Video - Bitcoin Nodes vs. Noses - An Axiom For Discussing Bitcoins Scaling And Centralization Problem

The block-size debate is a proxy war between centralization and anonymity. This needs to be solved before we end up with something worse than MasterCard. The good news (or great news) is that adding identity to mining can allow Bitcoin to scale almost infinitely, immediately.


Hello, I want to talk about a, I guess, a new discovery a new axiom that might help clarify Bitcoin's centralization problem. As most of us realized block size debate it's really a centralization debate so it's not that anyone wants smaller blocks. It's just that lager blocks create more centralization of these very dangerous type of centralization which is the Bitcoin hardware miners. So, mining pools, coin based all that are all types of centralization but they are not nearly so dangerous because we can elect to unsubscribe from mining pools and coin base. We can take our services away from them but folks who control an owned mining hardware are opt in. They choose to contribute to the system and if they decide to collude or even if one entity buys up all of the mining hardware then you have centralization. So, this is really the most dangerous type of centralization. And as I'll suggest that Gregory Maxwell's solution which proposed just a few days ago which is oh, well if they centralized too much will just change the hashing algorithm on them, it's very dangerous -- dangerous proposition especially when we look at the difficulties with the Bitcoin governance.

So, let's rewind quickly and just try and get through this idea. So, as we know Bitcoin or Satoshi's dream is really this idea that we have decentralized consensus. And I put decentralized here in quotes because we have no way of actually proving or showing decentralization of those entities that control consensus. So, the mining hardware the miners themselves which are at the center of the consensus cannot be shown to be decentralized. But it is this decentralized consensus that gives us censorship resistance. We get censorship resistance because we are not a Mastercard, Visa or something like that which has one central entity that can kick you off the network which can control your accounts and do whatever they choose. So, Bitcoins novelty is the fact that it is a decentralized consensus. And this in turn gives us this very important idea of censorship resistance if we sacrifice censorship resistance in any way we no longer have a currency.

Okay, if I send you 5 Bitcoins at some date and some entity can go and change that transaction later. Well, you don’t have a currency at all so all value in Bitcoin disintegrates to 0. But a lot of organizations including myself are really looking at the censorship resistance in terms of creating this distributed public ledger which allows tons of novel features. So, I've built my snowcaps, climate change proposal on this idea that you have a censorship resistant ledger upon which you can create smart contracts that are reliable and believed by everyone.

So, the core of this problem and really is this axiom can be shown just in a simple diagram right here. And it basically says there's really tragic and terrible thing for those of us who are big fans of decentralized consensus. And what this new axiom suggest is that you can either ensure absolute anonymity or you can ensure decentralization. But you cannot provably have both. So, the greater decentralization that you have the less likely you'll be able to pull off anonymity and the greater anonymity in this particular case of miners the less likely you are to have decentralization and you won't be able to prove decentralization at all.

So, the decentralization that Bitcoin requires is not a number of nodes because those nodes can all be controlled by one group or one person. The decentralization that Bitcoin requires with censorship resistance is the number of noses. The number of humans that are actually mining and invested in this idea of mining. So, we made a very tragic mistake by counting nodes which can be controlled by one entity and not noses. And a lot of people are going to get really upset with this but it's just a tragic -- tragic reality. The beauty of this though is it does not affect the anonymity of transactions themselves.

So, what we're talking about is the anonymity of miners. That is the fundamental mistake and indeed a mistake that's Satoshi made. This might be the fundamental mistake that he made when he's building Bitcoin. He assumed that decentralization of hardware would entail decentralization. And as we're starting to realize that's just not true. But there's some good news here. So, if we were able to add a little bit of identity so if we inject mining with a little bit of identity we could end up with some really fabulous solutions for scalability. So, I just like to use this hypothetical. What if one individual say some famous individual in Bitcoin say for example Andreas Antonopoulos mind one-tenth of all the transactions. Okay all the rest were done by proof of work and then we just had this proof of reputation mining one-tenth of the transactions. The beauty is that individual be it Andreas or not would be able to mine millions or billions of transactions very very quickly because they don't have to rely on proof of work.

So, we kept the regular proof of work chugging along and then after nine blocks this individual who's also contributing mining could pull together all the transactions perhaps even those without any mining reward whatsoever all the leftover transactions stuff them into a block and mine them very very quickly. Clearly this brings up some real problems with voting and electing representatives. But we might not have a choice. If this axiom is correct, the only way to establish and ensure decentralization is to inject some identity into the system. And again, if we sacrifice decentralization. We are likely going to sacrifice this idea of a decentralized consensus which gives a censorship resistance which gives us a currency.

Okay. So, this is the most fundamental property of Bitcoin. This is not, okay, the anonymity of miners is great, right. Say all the miners were in one house working together and they were all folks you could trust. It doesn't stop a big centralized organization coming down and trying to kill them or influence them or intimidate them. So, the anonymity of miners is great because it protects them from intimidation. So, it's great to have a dose of proof of work in the system. But until you establish some identity we won't be able to establish true decentralization. And just think about say Andreas was mining one-tenth of the blocks. Well, it's clear we've already dramatically decentralized the incentive structure from mining in general. Most miners have a very specific agenda. And usually just to make as much money as possible. Prints as many Bitcoins as possible. All of a sudden driving on Andreas who believes in Bitcoins for a much bigger purpose, much bigger reason and he's kind of done his proof of work. We know that he really cares. So, it's a proof reputation similar to the house of representatives Andreas would be protecting Bitcoin because he has different incentives from the miners in general.

And clearly could take this a bit further. One out of every four blocks could be mined by one particular representative. And I would suggest it's good to have something like four hundred to a thousand different representatives in different geographic locations. Very very tough to do but it might be the only way to remove the quotes here off of the word decentralized. Because right now someone tells you were using a decentralized consensus you need to turn to them and go where are your facts. You cannot prove decentralization. You cannot even show it all of mining right now could be controlled by one or two various entities. We don't know that we can't stop it and we've got no actions to prevent this from happening.

Okay so, Gregory Maxwell the other day spoke about how well if they centralize because right now people assume that more than 50% of the mining hardware is in China. If China decides and they can decide in one minute to go down a control of the mining hardware and run 51% attack on Bitcoin. Bitcoins destroyed by the government of China by the miners of China. Okay and so, what Gregory Maxwell suggested just immediately switch away from SHA256. Well, I've got a funny feeling that just won't go nearly as well as planned and be a real nightmare. But inside of what he said I think there's something fairly interesting, not quite sure how to pull it off. What if Bitcoin instead of relying on SHA256 had every block mined by a different algorithm. New algorithm SHA257, SHA333, SHA3, MDA512, who knows.

But if every block was mined by some new and random hashing algorithm and there were enough of them it would be very very hard for someone to build hardware that would be able to centralize that. We would return to the requirement for having a general processing unit. Doing most of the mining and then all of a sudden allows everyone with a computer to contribute and compete for those coins again like we were doing in 2013 when almost everyone I knew was mining. And now today, as we know almost everyone I know is not mining. In fact, I don't know anyone is mine. 2013, my nephew mined. He was 11 years old. Now, there are no 11-year-old miners. Everybody's professional industrialized miner and this is very very dangerous stuff. So, we really need to consider this new axiom and start to look at how we're going to address this problem.

Written by James DeAngelo on November 15, 2015.