Overcoming Corruption

Blockchain technology allows us, for the first time in history, to coordinate communication without a centralized authority in a secure and redundant fashion, to keep track of every important contribution, and to store those records eternally for future review. This review gives us the chance to analyze the consequences of every action and fairly reward the positive contributions to the group and punish the negative. One of the major efficiencies of a decentralized organization is that all rewards can be shared with the people who contribute to the success of the group, fairly, according to their merit.

We broadly define corruption in an organization as any action that benefits a minority at the greater expense of the group. Corruption arises when power is distributed inappropriately as measured by efficiency. Corruption arises in the dark, when people or institutions have the power to act without review. This is also the case in anonymous decentralized systems. Yet, blockchain technology also provides the opportunity to review actions and the power to judge their value, in order to inhibit corruption, distributing power efficiently.

Just because we can cite every action from a company on a blockchain, however, does not mean we have conquered corruption. More information does not ensure more productive collaboration. To discern the signal through the noise, DAO members must be properly motivated to behave correctly and to police corrupt behavior when it happens.

The immediate goal of obtaining perfectly fungible currency is not the proper incentive for conducting good business or good government. The cartoon scenario of a DAO consisting of anonymous members who stake their currency on rigid smart contracts gives the worst imaginable incentives for collaboration. If the entire proximal motivation of a transaction is currency, all participants will naturally behave in the most selfish manner possible, exploiting any opportunity for individual profit at the expense of the group.

Further, rigid “code is law” smart contracts involving fungible currency exchanges are designed to guarantee irreversible unreviewable programmed self-execution. Very rarely does any business venture proceed exactly as imagined from the beginning. In fact, according to New Institutional Economics (NIE) incomplete contract theory, contracts are seen as a form of experimentation by way of observation for the learning process in the incomplete contract model. To enter into any business arrangement, parties need to have confidence that fair resolution will occur when transactions do not go according to plan—whether to continue business with good-faith actors after an unforeseen event, or to resolve disputes in the case business breaks down completely.

Moreover, rigid “code is law” smart contracts will need to be extremely complex to account for the many possible eventualities of practical business situations. The insurmountable reality is that bugs or hacks can never be certainly precluded in any programmable contract. Beyond this, long term attacks will always be possible for any complicated and rigid set of rules. If the entire reward is short-term currency profit, all parties are naturally motivated to push the contract as far as possible in their favor, and they will certainly exploit any weakness in the contract design to their advantage.

The motivation for all parties to behave as selfishly as possible creates instability for any business transaction. More harmonious collaboration is motivated when improved reputation is the proximal goal. The situation is no longer a zero-sum game when either or both parties can create valuable long-term reputation from the transaction.

Secure and valuable reputation

The main problem with all current blockchain governance structures, especially on-chain governance structures, is the lack of proper incentivization. In business or in government all participants need to be motivated to improve the whole organization over the long term. Otherwise, when short-term fungible rewards are available, the natural and healthy response is to always attempt to game the system for maximal personal profit and to ignore how your actions may damage the group.

The proper incentive, in business and government, is reputation. An actor’s ​reputation ​is defined to be an evaluation of the history of their actions, as a predictor of their future behavior. If valid and accurate reputation can be established on a stable platform with a clear history, that reputation will be valuable for attracting future business.

A platform which vests users with valid and valuable reputation gives us the opportunity to effectively filter the voluminous information contemporary technology produces. People with higher reputation will have greater voice.

Any such reputation system must address the many problems, such as Sybil attacks and tyranny of the majority, that have undermined the meaning and value of digital reputation on all previous platforms. Further, it needs to be a dynamic and evolutionary system, able to anticipate and react to gaming that will inevitably result from whatever rules are in place.

The Semada platform is a decentralized autonomous DApp which vests users in verified reputation, which is designed to be secure against all the known attacks on digital reputation—such as Sybil attacks and tyranny of the majority problems—that undermine centralized platforms like Amazon or Alibaba.

The Semada platform vests users with ​domain-specific ​reputation tokens, meaning that for different expertises, there are separate reputation tokens. The reputation tokens are called sem​, and there are many different types, each of which is only powerful within the DAO containing the experts with a skill specific to that token. So there may be different sem tokens for developing smart contracts, or for advertising products, or for making governance decisions in each DAO. Each user will likely own several different types of tokens related to their individual expertises.

Each sem token gives its owner the power to 1) solicit work for earning more sem tokens (by posting a smart contract which indicates the owner’s availability to do off-chain work), 2) earn more sem by policing other members actions by voting during a validation pool which evaluates off-platform work, 3) post comments, including smart contract and protocol development suggestions, and 4) earn reputation-weighted salaries derived from the currency the DAO earns. Each of these powers focuses users on the long-term goal of improving the organization, instead of the proximal goal of gaining more perfectly fungible currency.

To give the tokens foundational value, the first requirement is that sem token creation must only be connected with actions which bring fungible currency fees to the platform. So each sem token is minted only when fees are sent to the core. (These fees are then distributed to the membership in sem-weighted salaries.) Secondly, sem token distribution must have a clear protocol which is enforced by the weighted democratic membership. Thirdly, sem token valuation must be reviewable, which is achieved by references from future posts which create sem tokens.

Decisions made by weighted reputational power protect the tokens’ value from Sybil attacks. A Sybil attack​ occurs when a single user creates multiple pseudonymous accounts in order to gain power. As soon as it costs less to obtain tokens with multiple accounts than it does with one account, the market will react and Sybil attacks will naturally and inevitably occur. Therefore, all democratic power must be weighted by sem holdings. If every decision is made by a sem-weighted democratic vote, then 1 account with 100 sem tokens has exactly the same power as 100 accounts with 1 sem token, and Sybil attacks are completely inhibited.

Decisions made by weighted reputational power also protect the tokens’ value from​ tyranny of the majority​. Under a one-person-one-vote system, half of the members have less than average expertise, but equal power. This incentivizes experts to gain power by catering to less expert prejudices instead of following the most effective decisions. Under reputation-weighted voting, the incentive changes to follow the majority of expertise, which changes weight based on the success of actions.

Decisions made by weighted reputational power also protect the tokens’ value from various tragedy of the commons26 problems. Tragedy of the commons occurs when members are not properly incentivized to police the evolution of the organization. For example, if complicated technical smart contract improvements are put to a vote by a large organization, very few people will have the ability or interest to participate in the debate. When they vote, they will not have the expertise to make a sound judgment and so experts are again incentivized to manipulate non-experts with sophistry. Therefore, many different types of sem tokens are needed for the many types of expertise. Then, the validation pools which oversee every action in the DAO incentivize experts to participate or else they risk losing the opportunity to gain more sem tokens and maintain or increase their relative power. Sharing the newly minted tokens with those who police the action inhibits this free-rider problem27 of non-participation.

***

Once reputation is the focus instead of more fungible currency rewards, damaging arbitrage opportunities are minimized. The next major concern for creating a healthy governance process is to decide how to reach consensus on protocols for behavior and how to enforce those protocols.