4 ms·

If you rely on systems that involve implicit custodial trust (i.e. not crypto) to coordinate with your homies, and you have powerful adversaries, your adversar

If you rely on systems that involve implicit custodial trust (i.e. not crypto) to coordinate with your homies, and you have powerful adversaries, your adversaries will compel those custodians to prevent you from communicating with your homies. They also might find you and hurt you. This happened with thepiratebay and DNS. It happened with wikileaks and paypal. It happened at the ISP level with Signal during recent unrest in Iran. It happens with the US banking system anytime somebody manages to get somebody else on one of the OFAC lists. It happens to would-be radicals on twitter all the time. On the other hand, systems that don't involve implicit custodial trust don't have these problems, which is why journalists in China occasionally publish on Ethereum, because the Chinese censors don't have a single-point-of-failure to attack. I don't think that currencies are an especially interesting application (although the wikileaks case is an example where they are relevant to group coordination). It's just that the currencies are what gets hyped because they're relevant to the investment side--without them, there's less money to build the other stuff. Another example which I expect to emerge, but which I don't think has been developed yet, would be the use of NFT's in a successor system. If the leader of a movement gets apprehended or incapacitated and they don't check in after a certain number of blocks, then a contract moves the leadership token to their successor. This denotes the successor's signing keys as authoritative, and the leader's (presumably compromised) keys as deprecated. If you keep signatures on-chain you can also tell which ones were registered at a time when they were valid. You couldn't do this on AWS, somebody would just compel AWS to remove the data they don't like, you need something like a blockchain. But it only works if enough people are running nodes to make attacking all of the node operators infeasible. For that you need to provide incentives, and now we're back to cryptocurrency. For me it comes down to whether you think dissent is more important than preventing all of the shady business that goes on in secret online. I do, so I'm pro-crypto. The currency bit is an ugly necessity which I hope we can find a way to get rid of.

JohnFen 3y ago
> If you rely on systems that involve implicit custodial trust (i.e. not crypto) to coordinate with your homies But there are myriad ways of coordinating without relying on a systems that involve implicit custodial trust. More so now than ever. But I get the sense we're talking about different things here. I was talking about coordinating activities with groups of people. You seems to be talking about payment systems. > If the leader of a movement gets apprehended or incapacitated and they don't check in after a certain number of blocks, then a contract moves the leadership token to their successor. You could use blockchain for this, but equally you can use any of several other methods to achieve the same effect. > For me it comes down to whether you think dissent is more important than preventing all of the shady business that goes on in secret online. For me, this doesn't enter into is. I do think dissent is critically important. I don't see how blockchain is essential to doing that. Why bring in something as tainted as cryptocurrency when there are other options?
- __MatrixMan__ 3y ago
  You keep saying that there are many alternatives, but I don't know what you mean. I'd like to outline a scenario and you can tell me which technology you think would scratch the itch: Suppose you're a journalist in a place with an oppressive government. You have information about government behavior that you expect will cause protests. You're not an activist though, so you don't want to actually plan those protests. Suppose I'm an activist in the same country--I'll plan the protest--but we don't know each other. We've got a friend-of-a-friend-of-a-friend... kind of relationship. Ideally you could post the information somewhere, and then our mutual friends could post annotations which validate it. They'd be attaching "I was there and I saw this happen" kind of information. I would then see this information because I explicitly/transitively trust people who have validated it. The government can inject disinformation about it, but it just gets ignored unless they manage to get us to explicitly trust their agents about this kind of thing (presumably this will be prohibitively expensive). The government will find us and hurt us if they catch three or four of us coordinating in this way. But if there are thousands of us, it will be a different story. So we need to collect a critical mass of validations, and we need to plan the protest, and we need to do it in a way that the government can't prevent and which doesn't betray our actual identities to the government--at least not until we all show up at the protest site. I'm working on this app, by the way. I intend to have pluggable backends so it won't be coupled to a specific blockchain--you could use it over ssh for instance, or a thumb drive, or via a shared S3 bucket--but it's the blockchain backends that I expect to be most useful in situations with many stakeholders and truly aggresive adversaries because it's not realistic to expect users to maintain the system of file transfers necessary to ensure that the annotations dataset converges in a way that gets the message and its validations from journalist to activist, or from activist to other activist. Sneakernet / file transfer will be the first backend that I support. What should I support second? What is tamper-resistant enough that the government can't remove the data, lacks dependency on a naming system which can be tampered with to hide the data, and is also public enough that all parties can access the data and know that the version they're accessing is the real one, if not a public blockchain? Granted, there may be solutions that are not technically a blockchain. I'm curious about Kademlia DHT's, for instance. I wonder if SSB can be used here. But so far as I can tell, all serious contenders are, if not blockchains, blockchain-like.

JohnFen 3y ago

> If you rely on systems that involve implicit custodial trust (i.e. not crypto) to coordinate with your homies But there are myriad ways of coordinating without relying on a systems that involve implicit custodial trust. More so now than ever. But I get the sense we're talking about different things here. I was talking about coordinating activities with groups of people. You seems to be talking about payment systems. > If the leader of a movement gets apprehended or incapacitated and they don't check in after a certain number of blocks, then a contract moves the leadership token to their successor. You could use blockchain for this, but equally you can use any of several other methods to achieve the same effect. > For me it comes down to whether you think dissent is more important than preventing all of the shady business that goes on in secret online. For me, this doesn't enter into is. I do think dissent is critically important. I don't see how blockchain is essential to doing that. Why bring in something as tainted as cryptocurrency when there are other options?

__MatrixMan__ 3y ago
You keep saying that there are many alternatives, but I don't know what you mean. I'd like to outline a scenario and you can tell me which technology you think would scratch the itch: Suppose you're a journalist in a place with an oppressive government. You have information about government behavior that you expect will cause protests. You're not an activist though, so you don't want to actually plan those protests. Suppose I'm an activist in the same country--I'll plan the protest--but we don't know each other. We've got a friend-of-a-friend-of-a-friend... kind of relationship. Ideally you could post the information somewhere, and then our mutual friends could post annotations which validate it. They'd be attaching "I was there and I saw this happen" kind of information. I would then see this information because I explicitly/transitively trust people who have validated it. The government can inject disinformation about it, but it just gets ignored unless they manage to get us to explicitly trust their agents about this kind of thing (presumably this will be prohibitively expensive). The government will find us and hurt us if they catch three or four of us coordinating in this way. But if there are thousands of us, it will be a different story. So we need to collect a critical mass of validations, and we need to plan the protest, and we need to do it in a way that the government can't prevent and which doesn't betray our actual identities to the government--at least not until we all show up at the protest site. I'm working on this app, by the way. I intend to have pluggable backends so it won't be coupled to a specific blockchain--you could use it over ssh for instance, or a thumb drive, or via a shared S3 bucket--but it's the blockchain backends that I expect to be most useful in situations with many stakeholders and truly aggresive adversaries because it's not realistic to expect users to maintain the system of file transfers necessary to ensure that the annotations dataset converges in a way that gets the message and its validations from journalist to activist, or from activist to other activist. Sneakernet / file transfer will be the first backend that I support. What should I support second? What is tamper-resistant enough that the government can't remove the data, lacks dependency on a naming system which can be tampered with to hide the data, and is also public enough that all parties can access the data and know that the version they're accessing is the real one, if not a public blockchain? Granted, there may be solutions that are not technically a blockchain. I'm curious about Kademlia DHT's, for instance. I wonder if SSB can be used here. But so far as I can tell, all serious contenders are, if not blockchains, blockchain-like.