LXDAO: Gitcoin Founder Discusses Nearly 30 Mechanisms for Coordinating Collective Intelligence

Translation: Yu Xing, LXDAO

Proofreading: Ray

Introduction

In episode 65 of the GreenPill podcast, Griff Green from Giveth and Kevin Owocki, the founder of Gitcoin, discussed nearly 30 different mechanisms for coordinating collective intelligence aimed at achieving better collective resource allocation. This article distills their conversation.

Main Text

The coordination mechanism chart was originally published by Kevin Owocki on Twitter, sourced from Octavian.

When you can achieve better collective intelligence, you can achieve better collective resource allocation. This is an amazing thing when you are trying to fund public affairs and build a regenerative financial system. - Kevin Owocki

Let’s start with a few classic examples of mechanisms:

Tithing ---

Tithing - This is an ancient religious practice where a portion of income is donated to churches, nonprofits, etc. It is simple and easy to understand; however, it lacks feedback, as people do not know what their money is doing. There is no enforcement or checks and balances, and it requires a strong cultural drive to make it work.

Taxes ---

Taxes - This is like a "mandatory" tithing, where you are required to give a portion of your money to the government. The downside of taxes is that power leads to corruption, and centralized authorities decide how funds are used from the top down. The only feedback mechanism is voting, which is a very slow process, but it must be acknowledged that this is currently the best way to fund public goods.

Voting ---

Voting - There are various forms of voting, typically expressed as majority rule, and representative democracy is also a form of majority rule. Another way is consensus voting, which essentially requires absolute agreement to make progress.

Taxes are a funding mechanism, while voting is a distribution mechanism. There are many powerful mechanisms to raise funds for public resources, but the distribution mechanisms seem less effective. Voting can be corrupted by money, so how about the "market" as a distribution mechanism?

What if there was a market for "clean rivers"? This would allow different companies to compete in this new market. The collective could use supply and demand to choose the public goods we want and reward value creators.

Markets ---

Markets - Generally seen as the best way to allocate resources, but not suitable for collective resource allocation. Markets pursue profit maximization and only serve those who can afford the purchase price, applicable only to highly exclusive goods. Market failures are evident in public goods funding; however, markets themselves are a decentralized collective intelligence mechanism, where any good that has consumer demand and is willing to pay will be produced, which simultaneously incentivizes innovation.

"The coolest thing about markets is that it's win-win. Nobody is losing. If you're creating value, you win, which is amazing. The real problem is commons don't have markets to push back." --- Griff Green

The coolest thing about markets is that it's win-win. Nobody is losing. If you create value, you win something, which is amazing. The real problem is that public goods do not have markets for feedback regulation.

Web-based Mechanisms

As we transition from the industrial age to the information age, there is now an opportunity to implement more effective governance mechanisms that can provide funding at a scale similar to markets while also possessing the legitimacy and public interest that taxes and coordination bring, truly democratizing the economy. Whenever people create a high-resolution image of public product demand, collective intelligence will invest funds to realize it.

Here are 27 mechanisms discussed by Griff and Owocki, of course, each with its limitations.

Quadratic Funding ---

Quadratic Funding - Created by Glen Weyl from Microsoft and Vitalik Buterin, the founder of Ethereum, this mechanism raises funds through a central funding pool, distributing funds based on the level of public support for a project rather than the amount of capital contributed. This is in contrast to a 1:1 method, which is dominated by the wealthy and only reflects capital demand. Quadratic funding is based on the number of contributors rather than the amount of funds, reducing the influence of large donors and leaning more towards democracy.

Example: You initiate a fundraiser and raise $100 from 100 donors, while I initiate a fundraiser and raise $100 from just one donor. Although both fundraisers received $100 in donations, you will receive 99% of the matching funds pool, which helps direct funds to "the poor and the many." Even if you only donated one dollar, you are actually part of the public will, guiding the direction of public fund allocation.

Issue: It may be subject to "witch attacks" and requires raising a public funds pool.

Conviction Voting ---

Conviction Voting - This is a mechanism similar to quadratic funding that also mitigates the influence of wealthy politics but requires time. The rule is: you express support for a proposal by staking tokens, which can be withdrawn or reallocated at any time; the weight of the vote increases with the duration of the token stake, and once a proposal reaches a threshold, it passes.

If you own more tokens, you have more say, but even if you only own a small share, you can still advance a large proposal, just requiring more time. This applies to competing proposals of the same level; among all proposals, you can allocate your tokens to your preferred proposals in various ways. This concept comes from Michael Zargham and others and is implemented by 1Hive. It works very well when there are many competing proposals and a clear budget, allowing the community to allocate the desired amount within a given time.

Issue: Requires many proposals to compete; otherwise, it is easy to pass a proposal. The good thing is that minority opinions can actually pass proposals.

Retroactive Public Goods Funding ---

Retroactive Public Goods Funding - Inspired by Karl Floersch, this mechanism has been implemented on Optimism. They create an expert committee and allocate revenue from the Optimism network to this committee, which then distributes funds to things that provide value to the ecosystem.

Benefits: More talents can develop public goods based on expected future revenues, while funders do not have to worry about the difficulty of measuring funding outcomes.

Network Goods ---

Network Goods - Private goods are competitive; the iPhone is an example: you must be able to afford it to own it. Public goods are non-competitive, such as breathing air: everyone can access air regardless of financial status. Network goods become more valuable as more people consume them, just like open-source software. The more attention NFT art receives, the more valuable it becomes. By issuing NFT artworks for public goods, the value of NFT artworks can be increased through the dissemination of public goods, achieving funding goals.

Effective Altruism ---

Effective Altruism - Effective altruism follows the principle of cost-effectiveness, aiming to fund the most efficient and best-performing groups, getting the most results for every dollar spent.

Kevin: Effective altruism is based on the theory that we are not actually donating to make an impact but to feel better about ourselves. Effective altruism, as a philosophical thought, operates differently for everyone. If we categorize public goods from a utilitarian perspective, how do we determine which cause can produce the greatest benefit?

Suppose I donate $5 to buy 5 mosquito nets, which can save 100 people from malaria in a certain area over 5 to 10 years; while the same amount of money funds another project that can only save 2 people. Ranking funded public goods using a similar mechanism and donating our money to the most effective public goods is very interesting.

The limitation of effective altruism is that this model cannot identify projects that may prove effective in the future, so it is suitable for large-scale operations, which can create integration effects.

Griff: The problem with effective altruism is that analyzing impact data is very difficult. Impact is essentially qualitative, but we are trying to quantify it. However, some values are unquantifiable, like how much is a sunset worth? These are all monetary vectors. A more direct ethical dilemma is the trolley problem: how do you quantify the value of a life?

Hypercerts ---

Hypercerts - Owocki is working with Protocol Labs to develop this mechanism. The principle of hypercerts is as follows: suppose I collect ten tons of carbon from the atmosphere and receive a certificate; that certificate may have value in the market because another person might be able to emit ten more tons.

Hypercerts can promote similar impact factor contributions to any impact vector. I saved ten people, picked up ten tons of trash, helped ten elderly people cross the street; any positive impact can be issued a hypercert, and then a market for impact assessors can be initiated to attract those who want to collect hypercerts to showcase their virtues. Web3 provides tangible proof of virtue.

We are launching a three-party market where any DAO that generates positive impact can issue hypercerts, impact assessors can certify them, and interested parties can purchase them.

The overall value is that any DAO generating positive impact can shift from worrying about how much impact it can capture to how much impact it can create, fundamentally changing the incentives for impact DAOs, allowing them to build business models, which means more capital and talent can flow into impact DAOs.

At the same time, this also addresses the issue of effective altruism mentioned above—if you can create a decentralized market regulated by collective intelligence where people can issue, assess, and purchase hypercerts, then you can have a decentralized data source on impact to perceive whether the impact occurring in the future is happening.

Why didn’t they work in the past? Impact markets existed before NFTs and are not new, but they were not often discussed in traditional systems. What’s different now?

A globally transparent, programmable, and tamper-proof ledger is key. For example, carbon emissions are an impact that flows within a global system; previously, people lacked a basis for consensus. At the same time, one thing I learned from carbon emissions is that initiating purchasing pressure will be the most challenging part. Carbon emissions can be seen as a subset of the large-scale application of hypercerts, with regulations corresponding to carbon emission shares, 99% of which come from government regulations. But in a decentralized market, how do we exert this purchasing pressure? Should we have Vitalik tweet about it? Or should we make the number of hypercerts a hard requirement for funding? I think to achieve large-scale application, the most likely thing is to leverage social and cultural forces to make collecting hypercerts a trendy thing. This may be somewhat possible, but before that, we can still establish a market, a donation game.

Augmented Bonding Curve ---

Augmented Bonding Curve - A bonding curve is a smart contract that serves as a means to initiate a market. The approach is to mint corresponding tokens for reserve assets; each time assets are added, it mints tokens, and conversely, it burns tokens to release collateral. The trading object of the tokens is the contract, so there is no need for a buyer when selling. The price of the tokens rises with each minting and falls with burning. By improving the liquidity base, it addresses the price discovery issue during cold starts.

The augmented bonding curve (ABC) adds fees during the minting and burning process, so when assets are sent to the ABC to mint tokens, a portion of the funds goes into a pool managed by token holders, while the rest is used for minting tokens.

If someone destroys it by sending tokens to the ABC, a portion of the released funds will enter the collective. This is a market-based continuous funding flow; as long as there are buy and sell transactions, fluctuations in asset price discovery will occur.

ABC also adds a "Hatch" in its initialization to collect funds before issuing the bonding curve, where a portion of the funds is used to mint tokens for all participants at the same price, and another portion is used for the collective.

Dominant Assurance Contracts ---

Dominant Assurance Contracts - Assurance contracts are a funding commitment based on a precondition, i.e., "I want to fund this project, but only if others do too." If the fundraising does not reach the threshold within x weeks, all funds will be refunded. Kickstarter is famous for this model.

Dominant assurance contracts go further; it is a bet: if the contract reaches its goal, the main funders will get back their funds plus some profit, but if the project does not get funded, their money will be distributed to all minor holders. So this is a game of profit-making, donating, and insight, where both self-interest and altruism have enough incentive to participate, promoting the public to fund public goods in an engaging way. This model originates from Alex Tabarrok's development of the concept.

Donation Mining ---

Donation Mining - Giveth.io is a community-driven organization aimed at funding nonprofits and social causes. GIVbacks is its previous donor reward mechanism. Every two weeks, GIVbacks rewards GIV governance tokens to donors of verified impact projects, allowing donors to build their influence within the organization while funding the development of public goods, and expecting to increase GIV value through network effects for returns. In addition to GIVback, participants can also engage in GIVpower, increasing the exposure of a project awaiting funding by staking GIV tokens and earning mining rewards.

Pairwise (formerly Budget Box) ---

Pairwise (formerly Budget Box) - This tool comes from the Colony ecosystem (2018). Pairwise evaluation refers to two options appearing on the screen, and you choose the one you prefer; in the end, you will get a ranked list based on your preferences. Your ranking can be merged with others' rankings, and you can ultimately get an ordered list, a preferred option, or a weighted list for budget allocation.

Collaborative filtering: It pairs you with other users who have similar preferences. Through large-scale datasets and collective choice preferences, once you have enough user profile data, you can use collaborative filtering to evaluate other choices they might like without going through various complex selections.

Moreover, voting should essentially be signal aggregation. Having a personal server running your personal AI bot that trains itself to vote for you by sending all online content and even personal messages, you only need to occasionally check its availability; how cool would that be?

Futurarchy ---

Futurarchy - Created by Robin Hanson, futurarchy is a mechanism for betting on how the future will impact certain proposals. You declare that a proposal will have an impact, and if it does, you will be rewarded; otherwise, you lose your bet. This is like incorporating the impact of proposals on the economy or other projects you are voting on into your voting choices, introducing voters into prediction markets, allowing the market to filter which policies will create indicators for national welfare. This leverages the power of the market, but the danger is that once an indicator becomes a target, it is no longer a good indicator.

Stigmergy ---

Stigmergy - This is an indirect coordination mechanism, or rather a characteristic shared by all coordination mechanisms, rather than the mechanism itself. Stigmergy is a historical trace of actions that constrains the consensus range, ultimately achieving coordination goals. The principle is that individual actions leave traces in the environment that stimulate the same or different agents to perform subsequent actions. This is how ants coordinate their actions; they leave stimulating traces that guide other ants to achieve their goals.

If you can establish a trust network, telling others who these trustworthy people are, they have high credibility, and the more good deeds they do, the higher their credibility; for example, when people share their donations on Gitcoin, that is a pheromone behavior that encourages others to donate; this mechanism has a self-reinforcing property.

Praise ---

Praise - A bottom-up, peer-to-peer collaborative reward/reputation system, praise maintains a reward fund by giving appreciation to community members. Praise is automatically calculated and recorded, and every two weeks, a group of quantifiers scores the value of each praise. Based on these quantification results, tokens are awarded to the praised individuals, linking each token distribution to each praise. The motivation for praise is gratitude, elevated emotions, positivity, etc., while this rich data can be used to incentivize contributions, inform the community about who is doing what, and so on. People vote for each other, informing everyone who they think is valuable and who is adding value, creating a similar trust network model. Now we can issue tokens on the trust network, which is even cooler.

In Regen web3, we are linking financial capital with other types of capital, such as social and cultural capital, while also making finance serve social and cultural capital. Referencing Gregory Landua and his theory of eight forms of capital.

Demurrage ---

Demurrage - It is somewhat like inflation—the increase in the total amount of tokens reduces the unit value, but the purpose is the same (currency depreciates over time). In demurrage, if you have 1 unit of currency and hold it, over time, you will eventually only have 0.99 units left, with the other 0.01 units being put into a common fund pool. This mechanism aims to incentivize the velocity of currency circulation, mainly referencing the characteristics of perishable assets, such as rice or flour, which decay over time. Therefore, the church would issue a stamp indicating that the rice is decaying, meaning its value diminishes over time. Demurrage originally comes from shipping, meaning "stay" (Old French), referring to the time loss required to unload goods from a ship. This is "liquidation damage compensation." While there are some relatively successful examples of demurrage-style currencies, they have led to similar results as inflation, but the latter seems to provide a better user experience. Demurrage is somewhat like a forced tax.

Proposal Inverter ---

Proposal Inverter - The proposal inverter emerged from research on DAO-to-DAO coordination. Typically, a proposer faces a DAO with a proposal; now the proposal inverter acts as middleware, allowing a proposal to face multiple funders or DAOs simultaneously. If a proposal can solve problems for multiple DAOs, they can co-fund it, and then the funds are collected and distributed according to agreed milestones.

Prop House ---

Prop House - Nouns DAO created this funding mechanism, where the community raises funds in their community pool. During these fundraising processes, builders can propose ideas, and each round of fundraising has a certain amount of funds that can be won by a certain number of proposals. Then, community token holders vote to select the best proposal to receive a certain amount of funding.

Quadratic Voting ---

Quadratic Voting - Quadratic funding is a branch of quadratic voting that can measure not only the preferences of the population but also the intensity of those preferences, serving as an alternative to the one-person-one-vote system. You will receive a certain number of voting credits and can vote on any number of issues, adding weight to each proposal based on the square root of the voting credits you bet on that issue. Based on the number of voting credits someone allocates to a specific issue, you can see not only the preferences of voters but also the intensity of their preferences on that issue. This avoids the outcome of wealthy domination, where louder voices influence judgments on what voters prefer; instead, the broad consensus of the crowd sways the issue. It also encourages voting on multiple proposals rather than putting all voting credits on one proposal.

JokeRace ---

Created by JokeDAO, this mechanism is a bottom-up on-chain governance method. You can tell a joke in the JokeDAO channel, and if it's funny enough, you can receive funding that can be used for anything, including planning community roadmaps, generating ideas, supporting reward tasks, etc. There are different forms of voting to determine what is funny, including 1 token: 1 vote, time-decaying voting, and quadratic voting. The most fun part is that they hold a distributed decentralized joke competition every week.

Aside

$25 Trillion Opportunity ---

$25 trillion opportunity - The public goods sector is serious business! If we can improve the world better than the government, we can decide how to allocate the funds that are currently wasted by the government in providing value to society, with over $25 trillion spent on public goods each year. If we can create a more efficient system and capture a small portion of the value created, entrepreneurs can enter this field and innovate. We might even be able to transform private goods into public goods: food, water, shelter, and other basic needs—bringing them into the realm of public goods.

"We can invent higher resolution mechanisms for democracy. Regen web3 is a blue ocean of opportunity." --- Griff Green

We can create higher resolution mechanisms for democracy. Regen web3 is a blue ocean of opportunity.

Holographic Consensus ---

Holographic Consensus - Proposed by DAOStack, holographic consensus is a hybrid of futurarchy and conventional DAO voting. Anyone can propose an idea (very easily), and token holders can "bet" (vote) on that proposal. When they do so, they "enhance" the proposal to gain more attention and even expedite the voting process. If the proposal passes, those who bet will win tokens (Griff calls it "money"), while those who vote against will lose their stakes.

This considers the attention economy; not everyone has time to focus on all proposals and votes, but when someone you know votes for a proposal, it serves as a form of recognition for you, and you may follow up on it. When there are many proposals and too many things to pay attention to, this method can work very well.

Holographic consensus helps maintain scalability while also being resilient, which are two things that are hard to achieve simultaneously. In this way, those who bet play the role of market prediction.

Skeuomorphism ---

Skeuomorphism - This is a classification of mechanisms rather than the mechanisms themselves. Skeuomorphism divides coordination mechanisms into skeuomorphic and non-skeuomorphic. An example is the information retrieval methods chosen by Google and Yahoo. Yahoo is skeuomorphic because they adopted the traditional library card catalog system model, allowing you to search for content of interest in a parent-child topic manner, just like finding books in a physical library along classifications. Google is non-skeuomorphic, providing information retrieval results through a simple search box, which was unprecedented on the internet.

"The earliest ideas for blockchain ledger technology are going to be just to port over existing methods and ideas. The best, most interesting mechanisms are going to be non-skeuomorphic." --- Kevin Owocki

The earliest ideas for blockchain ledger technology will simply port over existing methods and ideas. The best, most interesting mechanisms will be non-skeuomorphic.

Sortition ---

Sortition - Rarely used in real life, jury duty is an exception. Sortition involves selecting a few people from a subset of potential voters to make decisions. It is somewhat like delegation but randomly selecting from a larger crowd of "ordinary people." Why is it not common in blockchain? No conclusion yet!

Aqueduct ---

Aqueduct - A channel in an irrigation system used to guide water from one system to another. If we are to establish a diversified public goods funding system, what would it look like to build bridges between different public goods funding mechanisms? Sending an aqueduct from your DAO to another DAO, connecting modules of public goods funding. An aqueduct that can connect any other mechanism in the ecosystem. Just like Radicle Drips - you can program a percentage of your income here and there based on your governance. This is a cyclical token flow. Set it up and forget it.

Owocki's example is Gitcoin Grant 2's new "aqueduct": a project will allocate a percentage of its DAO-governed funds to stream to the Gitcoin Grant's Matching Pool, which will run a quadratic funding round based on the project's vesting schedule.

Owocki's example is Gitcoin Grant 2's new "aqueduct": a project will allocate a percentage of its DAO-managed funds to stream to the Gitcoin Grant's matching pool, which will conduct a quadratic funding round based on the project's vesting schedule.

The high-level primitive here is simply "money (token) streaming."

Ranked Choice Voting ---

Ranked Choice Voting - Instead of choosing one representative to vote for, you can submit an ordered list (first choice, second choice, etc.). In practice, the ballots are counted in a way that allows seemingly minority candidates to have a chance and allows voters to express their true preferences rather than strategically voting for the "lesser of two evils." It eliminates the problem of third-party candidates ruining ballots. This is particularly useful for winner-takes-all voting with 4-8 candidates.

Staking/Slashing ---

Staking/Slashing - Staking on Ethereum means depositing 32 ETH to activate validator software and follow protocol consensus rules in exchange for earning a small amount of issuance interest. If you provide false information to the protocol or essentially do anything that could lead to a consensus fork, your stake may be slashed, meaning you will lose your deposit. If you are offline for a day or two, you may also be slashed a small portion for inactivity. This system creates a crypto-economic incentive because funds are at risk and can be staked, slashed, or increased through the protocol at any time. This is a way to bind capital to protocol behavior.

Downside: It requires capital and is to some extent subject to wealthy domination.

Proof-of-Work ---

Proof-of-Work - Proof-of-work is the predecessor of proof-of-stake and still operates in Bitcoin (unfortunately generating a lot of carbon emissions). In fact, it is used in many projects to allocate resources outside of the high-energy Bitcoin network. This resource is issuance ("printing money"). Griff gave several examples, such as CureCoin and FoldingCoin, which reward issuance (tokens) through useful proof-of-work (folding proteins) to find cures for cancer, Alzheimer's, and other diseases. Those who fold more proteins will receive more issuance.

Decentralized Identity ---

Decentralized Identity - We do not want oligarchic digital identities, privacy being crushed, or identity leaks; we want sovereign digital identities (assuming they are protected). An important opportunity is that we can build systems that do not treat participants as interchangeable. We can build mechanisms that adopt a one-person-one-vote approach, which is essentially more democratic, or a one-token-one-vote approach, or something in between (like quadratic voting).

To build a more regenerative crypto-economic internet, we need to be able to engage in positive-sum games and repeated interactions with each other—if you do something to help me, you earn a certificate, and vice versa. This will help us establish trust based on verifiable actions, which is very interesting. By continuously accumulating positive-sum reputation, we will open up the complexity of the space for combinatorial design of all the coordination mechanisms mentioned in this episode to accommodate a massive number of subtle positive behavior markers. Over time, the system will start to generate gravity, creating an economic gravity well that attracts more people to participate. This cycles and repeats until we have a regenerative crypto-economic internet.

Web3 Social ---

Decentralized social media applications are great because they have sovereignty, privacy, and the ability to transfer our social graphs from one site to another. Network effects need to be established, which takes time. Once we shift our attention away from Web2 social, owning our social graphs and being able to fork interfaces, and reflecting the commonality of our influence circles on the social graph, this will position you as one of the centers of a multi-centered collective coordination mechanism, which is also one of the futures of group coordination mechanisms.