When DAO Meets the Homeowners' Committee: How the "Happiness Index" Under the Merkel Tree Reshapes Grassroots Governance?

CN
PANews
Follow
2 hours ago

Recently, prediction markets have become very popular, and I propose a new concept that might support some cool experiments.

This idea is not original to me; it comes from a rather magical paper. The author is one of the "founding fathers" of Crypto, Ralph Merkle, who radically suggested using prediction markets to govern nations. Such a paper was surprisingly published in the journal "Cryonics."

When I first read it, I found the concept interesting, but its practicality seemed zero. Recently, upon rereading, I suddenly understood that if the scenario is not limited to national governance, it is actually a quite operational general solution.

If you don't remember who Merkle is, he is a co-inventor of "asymmetric encryption" (public and private keys) and the inventor of the "Merkle Tree."

Every transaction on the blockchain relies on public and private keys. Each block of Bitcoin is stamped with a "root" of a Merkle Tree (used to efficiently prove that all transactions within the block are complete and unaltered).

Background of the Paper

Merkle first launched a fierce attack: he believes that the "one person, one vote" democratic system is fundamentally a mess. This system forces the majority of ordinary people, who do not understand economics, political science, or sociology and are misled by the media, to vote on extremely complex various bills.

This is both unfair and inevitably leads to mediocre and poor decision-making. The governance machine designed in the paper (Merkle calls it DAO democracy) operates on a logic completely opposite to traditional voting systems.

Traditional voting is "decide first, then see the results" (people vote for A and then endure the good and bad consequences brought by A). Merkle's machine is "predict the results first, then make decisions." The operation of the machine relies on two core components:

1. Unique Goal: Citizens' "Happiness Index"

The machine has a unique and unalterable ultimate goal (protected by DAO contracts), called the "Happiness Index."

This index is obtained by all citizens "scoring afterward." Every year, all citizens score the past year, for example, from 0 (worst) to 1 (best). The average score of everyone becomes the "annual Happiness Index" for that year.

This score is the only metric the system pursues.

2. Decision Engine: Prediction Market

With a unique goal, decision-making becomes simple. When someone submits a new bill (for example, "Should we build a new high-speed rail?"), the system does not open voting but automatically sets up two parallel prediction markets:

Market A: Predict "If the bill passes," what will the long-term "Happiness Index" be?
Market B: Predict "If nothing is done," what will the long-term "Happiness Index" be?

Then, the machine waits for the prediction period to end and looks at the prices of markets A and B.

If the price of market A is higher than that of market B (predicted to be 0.72), the machine automatically determines: pass the bill. Otherwise, it is rejected.

The Ingenuity of the Design

The ingenuity of this design lies in transforming "decision-making" from a "political issue" filled with biases and populism into an "information problem" that rewards rationality and expertise.

In a prediction market, a person making random bets ("I don't care, I just hate high-speed rail!") will lose money. Those who can accurately predict "whether this bill will make most people happier in the future" are the ones who can truly make a profit.

It cleverly utilizes "greed," allowing rational voices rather than the loudest voices to dominate decision-making. Of course, the specific mechanisms are far more complex than I have described; those interested can refer to the paper themselves.

Back to Reality

Personally, I believe that using this machine to govern a country has zero practicality.

Merkle himself mentioned many challenges: for example, how to prevent the system from absurdly choosing "to give everyone hallucinogens" in pursuit of high scores? And how to handle bills that have a "10% chance of causing the apocalypse"?

In addition to these technical challenges, political friction also determines that no political system could apply this solution.

However, if it is not national governance but rather some narrower fields, through appropriate abstraction and carefully crafted conditions, I believe there could be a pathway.

A Simple Example

Decision-making in a community's "owners' committee." The "show-off faction" wants to spend 100,000 yuan to build a completely useless fountain. The "essential needs faction" wants to use this money to repair a leaking roof.

Under traditional voting, this matter ultimately becomes a victory for "the loudest person" rather than "the right person."

Using the "Merkle machine":

Goal: Annual resident satisfaction.
Submit two proposals and let the prediction market set the prices:

Market A: Predict what the average "satisfaction" score will be at the end of the year if the fountain is built.
Market B: Predict what the average "satisfaction" score will be at the end of the year if the roof is repaired.

Those homeowners whose roofs are currently leaking (they are the true "experts" on this issue) only have 1 vote in traditional voting. But in this market, they are 100% certain that repairing the roof will improve satisfaction, so they dare to bet heavily on "Market B." The system sees that the price of "Market B" (predicted satisfaction) is higher than that of "Market A," and automatically approves the proposal to repair the roof.

Settlement: By the end of the year, all homeowners score. Those (now not leaking) homeowners give high scores. Those who bet on repairing the roof win the money from those who bet on building the fountain.

The actual application design may be a bit more complex than this, but the basic logic is as follows.

The essence is to hand over a publicly torn, subjective community decision to a similarly public, prediction machine driven by real money for arbitration. And the democratic one person, one vote has not disappeared; it has merely changed form, allowing the entire mechanism to operate.

This product could even become a "governance-as-a-service" platform. The platform itself does not decide any KPIs or proposals; it only provides a neutral "toolbox" (such as DAO contracts, prediction markets, and Oracles).

Any organization, from owners' committees to open-source communities, can register to use it and then "insert" their unique KPIs (such as "satisfaction" or "download volume") and specific proposals.

The platform is only responsible for running the market and returning the "optimal solution"; it acts as a neutral "referee," providing a plug-and-play decision machine for all organizations that need to make difficult, public decisions.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink