3 ms·
The fundamental problem that needs to be solved is voting power allocation. You have a group of people that wants to use a cryptocurrency and you want to allow
by danbruc 5y ago
The fundamental problem that needs to be solved is voting power allocation. You have a group of people that wants to use a cryptocurrency and you want to allow everyone to voice his opinion about what transaction are valid and should be recorded and which are not so that you can determine the consensus.
Your first problem is to decide who gets to vote. Does everyone get to vote or do you have to participate in the system, e.g. own at least some coins in that system? Do only humans get to vote or do legal entities like banks and companies get to vote? Does everyone get the same vote or does the number of votes vary, e.g. based on the amount of coins you own?
The second problem is the hard one, you want this system to be anonymous which means you can not have a directory of voters but you have to prevent voters from casting more votes than they are supposed to have. That means you have to find some proxy for voters that is ideally impossible but in practice at least hard do duplicate. On the other hand this proxy should be easy and cheap to obtain so that every voter can make use of his vote.
For the example of proof of work the assumption is that every human can own and dedicate about the same amount of computing power. It is not too easy to duplicate because you have to spend money in order to buy more computing power and in turn more votes which is better than giving one vote to each username or email address which are much cheaper to duplicate.
But it turned out that it does not work all to well, people can own and dedicate vastly different amounts of computing power after all so that now only a tiny fraction of all Bitcoin users make use of their vote. Everyone else now trusts the miners on the basis of aligned incentives, i.e. malice miners would destroy the trust and usefulness of Bitcoin, driving its value down, and with that undermine their mining revenue.
Personally I am quite skeptical that one can find something that works well. It has to essentially proof that you are a specific human without revealing your identity and still be unforgeable. Proxys like computing power, storage space, or money are generally duplicable by investing more money, so that it seems hard to come up with a system where voting power is not concentrated on the wealthy.
This is not a fundamental problem but there are certainly people that would consider this to undermine the promise of cryptocurrencies to a certain extend because the votes and therefore the power is not spread evenly among the users and so this requires some amount of trust.
- SheinhardtWigCo 5y agoAlgorand solves this by using a verifiable random function, with the output weighted according to the user's stake. It's quite elegant. Paper: https://eprint.iacr.org/2018/377 https://eprint.iacr.org/2018/377
- jl2718 5y agoWhich is ... proof of stake. Literally vote allocation by wealth.
- SheinhardtWigCo 5y agoIsn't that desirable for a cryptocurrency?
- danbruc 5y agoAs long as you want the wealthy to make the decisions while getting more wealthy in the process, sure.
- Kbelicius 5y agoIsn't that also what proof of work is? I mean, it is far easier for me, a non wealthy person, to stake some pos coin than to mine some bitcoin.
- jl2718 5y ago> For the example of proof of work the assumption is that every human can own and dedicate about the same amount of computing power. I'm not aware of anybody proposing that bitcoin would be a democratic protocol. There were other proposals that were intended to be democratic, but this was explicitly rejected for bitcoin. You might say that bitcoin was intended to be computationally secure against democracy itself. > Everyone else now trusts the miners on the basis of aligned incentives Not really. It's on the basis of cost. Let's say you want to reverse a transaction. First you have to have >50% of the network hash power, which means you need to procure >100% of the current hash rate. That would take at least 2 years and $10B. But let's say you bought 300% of the network hash rate, so you own 75% of it. Great. Now you're two years behind, and at 75% it will take you 4 years for your fork to catch up, so 6 years total. Congratulations, now you've caught up and have the longest chain. Oops, nobody cares; only the main chain coins can be traded for coins, cash, goods, or services. You just spent a huge amount of money to prove that there is no such thing as a 51% attack on a UTXO blockchain. BTW, not so for a stateful blockchain like ETH/ETC. 51% can take control immediately. The difference is that verification on a stateful blockchain doesn't require proof of origin, just consistency with state. A 51% attacker can easily manipulate the state if they maintain control long enough, just by broadcasting their chosen state to any incumbents. Sure, they could do full verification, but, nobody does that. Most full nodes take a while just to sync from the snapshots, never mind genesis.