is not fungible.
> 2) In order to demonstrate that you have it you generally speaking
>    have to have already given it away.
> 3) It can decay in value, unpredictably. My inside information that
>    Joe Blow is a communist spy is valuable today and might become
>    worthless tomorrow.
> 4) It cannot be effectively loaned or borrowed.
> 5) It has highly unpredictable value. Two pieces of information might
>    be worth the same number of pieces of gold from me, but you may
>    find one of them worthless and the other very worthwhile.

Hmm.  It seems to me that a bunch of these characteristics you've described
seem very similar to a stock market situation.  I would use the analogy of
information as shares...  It also seems that number two is a typical zero
knowledge situation - plus the fact that if I tell you I have a piece of
code that does x - you want the code, and knowing what it does has no real
value to you, if you just want it for its functionality.

BTW, what is fungible?  I've seen this term used several times, but have no
idea what it means.

> Dollars are a natural currency for use in internet trade. So are gold,
> D-Marks, Yen, etc. There is nothing wrong with these things.

> I'll agree that I don't like government sponsored currencies, but
> since everything is denominated in them right now I'd say that they
> are perfectly fine.

I think this is a key point - there has to be a common sponsoring agency,
a "data bank" or something that holds all the keys, and has all the info.
Making a currency isn't really the hard part here - someone could just 
encrypt a textfile that says "This is a five point cyphermark".  All that's
important is the key authentication at the bank, who will be the party who
trades it around ultimately - it's getting people to agree on it, and give it
value that's the issue...


Zen, philosopher-at-large

"Michael V. Caprio Jr." says:
> 
> > 2) In order to demonstrate that you have it you generally speaking
> >    have to have already given it away.
> 
> I would use the analogy of information as shares...  It also seems
> that number two is a typical zero knowledge situation -

No its not. Its easy to conduct a zero knowledge interactive theorem
proof for things that are mathematically expressable, like "I know a
Hamiltonial circuit of this graph", but it won't work for anything
that can't be expressed that way.

Example: construct a zero knowledge proof for the proposition "I know
something interesting about George Bush that you would be willing to
pay $100 to know".

> BTW, what is fungible?

A fungible thing, sometimes called "a commodity", is one for which the
all are oblivious to substitution. As an example, when you request a
dollar bill from me, you don't care WHICH dollar bill you get. When
you ask for a one kilo gold ingot, which ingot from the space of all
ingots doesn't matter to you.

Only fungibles can be traded in securities markets or deposited in
accounts. I can trade shares of IBM because you have no care which 100
shares of IBM you get. I can trade futures contracts for West Texas
Intermediate Crude because thats a very well specified substance.

Currency is ALWAYS fungible. That which is not fungible cannot be used
as a currency. In particular, "information" is not fungible. It is not
a commodity. Two pieces of information are not indistinguishable.

Perry