[1/18/2019] How To Enable Democratic Blockchain Governance Through One-Entity-One-Vote

Many chains struggle with plutocracy - a de facto form of centralization. In this piece, we explore how Storecoin could implement a one-entity-one-vote system.

Read the full piece https://research.storecoin.com/how-to-enable-democratic-blockchain-governance-through-one-entity-one-vote

The biggest concern with identity based systems is the sybil attack. If you’re relying on a third party to validate individuals, what prevents me from, say, buying or bribing that entity and having it validate that I am 4,000 separate individuals? If you allow multiple third party identity verifiers, you have to worry about the “weakest link”, the one that might get compromised. A related question, who determines which third parties are allowed to verify identity? Whoever has the power to define the identity validator effectively controls a “1 person 1 vote” network?

Ari,

You’re right. To credibly secure a decentralized democracy, we need to solve for the likeliness of sybil attacks on identity. Our approach here is really unique.

The foundation of a democracy is one-person, one vote. This isn’t possible in computing networks. With p2p, an entity can be represented by anything, shell companies included. To secure against sybil attacks on voting, we’re solving “identity” by not trying to solve identity. We’re solving it by trying to solve for trust in an p2p economic network. We’d love to work with you on this new research that we are calling Fault Tolerant Trust or trust-your-network. TYN for short (“ten”).

On some level, think of TYN as the BFT consensus mechanism for trusting governance itself.

We invite you to review our v0.1 research:
Two-page overview of Fault Tolerant Trust (TYN) https://www.dropbox.com/s/u4z4fjj5iwtygdo/Fault%20Tolerant%20Identity%20(TYN)%20Overview%20%20-%20Storecoin%20Research%20(2).pdf?dl=0

Full v0.1 research paper for TYN: https://www.dropbox.com/s/xn9gaabzp12myde/How%20Fault%20Tolerant%20Identity%20(TYN)%20works%20-%20Storecoin%20Research%20(2).pdf?dl=0

As Chris and Rag say, “highlight why trust-building mechanism is critical to the success of decentralized democracy.” There are several IRL trust-building concepts and one failure I’d like to discuss equivalents within TYN and sybil-resistant networks:

  1. Test votes
  2. Campaign finance reform
  3. Network corruption for illicit purposes

First, test votes are low stakes votes, like quorum calls that prove that sufficient numbers of legislators are present. In mature legislatures, caucuses (i.e. voting rings) serve as a primary governance vehicle within the entire system. Quroum calls can be used to consume time in presidential democracies, but in parliamentary democracies, quorum calls and their equivalents can be used to test the legislative majority short of a vote of no-confidence. Quorum votes would translate as asking “do you still exist” to nodes and miners as part of proving that the network is still active.

Such votes can be used to boost trust in a network to prove a strong governance majority. One level deeper, a decaying majority can be a yellow flag potentially indicating a sybil attack degrading network integrity (or a need for a policy shift). Now, test votes can be weaponized against a governing majority but can also be twisted back around again by a strong legislative leader, both of which are demonstrated through this story of the American House of Representatives in the late 1800s: https://history.house.gov/Historical-Highlights/1851-1900/Speaker-Thomas-Brackett-Reed-of-Maine-proceeded-against-the-“disappearing-quorum”/

Second, one of the TYN failures despite ‘IRL one voice, one vote’ is that of campaign finance. Campaign finance reform tries to decouple donations from voting motivation. By moving to a decentralized platform, we can take a fresh look at campaign finance reform. What would be the equivalent of disclosing your donors? What about justifying your votes? How would we restrict the buying of votes? This is the fundamental set of questions that TYN seeks to answer, so perhaps the long-running debates on campaign finance reform offer some adjacent thinking that we can learn from.

Third, the corruption of non-obvious/private networks to certain ends. Disrupting such corrupt, plutocratic networks through TYN has intrinsic value. Take the cases of the networks Harvey Weinstein and other predators were able to call to their defenses, as described through new reporting by Ronen Farrow. Such shadow networks emerge through shared experiences (e.g. galas and conference retreats), shared boards, and shared funding channels, among other aspects. Good decentralized governance both understands the value of such lack of transparency in enabling action/flow (e.g. the elimination of earmarks has markedly slowed down Congressional throughput) and the risk of it (e.g. corrupt backroom deals that harm the general pulic). These leveraged networks are examples of one voice, many votes that Chris and Rag are looking to prevent.

Finally and as a non-IRL example, non-monetary, user-dominated networks like Quora use onboarding to build a de facto TYN process (with many, many failures) to secure new users understand and comply with rules like ‘be nice, be respectful,’ separate from the role of the community to self-enforce through reporting (and the old role of user moderators). How would we adapt something similar to onboarding for the purpose of trust-building of new miners or nodes within a TYN structure?One thing I wonder if TYN can integrate into its implementation is the concept of test votes. In mature legislatures, caucuses (i.e. voting ring) serve as the primary governance vehicle within the entire system. There are coercive and non-coercive tools to maintain cohesiveness among a governing caucus. Votes like quorum calls can be used to both illegitimately inflated “voted with X” percentages in campaign ads but also serve as test votes to confirm the cohesiveness of the governing coalition.

1 Like

As a small point, Caltech’s dorm-equivalents were also social clubs, and membership expansion was controlled by its members. At a basic level, we found great value in adding a third option between “yes” and “no,” that of “who?” The “who?” vote indicated that they weren’t a justified no but also had insufficient justification to say yes. In terms of voter psychology, this prevented a forcing of uncertainty.

For governance construction, I would recommend consulting with an expert in polling to understand the psychological construct of vote options (e.g. Dan Ariely is the best in the field).

2 Likes

I generally like it. At a high level, it’s a reasonable approach, and it’s nicely framed within the models already trusted and validated in the cryptocurrency world. I don’t have a current opinion on the more subtle specifications; we’ve got quite a while to discuss those points over time.

1 Like

Great thoughts! Apparently posts have to be 20 characters, so adding this filler (-:

One “abuse” of TYN could come from adversaries proposing irrelevant items for voting and thus overwhelming the voters with “too many issues to vote”. An example of such “voter overwhelming” (although not intentionally) is the ballot that happens in November where everything from electing the next president to local tax reforms are tabled for voting. It is impossible for an average voter to “trust” the details of every measure and vote confidently. So, most voters either not vote at all or “guess” who/what might be better based on what they have heard or read. Both result in subpar outcomes. So, a decentralized democracy must somehow ensure that this doesn’t happen. “Test votes” described Craig is an excellent solution, but it can be strengthened further by requiring “proposers” of any measures to “stake” $STORE for tabling any proposals for voting. If sufficient quorum is not reached, the proposers would lose their stakes. This has two benefits.

1/ Discourages “let’s have it on the ballot and see what happens” mindset. It forces some hustling by the proposers of measures to do their homework.

2/ It creates some noise prior to actual voting, so people who care for the proposed measures (for or against) can get involved before the voting actually happens.

This also naturally brings “transparency”. Any pooling or collusion to support or defeat proposals will be “open” because of the staking requirements and “one entity, one vote” nature of TYN. So, even before the formal voting, the “trend” (at least from those who are vocal) will be obvious. Of course, the formal voting can bring opposite results, but the process has more opportunity to remain transparent.

Would be interesting to see what the community feels about minimum staking requirements to table proposals for voting.

Here’s the public link to comment on the TYN research document itself: https://docs.google.com/document/d/1wnW0YLS1G3kGyZldw7QsszE79YlfWGjxbdQUinz2Qx0/edit?usp=sharing

Let us know when you comment as we otherwise won’t get notified by Google. Once you do, we’ll dive back in to the document (which is now open for comment).

Here’s the public link to comment on the TYN research document itself: https://docs.google.com/document/d/1wnW0YLS1G3kGyZldw7QsszE79YlfWGjxbdQUinz2Qx0/edit?usp=sharing

Let us know when you comment. Once you do, we’ll dive back in to the now public research document.

Ideally, we have these conversations in specific areas of the research document itself while high-level breakthroughs and improvements can be shared on this thread.

Yes, Fault Tolerant Trust (TYN) attempts to solve for this.

Think of TYN as a two-tier network. It separates trust from a vote. Once trust on a fact or a set of facts is reached, a vote can be taken using the 2/3 framework. This “division of knowledge” is what makes TYN a realistic consensus mechanism for trust-of-governance for more than just public blockchain systems (p2p economic networks).

This logic compounds on itself so a vote theoretically could be based upon a network of 100s of TYNs from 100s of networks. A computer could read the “trust score” of each TYN to either assist a decision or algorithmically make a decision.

Think of a trust score as the score of a TYN itself. If 76% of the network trusts something, then the trust score is 76%. The implications for structured data (a trust score) building on networked structured data (trust scores of multiple TYNs) to create advanced decision-making models is very fascinating to me.

In summary, the possibilities are quite large with TYN as a mechanism for finding truth on a fact, a mechanism for finding truth on a set of facts before moving to a vote, and a mechanism for reaching consensus on a decision (a formal vote). It could also help form a super-intelligence that we can trust isn’t sybil! :wink: