From: wcs@anchor.ho.att.com (Bill_Stewart_HOY002_1305)
Date: Wed, 6 Oct 93 15:05:22 PDT
To: cypherpunks@toad.com
Subject: Re: digicash
Message-ID: <9310062202.AA15494@anchor.ho.att.com>
MIME-Version: 1.0
Content-Type: text


Yes, as Bob said, digicash *does* sound like play money,
as any fiat currency is.  Gold and silver are, in many cultures,
useful real money, partly because they're useful or decorative,
durable, and hard to make more of, occasional gold rushes excepted :-)
It's also hard to debase, though some coin-makers have short-weighted their
coinage, or even talked people into accepting cheaper metals.
Leaves are also pretty and conveniently grow on trees, but aren't good money.
Ultimately, it's worth whatever goods and services people are willing
to give you in exchange for it.

Paper money with numbers on it is easy to print more of;
some governments realize that and abuse people's trust of them,
leading to mild-to-medium inflation rates in many countries,
and hyper-inflation in countries that think they can print it faster
than people can switch to less volatile currencies (Israel went through
a period where their pound was inflating so fast that most people
switched over to dollar-based price indexing until the government started
behaving again; governments like the U.S. realize they can steal a lot more
money if they do it slowly.)  I once saw a board game that used real
Nicaraguan 50,000 cordoba notes as money - they were cheaper than 
printing good-looking play money from scratch :-)

Bits are also easy to make more of, though it may be more work if you want
them to be prime numbers or whatever.  In order to make them useful and
not inflated, they either have to be directly pegged to a good or service
that people want, and offered by someone reasonably trustworthy,
or otherwise tied to something of stable value.  Metro or BART tickets
are magnetic cards with bits on them, which people are willing to pay money for
because they know they'll get subway rides in return for them,
and aren't likely to be defaulted on.  Telephone cards are 
easy for the phone to make, but people accept them for the same reason.

For digicash tokens that _aren't_ tied to a single service provider,
you typically need some other reason to trust the issuers not to make 
lots of them - people accept EFT, money orders, travellers' checks, and personal
checks from a variety of issuers because they know they can get regular cash
in return for them, and because business and legal relationships are there
to such an extent that people don't usually default on them, though
merchants often charge service charges to make up for the costs of redemption
and the risk of defaults, especially for foreign transactions.
Digicash could work the same way, only instead of Western Union accepting cash
at one office and telegraphing a request to pay cash at another office,
we're hauling the email around ourselves.  

Authentication theory is fun, and a good subject for university research,
and the economics are important as well, since you'd need to make a profit
to bother running all the computation that the system costs,
but ultimately, it won't happen until some entrepreneur or group of
fanatics :-) decides to get down and do it.  Most of the current similar 
systems either don't give you anonymity, or do so by using pre-paid tokens;
anything fancier is cutting into relatively untapped markets.

			Bill