Smart Money
July 2 1999
Anselm Hook
Introduction,
Imagine a form of money which could only be spent on causes
that you believed in. That if you hired
somebody, paid them money for their labour, you could still have some
fractional say in how they choose to in turn spend the dollars you gave
them. By such a mechanism you could in
some small measure steer your work energy - the dollars you traded - away from
causes that were harmful to yourself.
This essay argues that even if you give somebody dollars, that those dollars
are not entirely freed and liberated, that those dollars ultimately still
reflect your work energy and that they should in some small way still reflect
your code of ethics even if they are out of your hands.
Consider what a strong role money has in our society. Money has become the glue which ties
together the rest of our lives.
Everything has become an expression of the dollar. When we intend goodwill to a third party we
often donate money to their cause. We
spend money on things we are passionate about, and we use the dollar to vote against
causes we do not favour.
Money hasn't always been so much at the core of all
transactions. Even a hundred years ago
things were different for many people.
It used to be that our fial relationships, our affiliations and
associations were built on top of the long term acquaintance of the respected
individuals of the communities we lived our whole lives within. We would contribute energy to persons we
knew were of sound moral character, and we knew that they wouldn't use our
contributions in a way that would harm us.
The small towns of our youth were rich in alternative currencies - debts
of honor, relationships of trust. Today
that is in a large sense gone - the sheer liquidity of cash means that if we
give a person a dollar, there is some chance that they might use that dollar in
a way that we would never have agreed to.
Our society lacks the follow-through of the implicit values and trust we
placed in other persons. We believe
today that a person can do whatever they choose with their own dollars and
trust in the broader legal system to protect and defend our principles and
freedoms. Because of this we can't
choose to favor certain kinds of work or behavior over other kinds of work or
behavior. If a neighbour was going to
dam a local stream one might choose not to contribute to that effort, however
if that neighbour earned dollars - rather than a debt - they have lower
impedance barriers to acting in directions that might be harmful to
oneself. We see this every day in our
society. Recently in California there
has been an effort to buy back some of the redwood forests that form a local
watershed. The forestry company in
question technically owns these resources and has few impendances to doing what
they wish with them, even though the total cost to the community is
unreasonably high.
Today we are on the verge of having the kind of technical
infrastructure which will allow us to have smarter money, money that can carry
a set of desires and ethics along with it, and which can in some degree continue
to reflect your own moral principles and values even when no longer directly in
your hands. This is a kind of money
that could report back to you what it was spent on, or could create
transactional penalties for being translated into uses which you fundamentally
disagreed with. Such currencies could
form a powerful basis for decision making, and could help our society engage on
large scale endeavors such as improved Internet capability, manned
space-travel, aid for third world countries, improvements in health and
education local to your community.
These kinds of currencies could also have a reciprocal effect - people
who believed in certain ideas would have to back up their arguments - to be
willing to create dollar encumberances that might make their dollars somewhat
more expensive to certain buyers in their community. The values of a geospatial or internet community could become
explicitly bound by the licensing agreements and transactional exchange rates
associated with their dollars. The burden
of responsibility would shift away from formally declared constitutions and
legal structures to those values entrenched in the dollars themselves.
The
Exchange mechanics,
Technically the challenge is to design a currency exchange
mechanism that allows both buyers and sellers to express their values and the
work energy they are interested in, and to then allow the floatation of those
values in combination with work energy on a value exchange system. We want to formalize the colloquial values
of our community into the currency itself.
The currency itself becomes a digital "agent", advocating for
certain interests and negotiating on your behalf with other agents. At the end of successful transactions the
two parties trade ownership of their respective agents in such a way that the
identified goods the agent represent still exist but are now in differing
hands.
A smart money agent can represent a bid to offer. It is backed up by some discrete physical
resource such as access to computation, access to work labor, or access to
physical goods. Alternatively an agent
can represent a bid to purchase. In
this case there is a transaction that takes place. If one party offers labor then the other party typically offers
dollars in exchange, or service in kind.
When the transaction has taken place then the "ownership" of
the agents is exchanged. The same encumbrances
still apply but now the ownership has changed.
The agent resides in a computational sea that consists of
points of intense computation separated by relatively low bandwidth
connections. It behooves agents to be
distributed and to be fully self contained therefore since the cost of having
to always return to a home base to confirm statements of fact is too
expensive. Under this model the agent
is distributed over the Internet and can reside in a multiplicity of databases
- both central servers and on client machines.
Agents may exist in duplicate even though they may all refer to a single
offer. The servers themselves may only
be tailored towards accepting certain agents - for example there may be servers
that are only interested in contract positions in the computer engineering
industry. Agents may to replicate
themselves to all servers that allow them entry. Agents discover servers by interrogating server-to-server
databases or based on their own internally specified publishing limits as
defined by their author. Agents
discover each other based on a combination of database-managed queries and on
unique agent interrogation code.
Agents themselves may simply consist of a description of the
resource for sale and its encumbrances or may also have their own logic in
order to be able to evaluate and respond to purchase offers as quickly as
possible. The server itself knows how
to do basic evaluation on simple agent properties, but more complex queries
require the agent itself to do the evaluation based on internal code logic
distributed with the agent.
The entire system runs in a sandbox on a client’s computer,
and on servers distributed over the Internet.
There are boundaries specified so that agents cannot escape into the
broader Internet nor use up more computational resources than permitted. Also there are limits to prevent agents from
denying other agents access or services - each agent is throttled as to
resources and presence.
The sandbox however isn't necessarily designed to prevent
agents from doing real work. If an
agent transaction takes place, and trustworthy authorities reasonably certify
those agents, then the agents can autonomously engage in the transaction itself
- and this sometimes involves doing real work on a client computer or elsewhere
in the computational sea. Examples of
this may be to guarantee to present an advertisement banner to a certain number
of viewers, or to crack a key certificate or to deliver an mp3 file to a target
or to proofread a web page, to find five other potential friends within a five
mile radius who also are bored right now, or to translate a document from one
language to another...or any of an infinity of other types of genuine value
exchange.
Value based exchange and
currency washing,
Fundamentally the point behind a digital currency exchange
medium like this is to lower the barriers to sustaining more complex
transactions. Today we use the simplest
possible transactions and the dollar is largely severed from the value systems
of the parties exchanging those dollars.
The reason this occurs is because values are ultimately held
individually, while work energy is ultimately a pair wise relationship
only. A person alone has no need of
measuring their work energy - they work only according to their values. There is a mismatch between the two concepts
and enforcement of values has to be propagated explicitly.
Consider two parties, Alice and Bob. Alice is opposed to slash and burn cattle
farming. Alice knows that organizations
that are free and clear of this kind of behavior are denoted as such by
reputable authorities, and only hygienic agents will have the proper
certificates attesting to their legitimacy.
On the other hand Bob has a product to sell such as a pair of
shoes. If Bob purchased his leather
from a reputable source then the certificates that Bob has from those agents
will be attest to those facts. If Bob
in turn wishes to trade certificates with Alice, he must use the agents he has
on hand, or use some agents. Alice may
only accept a certain rate of dirty dollars.
If Alice is too strict in her demands then she may devalue her own
ability to make purchases. On the other
hand if Alice is too liberal then she dilutes her personal value systems and
contributes to a cause she doesn't believe in.
As Bob trades certificates with Alice, he acquires a larger percentage
of cleaner money (according to criteria set by Alice). This action washes his currency and he can
then use that in transactions with other parties who perhaps are more strict. Ultimately the market sets currency exchange
rates. Alice and Bob choose what their
values are, and the system automatically ranks the exchange rate of their
currency. A single item of currency may
have dozens of discrete encumbrances, and each may hamper or favor the value of
that currency in various markets.
User Liabilities,
Since smart money can reflect your values, and since it can
in fact do work in some circumstances, as an owner of those currencies you
become liable for the actions that currency takes. If you program a currency to engage in malice, or if that
currency by accident engages in malice you may have some degree of
liability. The best model for this
today is the common corporation. As a
CEO you have a group of employees acting on your behalf. Responsible CEO's setup some kind of
liability protection for themselves and their shareholders, but at the same
time not all liability can be completely avoided.
Overall the market sustains some risk in order to remain
attractive to users. As well the cost
of the market is increased somewhat by paying for the costs of abuse be it
accidental or intentional. This makes
all currencies more expensive however this cost is probably equivalent to the
same kinds of costs seen in current markets.
Interoperability with today's
currency,
Translation to less encumbered currencies such as the U.S.
Dollar can also be done, and in fact should be encouraged. Ultimately the entire system can be pegged
against the U.S. Dollar as a baseline.
Each form of currency can reflect some perturbation off the dollar and
the dollar can be the common ground between all currencies.
Governments tax the exchange of work energy. This serves to slow down the rate of the
economy such that it doesn't crash, burn up or blow out, and to pay for the
basic services and infrastructure support which are deemed necessary for the
benefit of the society as a whole. A
similar mechanism can to be in place here in order for this system to remain
legal, and also to impede the potentially rapid transaction rate - perfectly
liquid markets may be subject to crash and boom cycles much like traditional
ecological or economic systems but at a much faster rate.
Value Licenses,
Rather than users each inventing a unique license, many
users may subscribe to well-known licenses that cover a well-defined range of
interests. These licenses will have
undergone analysis by more eyeballs, and will be known to be more secure
against legal but opportunistic abuse.
Common licensing models will probably evolve out of existing licenses
such as the GNU Public License by Richard Stallman. Another popular license form may be the Berkeley License. Finally there will be a wide number of
certifying authorities such as GreenPeace or the Red Cross who certify certain
work and business practices of various organizations or even Governments.
People are still welcome to dial their own license, and
generally speaking products should provide some meta information about their
nature. It may be common for example to
have local community currencies which favor local community volunteerism, or
certain businesses may want to encourage customer loyalty by offering
currencies which favor their establishment over others. Of course the degree of restriction these
currencies choose to have does encumber their usefulness, but one can imagine
being given money specifically for one purpose, such as 'beer money' to only be
spent on a nights entertainment and not on anything serious.
There will need to be license authoring and editing tools to
examine closely all of the encumbrances associated with dollars being
acquired. There will be a need for
auditing programs to generally analyze uncommon currencies. As well there will need to be network infrastructure
to constantly rate and revalue popular currencies according to the market
demand.
Volatility and Niche
licensing,
Agents may naturally have expiry and may naturally decrease in value and this should be reflected in the valuation those agents have in the market as determined by other agents. One way to do this is to have some automatic trading of currencies - where even less used currencies are being traded. This forces the system to exercise, and may have to be enforced rather than a natural mechanism. The system should try to avoid having undue labor of evaluation placed on third party agents or else this will diminish the diversity of agent license models (the impedance barrier against evaluating niche licenses may be too high).
Exchange
Software services,
The software itself consists of a stock market distributed
over a vast number of servers. Servers
tend to be large databases at centralized locations on the Internet. The client also has a local server. Agents are aggregated and distributed
between servers using a pollination model of random dispersement favoring the
point of issuance. The basic commands
that a server needs to support should include,
Cryptographic product
authenticity determination,
This proposed system can only succeed if it is reasonably resistant to tampering by malicious or playful forces. 1) The authenticity of the products of the buyers and the sellers must be reasonably provable and 2) the moral values impregnated in the currency must be provable.
A liquid currency is backed by the government and by law. It constitutes a guarantee that your dollar can be traded for goods that you want and so on. Users of the system have legal recourse. Consider Alice and Bob simply trading goods by hand. If Bob offers to do some work for Alice, and Alice offers to pay, either Alice has to pay in advance or Bob has to do the work in advance, either way one of the parties take a risk however there are often verbal or written contractual agreements between both parties and the penalties for breaking those agreements are costly enough to make them usually not worthwhile.
In our
society we have legal recourse when we’re shortchanged in a transaction. Bob can take Alice to court. However over the Internet we may not
necessarily even know whom the other party is or in fact if their goods are
truly valid. On the other side of the
coin as individuals we may want to protect our own anonymity as well. The scenario on the Internet is more like some
anonymous party is offering to pay for some services rendered, and you have to
decide if you want to risk that liability.
Alice may disappear from Bob’s ability to discover her after receiving
her part of the transaction. Ideally we
want generalized reasonably robust protections for all transactions, outside of
specialized forums such as barter groups.
Our legal system can be extended to apply online. As well, similar systems with similar penalties can be created. In fact these kinds of penalty mechanisms works fine for online systems such as EBay since the communities are small, based on trust, and generally trade for goods in an industry and on a medium that has traditionally had a high integrity. EBay also has defensive mechanisms which hamper the ability to engage in future trades if one participant defects. Usually in a system of many iterations and where all parties are present over the long term it seems less likely that a given participant is going to be deceptive. However there are exceptions – participants may deliberately grow trust and then cause a massive deception in a short period of time.
Ultimately
this system should use existing infrastructure such as the legal system as a
starting point. This requires some
means of tracking down third parties. It
is probably good enough to simply be able to identify the other party in event
of a breech. Of course the legitimacy
of the currency increases its value and so certified money is more valuable at
the same time legitimacy costs money, and guaranteed or certified currencies
may be naturally more expensive
Parties in a smart money based system could acquire legitimacy if they owned unique unforgable certificates that were commonly recognized. As well all parties could register with a central authority which you could goto for relief and who could pursue them and/or reimburse you. If Alice is indeed untrackable then at least Bob is compensated for his loss – although there might be a penalty to Bob for having engaged in foolish risk.
Ultimately the documents certifying certain services do point back to the advertiser of those services. However the ownership information may be cryptographically secured against viewing by any one of a number of system maintained master keys. The manufacturer of a product or service acquires a ‘factory’ which permits them to sign a fixed number of documents and each of their documents must be serialized to indicate where in the series they are. Otherwise manufacturers could use old certificate factories to flood the market with fake currency.
Anybody can offer any services for sale, but as they fail to deliver on those services their rating goes down, and access to valid certificate factories might indeed be closed by some kind of oversight authority. Of course users can change pseudonyms however presumably new pseudonyms have inefficient value.
Cryptographic certification
of user values,
Third party certification of the intentions or moral values associated with a currency is required. There will be parties who certify that certain organizations have certain business practices. Manufacturers of products obtain a license to print certain kinds of certificates as granted by organizations who go in and inspect their methods, and of course this license can be revoked for later non-compliance. The certification parties may in themselves be subject to mass peer review, and the value of their certification may decrease in value, thus decreasing the value of documents or currencies signed with their identity.
In cases where a group of users collude to deceive a third party it may be necessary to have spot checks on random transactions on the system overall.
A smart money system acts to consolidate and automate a wide variety of transactional mechanisms that exist today. It can shift control of transaction away from a centralized model to a distributed model. It can also help better match buyers and sellers and help individuals better discover niches of opportunity. More than simply being about currency it can also embody labor itself, and may constitute a new way to deliver client applications or application time for single use applications. It acts to insulate the human participants from the day-to-day cost of evaluating transactions. It can globalize labor. It can spark a renewed interest preserving our planets ecological heritage and other such endeavors that are not traditionally weighted heavily in existing business models. It can create more direct reward mechanisms for product vendors to reach their audiences – audiences can be directly rewarded for listening to advertisements or providing information about their interests or needs – and if needs are appropriately met between buyers and sellers then advertising as a medium ceases to exist and simply becomes valued just-in-time information.
A system such as this is not difficult to build or grow and can immediately have an audience. Such systems may emerge over time but if done as commercial ventures they will be fragmented and piecemeal with divisive splits between the vendor-owned databases of products and services. In fact basic transactional mechanisms need to be supported as a freely developed endeavor that benefits all of society. Systems such as this cannot be built by narrow vested interests. Should such systems come to exist they will provide a solid bedding plane for a whole new type of work and behavior that we can’t possibly imagine today.
First Monday
Idea Futures
Agents
Community
currencies and local money
Retailer
discount coupons
Traditional money systems
The GPL Open Source License
The
Berkeley License