| bc76@midmaine.com 2005-02-22, 5:45 pm |
| If I have accidentally reposted this, please ignore this post ( reply
to the other one as it was posted first if indeed it was posted ).
However this post may be easer to read as it is hopefully formatted
better.
Perhaps some of you remember E-Cash/E-Coins that were supposed to
provide digital coins that could be spent online as untraceably as real
cash can be spent at a convenience store. The idea was for banks to
take deposits of real money and issue e-coins that could be spent
online.
The patents on E-Cash expire in 2006. This will make it possible for
free software to be written by imaginative people that uses this blind
signature technology.
The topic of this post is: Can this technology be used to help solve
one of the main problems of distributed computing - that of ensuring
that remote algorithms or services are performed correctly through
economic incentives?
The E-Cash business model was that banks would implement E-Cash to
attract depositors who shop online. This has not happened.
But the only neccessary requirement that someone needs to justify the
cost in bandwidth and processing power that would be required to run a
digital coin service is the ability to make a profit, and there is no
technological reason that digital coins must represent real currency
in an account. They could very well be 'minted' out of thin air by
anyone as not neccessarily representing any actual currency. The
coins would then become a mere logical convenience.
But, of course, money is valueless unless it is redemable for some
purpose. Governments ensure the value of the money they issue by
accepting it for payment of taxes. But a coupon for a free Big Mac at
McDonalds also has value and can be traded. So, it seems that the only
requirement that exists for someone to mint digital coins with value is
for them to accept them back in return for some useful service. Of
course everyone must pay taxes, and so everyone has a use for the coin
of their realm, but many of us have use for a big mac coupon too.
If someone opened up an exchange counter with a cash/big-mac-coupon
exchange rate, then it would make accepting big mac coupons in leu of
dollars even more attractive, although the inconvenience of doing the
exchange would tend to discount the value of big mac coupons.
Perhaps there might be a company that provides a service like say
network routing, or file caching, or maybe just access to some nudie
pictures, to the public in exchange for some digital tokens that the
company itself minted. That company could then offer to buy
distributed computing power or hard drive space from the public using
the tokens, selling that computing power/storage to researchers etc in
order to pay for bandwidth, servers and profit margin.
Of course that would give the public an incentive to cheat ( by not
providing the services, or not providing them correctly whilst still
getting paid in tokens ) but it would also create a pool of people
owning valuable tokens. If the software to store and spend the
digital tokens were freely available, then there is no reason why the
public might not spend the tokens amongst themselves in exchange for
other services ( like hard drive space, or nudie pictures ).
Except - for cheaters who take payment and do not provide the a
satisfactory service. In a real economy, there are examples of this:
You might get a Quarter Pounder With Cheese when you ordered a Big Mac
at the drive through. Most likely, you do not take the sandwich back,
you just eat it.
Or you might buy a computer game at Wal*Mart and find that it will not
install correctly on your computer even though your system meets the
requirements on the box. Even though you have the legal right to take
the software back, and get another game, the fact that you would have
to hire a lawyer to sue Wal*Mart for the price of the game to get them
to allow you to do anything but exchange for another copy of the same
defective game means that you most likely eat that too.
Legal recourse exists, but it is rarely used in comparison to the the
number of disputable transactions that occur.
Even so, I have personally never found a McDonalds that will not take a
mistaken order back and give you the right order. Once or twice I
have even gotten a free apple pie. Though nowadays I realise that I
may also be getting a hacked up loogie too, and just eat the mistake
rather than eat an on-purpose. But spitting in customers food is
against official McDonald's policy. If it happens in reality, forcing
me ( if I act prudently ) to eat a QPC sandwich sometimes when I
really wanted a Big Mac, then I can't really blame McDonalds the
company. Their heart, in this case, is in the right place.
The frequency of incorrect orders does affect McDonald's reputation.
Their willingness to correct a mistake, and the free apple pie also
affect their reputation. Wal*Mart's reputation is affected by it's
policies too. But 'some pseudonymous internet guy' has no reputation
whatsoever. Odds are, he's nothing but a honey pot designed to get
you to front your money who will then just vanish.
What to DO? If you blacklist the account, the twit will just create
another one.
The answer, in a word is Whitelists. Of course whitelists are a pain (
read costly ) to manage because they can not be done with any assurance
of quality without at least some human intervention. But no need to
worry: the lists will be maintained if it is profitable to do so. And
is could very well be profitable if the whitelist maintainer charged a
mediation fee. This would allow them to take a cut of the transaction
in exchange for maintaining reputation information on the two
pseudonymous parties.
Dirty dealing would lead one party to file a complaint which would
bring down the other parties credibility score. Perhaps premium-fee
reputation reporting agencies would provide decent human investigation
of complaints. An el-cheapo broker might leave someone *no recourse*
to an unfavorable rating without a seperate investigation fee.
The function of 'the police' would be being done privately in this
virtual arena. One's cooperation with the private investigators could
also affect ones reputation giving the investigators something like
'police powers'.
Would one reputation reporting agency tend to take over, or would the
economic incentive of facilitating the most transactions cause them to
share data? How would they be similar to/different from credit
reporting agencies currently are? How might two well regarded ( by
their own favorite reputation reporting agencies ) easily conduct a
transaction without being known to each other's reputation reporting
services? ( software obviously - but would the agents cooperate or try
to corner the market for transactions with their own 'customers' )
The profit motive would seem to be sufficiently tempting for
intelligent human entrepreneurs to work to solve the various problems
of and provide infrastructure for a distributed network based on
sharing of resources in exchange for digital coins.
Money seems to be logical mechanism of choice for the real world to
allocate resources to distributed networks of real needs, wants and
sources of value. It seems that digital currency might be able to do
the same for distributed networks of digital computers, but the virtual
economy need not resemble the real economy too closely.
In the real economy, government paper ( or electronic bookkeeping in
units of government paper ) is used almost exclusively. Might a totally
virtual economy have room for a diversity of currencies without
sacrificing convenience?
Of course anyone that mints a virtual coin gives it value by accepting
it in return for a service, but a minter of coins would still be able
to collect some segniorage from those who held their coins by minting
more. After all, a Pentium 1 can only provide so much computing power.
A 56k intermittant modem connection is only so valuable as a network
router. The number, and hence the value of the coins one mints would
be dependant on one's ability to make good on the promise to accept
them for some service.
Going to an exchange counter, looking up an exchange rate, deciding to
accept Yen for the candy bar you are selling, and then going back to
exchange the Yen for Dollars with which to buy other stuff is a pain in
the butt. Most convenience stores would not even bother to accept yen
for this reason. Because there is no exchange counter for McDonalds
free-big-mac coupons, you will not find a Burger King that lets you pay
for a Whopper with one.
But in the virtual world, exchange is easy. It might matter very
little what currency you have. You could probably spend it anywhere.
Electronic exchange counters could make the currency differences
totally transparent. And if that is the case, what is to prevent
everyone from minting their own electronic currency *and being able to
spend it easily*?
Exchange counters might also be directories of services, which would
then also be reputation reporting agencies. A person would create a
provisional pseudonym with an authentication service. The person would
use their provisional identifier to register with an exchange counter.
Because the authentication service needs to get paid, it might charge
per authentication fronting an initial login to create the
registration, or it might charge a monthly fee. The fee would be paid
in newly minted coins backed by the pseudonym which would then be
registered with the exchange.
The person with the newly created pseudonym would then advertise
services that their coins are redeemable for. Certain information
about the volume of service and type might be provided for a pseudonym
if the owner of the pseudonym chose to publicise that information to
make their coins more valuable. Circulating coins could be speculated
on in a currency market.
Suppose Megacorp wanted to provide packet routing as a service.
Likely, many packets would be routed by their million$ worth of
machinery and infrastructure and their volume statistics would show
that. Maybe they would choose to identify themselves as actually
being associated with the real company AT&T in a credible verifiable
way.
Suppose also that Joe Shmoe has a Gigabyte of free hard drive space.
His pseudonym is 'Some Guy' and over the last month he has managed to
sell space on it but only 20 megabytes worth. Suppose also, that he
uses a dialup connection, and was offline three of the last four times
the owners of the data tried to access it. If AT&T/Megacorp has
minted 1 million Megacorp bucks, and Joe Shmoe has also minted 1
million 'Some Guy' bucks then 1 Megacorp buck will most likely trade at
1 million 'Some Guy' bucks per each or probably more.
A directory of services would force service providers to appeal to
buyers which in many cases would be automated programs. The automated
programs might very well require certain types of info about the
service provider. For instance, a Kazaa-like client might only want to
buy songs from cable or above bandwidth servers only. The client might
know of a directory that provides reliable bandwidth information and
choose to shop only there.
But in the end, you could have all the privacy you could afford. Just
keep your information secret. It might be tough to make much e-money
that way though.
Of course you could have another less private pseudonymous 'sponsor'
that was really yourself. E-cash would allow some anonymity. You
could sell some of your own coins for a variety of other people's
coins and then give them to your anonymous ID. That ID could then spend
them without it being too obvious where they came from.
Then there are legal considerations: If someone exchanged real money
for e-coins, then that would strain the e-coins' status as a mere
logical convenience. It would be best to exchange real money for a
service capability and then use that capability to back minted e-coins.
But even then there may be no legal distinction... And taxes.
Yikes!
The thought of the Cops raiding someone's bedroom because their Rouge
stole some Wizard's virtual belongings seems comical, as does the
thought of taxing Everquest dough, but when real money is involved,
that stuff would happen.
Even with a non-free e-coin system with a "You agree waive all legal
recourse for virtual damages" TOS might not survive when the government
saw that most of the economy had gone virtual to avoid taxes. Does the
contract The People's Court has contestants sign really prevent anyone
from suing in real court if they don't like the verdict? If we all
agree that stealing is legal and tolerated, but only frowned upon, in
X virtual world with the expectation that one is responsible themselves
for defending against 'theives and con men', is it really legal to
steal virtual goods? Might someone really be arrested for stealing
Everquest dough when 'stealing' is meant to be part of the game?
This post was just meant to provoke conversation and ideas. Any
thoughts would be interesting.
|