LiquidityOfTrust is an idea for a possible solution to CommunityMayNotScale.

In a healthy society, trust should be at least as liquid as money. That is to say, trust (and distrust) should pass from mind to mind with at least the same ease as the power embodied in currency is transferred from hand to hand. The reason is can be summarized easily: SoftSecurity is preferable to HardSecurity.

• Money is a HardSecurity system, and its power derives from its liquidity -- the ease with which it is transferable from person to person, community to community, or society to society.

• In order to keep the power of money in check, it seems necessary to have an equally-powerful SoftSecurity system. The solution I'm imagining is to somehow make IteratedPrisonersDilemma work on the scale of a society. For this to be possible, a defection must be readily communicable from the recipient of the defect to the others with whom the would-be defector wishes to engage.

The games in real life are of far more complexity and subtlety than the archetypical prisoner's dilemna, as they involve two complicating factors: rules of ever-increasing complexity, and the presence of secret information. The game of bridge is the most interesting (i.e., complex and subtle) game I know that is labeled as such, but the evidence seems clear that there have been presidents who have treated America's political system as a game to be paid for personal profit; lawyers who treat the legal system as a game to be played for the gain of the class of lawyers; church leaders who treat the moral code as a game to be paid for their personal gratification; and so forth.

When trust is broken, what often happens is that one of the observers of the exchange suspects (with some level of certainty) that trust has been misplaced, but is often unaware of exactly in what way it was broken, and is rarely in possession of enough concrete evidence to lodge a formal complaint. Unable to speak openly, these observers engage in clandestine gossip, eventually leading to the cheater being well-known as such by the inside crowd long before the details of the scam have been discovered. Icesave was one example: the word had gone through the financial community that something was seriously amiss in Iceland, and had been acted upon (in the form of betting against Iceland's currency), a couple of years before the debacle was publicly exposed. For these reasons, I think it's essential that the atomic unit in the formal LiquidityOfTrust system be the "hunch".

There are many kinds of trust: trust that you will be kind to me; trust that you won't molest my child; trust that you will be an honest custodian of my money; trust that you will be intellectually honest; the list is endless. A common mistake is to confuse different kinds of trust. Con men know how to play on this confusion -- they depend on the "fuzziness" of trust, on the assumption that "people who act personable and friendly" are "trustworthy with money". The street con is only the most obvious form of this fraudulent activity -- fraudulent bankers, for instance, can profit from the assumption that institutions which project the appearance of wealth (and which have figured out how to avoid regulatory scrutiny) can be trusted with your money. To avoid abuses of the "fuzziness" of trust, I think it's essential that we treat trust as a multi-dimensional space, and that anyone be able to invent a new dimension.

For LiquidityOfTrust to work, there are a number of obstacles which must be overcome, mostly psychological and social. Two common patterns stand out:

• The common implicit ultimatum that "I/we will only associate with you if you won't expose my/our transgressions".
• The less-common but more damaging ultimatum that "I/we will punish you if you expose my/our transgressions".

From a technical standpoint, the basic idea is to create an accounting system up to the task of recording and communicating the information needed for people to make effective decisions about whom they can trust, and in what ways they can trust them.

Some related ideas are described on WebOfTrust, TrustMetric, WebOfTrustModeration, FunctionalAccessTrustMetric, and there's also an early external [exposition].

Helmut, I share with Nathaniel the intuition, that trust can really be characterized as a kind of fluid energy [GoogleSideWiki:MayBeThatTrustFlowsIntoSomeInvisiblePrivatePlace] that cannot be destructed, but only flows to other places, perhaps regressing into the private sphere. -- FridemarPache