5 ms·

The author doesn't understand the primary goal. It's precisely to have global consensus that proceeds according to its logic - so if you own something it can't

The author doesn't understand the primary goal. It's precisely to have global consensus that proceeds according to its logic - so if you own something it can't be taken away from you. Everyone running their own private blockchains isn't decentralization, it's federalization. In the internet context, we know exactly how it would look like - initially an ecosystem of independent forums, each a fiefdom of the moderating caste, or subreddits on an alternative reddit that doesn't have global administration. Network effects would inevitably concentrate most activity to few hubs (which is how we ended with facebook, reddit, and twitter consuming the old internet of phpbb forums). Except instead of posts, now the new powers-that-be would have absolute control over people's wealth. What he in effect, unintentionally, proposes, is to give the equivalent of facebook absolute control over people's wealth. In fact facebook itself tried exactly that with Libra - but fortunately got shut down by governments. Honestly, I would prefer a state-run blockchain over the inevitable final form of his proposal. At least it would be regulated by actual laws instead of T&C that give the company near absolute power. There were crypto experiments of this type, most notably EOS. They had their own 'court' that ended up blacklisting dozens of addresses based on weak claims of theft. Eventually all DPoS networks are going to decay to something like that - because with just few publicly known validators escaping legal liability for not enforcing confiscation and freeze orders from real world courts is impossible. The scalability argument is true in isolation - everyone agrees it's currently a problem, but has nothing to do with centralization or not.

ojkelly 5y ago
> The author doesn't understand the primary goal. It's precisely to have global consensus that proceeds according to its logic - so if you own something it can't be taken away from you. How is global consensus not centralisation though? Humans are notorious for disagreeing on everything. Having one global arbiter of truth/consensus seems implausible at best. Anything that could provide global consensus will need to be backed by something formed from a human brain. And our brains are not perfect. We keep updating code because we miss things. Who would maintain the system for global consensus? And how would they not just be our new rulers? As a perfect example you mention ownership as a feature of global consensus. What do you own with an NFT? Who enforces that ownership? What stops someone pirating an NFT image? Ultimately this screams of solving a people problem with a technical one. That never works. Not without strong co-operation of the people involved, and even then technology plays a very small role in the solution. > The scalability argument is true in isolation - everyone agrees it's currently a problem, but has nothing to do with centralization or not. It does in that a truly decentralised system inherently cannot suffer from the scaling problems cryptocurrencies have.
- nootropicat 5y ago
  >How is global consensus not centralisation though? In the same way reality itself is 'centralized'. Global consensus is meant to add a digital layer to reality. Reality is objective by definition. What's subjective is trying to measure it - but this problem doesn't exist for digital systems. >Who would maintain the system for global consensus? The practical implementation has to ensure that (1) consensus is maintained by a diverse set of potentially anonymous participants. Ideally, this would mean literally all humans, but that's of course impossible. The set should be large and diverse enough so that even powerful actors (especially governments) have no ability to force them to do something. DPoS, PoA fail here from the start. PoW is vulnerable over medium term: it has infinite economies of scale which leads to centralization and taking it over externally is always possible (because 'stake' in PoW is external and potentially infinite - mining hardware). (2) the most profitable course of action is for every validator to follow the protocol honestly. In ethereum's PoS that's achieved by slashing maliciously misbehaving validators. PoW has an orders of magnitude weaker punishment here (just mining rewards for the time spent on mining a minority chain). Most DPoS systems rely on the assumption that a majority is always honest, even if it makes financial sense to not be - eg. Cardano's Ouroboros. >Who would maintain the system for global consensus? And how would they not just be our new rulers? In the early period updates are necessary as system evolves - but eventually it has to be frozen, with the possible exception of scaling parameters (like block size, number of shards) which could be determined by a vote from validators. >What do you own with an NFT? Who enforces that ownership? There's no difference to real life here. "Owning" a house actually means there exists a publicly verifiable promise from an entity capable of wielding some form pf violence to enforce that property right. Therefore, an NFT that's connected to any external asset can only a different form of such a record. In the case of image NFTs, it's an improvement on physical art. The point of buying physical art is to either signal wealth, or to use it for money laundering. Image NFTs are infinitely superior for the first - because they are globally visible (as opposed to a physical piece of art) and fake NFTs are impossible (so you can't buy a cheap replica and pretend you spent a fortune on the original). They are much better for money laundering because they are inherently global (no transport issues) and evade identity checks in the banking system. Exactly like physical art, if you just like it and want to look at it - you can save a copy. A replica of physical art is free in the same way - you can save a photo of it. If you want a physical replica, in both cases you can print it. An actual replica of physical art would be more expensive, but in many cases orders of magnitude cheaper than the original. Therefore, nobody buys expensive existing art just because they like it, which leaves two reasons I mentioned. >Ultimately this screams of solving a people problem with a technical one. Yes. The entire history of human civilization consists of solving social problems with technology.

ojkelly 5y ago

> The author doesn't understand the primary goal. It's precisely to have global consensus that proceeds according to its logic - so if you own something it can't be taken away from you. How is global consensus not centralisation though? Humans are notorious for disagreeing on everything. Having one global arbiter of truth/consensus seems implausible at best. Anything that could provide global consensus will need to be backed by something formed from a human brain. And our brains are not perfect. We keep updating code because we miss things. Who would maintain the system for global consensus? And how would they not just be our new rulers? As a perfect example you mention ownership as a feature of global consensus. What do you own with an NFT? Who enforces that ownership? What stops someone pirating an NFT image? Ultimately this screams of solving a people problem with a technical one. That never works. Not without strong co-operation of the people involved, and even then technology plays a very small role in the solution. > The scalability argument is true in isolation - everyone agrees it's currently a problem, but has nothing to do with centralization or not. It does in that a truly decentralised system inherently cannot suffer from the scaling problems cryptocurrencies have.

nootropicat 5y ago
>How is global consensus not centralisation though? In the same way reality itself is 'centralized'. Global consensus is meant to add a digital layer to reality. Reality is objective by definition. What's subjective is trying to measure it - but this problem doesn't exist for digital systems. >Who would maintain the system for global consensus? The practical implementation has to ensure that (1) consensus is maintained by a diverse set of potentially anonymous participants. Ideally, this would mean literally all humans, but that's of course impossible. The set should be large and diverse enough so that even powerful actors (especially governments) have no ability to force them to do something. DPoS, PoA fail here from the start. PoW is vulnerable over medium term: it has infinite economies of scale which leads to centralization and taking it over externally is always possible (because 'stake' in PoW is external and potentially infinite - mining hardware). (2) the most profitable course of action is for every validator to follow the protocol honestly. In ethereum's PoS that's achieved by slashing maliciously misbehaving validators. PoW has an orders of magnitude weaker punishment here (just mining rewards for the time spent on mining a minority chain). Most DPoS systems rely on the assumption that a majority is always honest, even if it makes financial sense to not be - eg. Cardano's Ouroboros. >Who would maintain the system for global consensus? And how would they not just be our new rulers? In the early period updates are necessary as system evolves - but eventually it has to be frozen, with the possible exception of scaling parameters (like block size, number of shards) which could be determined by a vote from validators. >What do you own with an NFT? Who enforces that ownership? There's no difference to real life here. "Owning" a house actually means there exists a publicly verifiable promise from an entity capable of wielding some form pf violence to enforce that property right. Therefore, an NFT that's connected to any external asset can only a different form of such a record. In the case of image NFTs, it's an improvement on physical art. The point of buying physical art is to either signal wealth, or to use it for money laundering. Image NFTs are infinitely superior for the first - because they are globally visible (as opposed to a physical piece of art) and fake NFTs are impossible (so you can't buy a cheap replica and pretend you spent a fortune on the original). They are much better for money laundering because they are inherently global (no transport issues) and evade identity checks in the banking system. Exactly like physical art, if you just like it and want to look at it - you can save a copy. A replica of physical art is free in the same way - you can save a photo of it. If you want a physical replica, in both cases you can print it. An actual replica of physical art would be more expensive, but in many cases orders of magnitude cheaper than the original. Therefore, nobody buys expensive existing art just because they like it, which leaves two reasons I mentioned. >Ultimately this screams of solving a people problem with a technical one. Yes. The entire history of human civilization consists of solving social problems with technology.