5 ms·

I found it excellent. It summarizes the main paradox of crypto: if you make it truly decentralized, it costs too much and is useless as a mass medium of exchang

by spupe 5y ago

I found it excellent. It summarizes the main paradox of crypto: if you make it truly decentralized, it costs too much and is useless as a mass medium of exchange, while also being a haven for fraud and theft. If you rely on any sort of ledger/layer 2 system, you are back to trusting a third party, in which case you would be better off with traditional banking.

PretzelPirate 5y ago
> If you rely on any sort of ledger/layer 2 system, you are back to trusting a third party It’s important to have definition of “trust” for this statement. If we use an Etheruem optimistic rollup with on-chain data availability guarantees as an example and we assume it’s out of beta, we have three situations where I can be in trouble: 1. Censoring my transactions 2. Refusal to allow me to withdraw my funds to L1 3. The network itself shuts down and there are no longer any operators If I am censored, I can force transaction inclusion by directly submitting the layer2 transaction to the inbox on Ethereum (the validator will get punished if they don’t include it). This will be more expensive for me since I have to interact with Ethereum, but it should only be done very rarely and my next rep should be to exit the L2 network. If the network refuses my withdraw transaction, I can perform an emergency exit by making a call to be L1 contract and demanding an exit happens (I can also use the strategy in the censoring case). If the network shuts down, then my only choice is to perform an emergency exit, which has probably already been started by other participants on the network. These options do get more complicated when you start using layer 2s with off-chain data availability committees (validiums and volitions) and that’s where you do need to trust third parties to ensure you can get the correct transaction data, but it only requires that one honest party is serving that transaction data. The trust requirements are actually pretty low for “out of beta” L2 networks while keeping the costs low enough that people can afford to use them. ZKRollups have similar features and even lower costs but ZK-proofs are still very new and harder to reason about.
- spupe 5y ago
  Sure, but if you want to qualify the "trust" to such minute detail, you can also do the same with traditional banking. I don't have to "trust" the banks, I have legal guarantees and transparency mandates enforced by governments. Why should I believe that the quasi-experimental procedure you describe operates smoothly, rather than believe that the traditional banking system will uphold its obligations, as it has for decades?
  - PretzelPirate 5y ago
    > Why should I believe that the quasi-experimental procedure you describe operates smoothly You don’t have to trust that the process works when you can see that it works by looking at the code and even become a validator/data provider yourself to ensure there’s always at least one honest participant (from your perspective). There was even uncertainty around the US forming a central bank with some worrying about the power it would have. Unless it would have been your approach already, not having confidence in L2s because they’re new technology isn’t a reason to dismiss them and decide to let the government manage all of your financial guarantees. > I don't have to "trust" the banks, I have legal guarantees and transparency mandates enforced by governments In this scenario, banks = L2 and government = L1. Are there scenarios where a bank (or other financial institution) can steal/withhold funds from you, or where they can censor your ability to transact with your own money? What is the recourse you have in those situations. What if the L1 itself uses its “right” to take your money through civil forfeiture or eminent domain? Are there cases where you may not trust the government (or where someone may not trust their government) because it is controlled by a small number of appointed or elected individuals? In that case, you can compare it to an L1/L2 blockchain setup where the L1 is controlled by a small number of validators (say 1000). That system would be easily corruptible (I can bribe 1k people) and I’d never treat it as a safe settlement system. Blockchain is all about your choice of who and what to trust. In the traditional financial system, some people will be fine with fully trusting their government and large financial institutions, whiles others will bury precious metals in the backyard and surround them with land mines. There’s all sorts of middle ground and blockchains let you choose where you want to sit on that spectrum.
    - spupe 5y ago
      > Are there scenarios where a bank (or other financial institution) can steal/withhold funds from you, or where they can censor your ability to transact with your own money? What is the recourse you have in those situations. What if the L1 itself uses its “right” to take your money through civil forfeiture or eminent domain? Eminent domain is about land use. Civil forfeiture and eminent domain are well described in the laws of all developed countries. And they exist for very good reasons, we don't want to get rid of this at all. The aim of the article is a critique of the potential for cryptocurrencies to be a true exchange medium equivalent to currencies as we have today. That is the illusion that feeds speculation and big dreams. If you are proposing that the use case for cryptos is individuals that have a reason to fear their own government, that leaves us criminals, political dissidents and wealthy oligarchs in third world countries. And it's still unclear whether they would be better served by storing their wealth in such a volatile and fraud-prone medium, versus actually digging a hole in the backyard as you describe. Or, you know, trusting another government/banking system like Switzerland like they all do.

by spupe 5y ago

PretzelPirate 5y ago

> If you rely on any sort of ledger/layer 2 system, you are back to trusting a third party It’s important to have definition of “trust” for this statement. If we use an Etheruem optimistic rollup with on-chain data availability guarantees as an example and we assume it’s out of beta, we have three situations where I can be in trouble: 1. Censoring my transactions 2. Refusal to allow me to withdraw my funds to L1 3. The network itself shuts down and there are no longer any operators If I am censored, I can force transaction inclusion by directly submitting the layer2 transaction to the inbox on Ethereum (the validator will get punished if they don’t include it). This will be more expensive for me since I have to interact with Ethereum, but it should only be done very rarely and my next rep should be to exit the L2 network. If the network refuses my withdraw transaction, I can perform an emergency exit by making a call to be L1 contract and demanding an exit happens (I can also use the strategy in the censoring case). If the network shuts down, then my only choice is to perform an emergency exit, which has probably already been started by other participants on the network. These options do get more complicated when you start using layer 2s with off-chain data availability committees (validiums and volitions) and that’s where you do need to trust third parties to ensure you can get the correct transaction data, but it only requires that one honest party is serving that transaction data. The trust requirements are actually pretty low for “out of beta” L2 networks while keeping the costs low enough that people can afford to use them. ZKRollups have similar features and even lower costs but ZK-proofs are still very new and harder to reason about.

spupe 5y ago
Sure, but if you want to qualify the "trust" to such minute detail, you can also do the same with traditional banking. I don't have to "trust" the banks, I have legal guarantees and transparency mandates enforced by governments. Why should I believe that the quasi-experimental procedure you describe operates smoothly, rather than believe that the traditional banking system will uphold its obligations, as it has for decades?
- PretzelPirate 5y ago
  > Why should I believe that the quasi-experimental procedure you describe operates smoothly You don’t have to trust that the process works when you can see that it works by looking at the code and even become a validator/data provider yourself to ensure there’s always at least one honest participant (from your perspective). There was even uncertainty around the US forming a central bank with some worrying about the power it would have. Unless it would have been your approach already, not having confidence in L2s because they’re new technology isn’t a reason to dismiss them and decide to let the government manage all of your financial guarantees. > I don't have to "trust" the banks, I have legal guarantees and transparency mandates enforced by governments In this scenario, banks = L2 and government = L1. Are there scenarios where a bank (or other financial institution) can steal/withhold funds from you, or where they can censor your ability to transact with your own money? What is the recourse you have in those situations. What if the L1 itself uses its “right” to take your money through civil forfeiture or eminent domain? Are there cases where you may not trust the government (or where someone may not trust their government) because it is controlled by a small number of appointed or elected individuals? In that case, you can compare it to an L1/L2 blockchain setup where the L1 is controlled by a small number of validators (say 1000). That system would be easily corruptible (I can bribe 1k people) and I’d never treat it as a safe settlement system. Blockchain is all about your choice of who and what to trust. In the traditional financial system, some people will be fine with fully trusting their government and large financial institutions, whiles others will bury precious metals in the backyard and surround them with land mines. There’s all sorts of middle ground and blockchains let you choose where you want to sit on that spectrum.
  - spupe 5y ago
    > Are there scenarios where a bank (or other financial institution) can steal/withhold funds from you, or where they can censor your ability to transact with your own money? What is the recourse you have in those situations. What if the L1 itself uses its “right” to take your money through civil forfeiture or eminent domain? Eminent domain is about land use. Civil forfeiture and eminent domain are well described in the laws of all developed countries. And they exist for very good reasons, we don't want to get rid of this at all. The aim of the article is a critique of the potential for cryptocurrencies to be a true exchange medium equivalent to currencies as we have today. That is the illusion that feeds speculation and big dreams. If you are proposing that the use case for cryptos is individuals that have a reason to fear their own government, that leaves us criminals, political dissidents and wealthy oligarchs in third world countries. And it's still unclear whether they would be better served by storing their wealth in such a volatile and fraud-prone medium, versus actually digging a hole in the backyard as you describe. Or, you know, trusting another government/banking system like Switzerland like they all do.