3 ms·

Its less about a bunch of people checking for crypto, as it is about designing the protocol to make it such that doing the right thing is inherently in the econ

Its less about a bunch of people checking for crypto, as it is about designing the protocol to make it such that doing the right thing is inherently in the economic interest of the miners/stakers. Basically, unless you are able to get over 50% of the hashrate (or over 50% of the amount staked) you will be expected to make the most money by properly packing as many transactions into the block as you can (or more accurately packing in the set transactions that will net you the largest transaction fee). If you try to cheat your expected mining income will be less. For example if violate certain protocol rules no other node in the network will consider you block valid, so you get zero for your mining effort. Deliberately trying to mine based on an old block is technically legal. But the odds of your deliberate forked block becoming part of the longest chain is far less than the odds if you mine on top of the latest block, so again, it is in your economic interest to mine on the latest. Deliberately refusing to include valid transactions from specific users in the blocks you mine is possible, but (assuming those transactions have a large enough transaction fee to be sensible to include in the block) you are economically better off to include it, since otherwise one of your competitors will include it in a later block and get that money you lost out on. That is really the whole idea, to make it so that the economically rational thing to do is the correct thing. Now the system isn't perfect. if you can get over half the miners/stakers to collude with you, breaking some of the rules can become the more profitable approach. (and for some systems the relevant attack ratios could be larger or smaller than 50%, but the basic idea stands). The other part of the puzzle is that that collusion is a sort of prisoner's dilemma. Other people are not likely to want to join with you to collude unless they are pretty sure you will be able to get a large enough fraction of the hashrate to join with you, as otherwise they are risking making less money by joining with you. which is to say that the non-colluding case is a Nash Equilibrium. I don't see an obviously way to employ such an economic self interest approach against a single centralized server.

RC_ITR 4y ago
>That is really the whole idea, to make it so that the economically rational thing to do is the correct thing. Why is that not the case for Visa or any other centralized entity? You think there's a world in which they incorrectly process transactions and end up wealthier? Like you can keep re-explaining how proof-of-work functions (and trust me I get that), but why is it better than centralization?
- jsmith45 4y ago
  Yes VISA 100% without a doubt could make more money by processing transactions incorrectly in some cases. Foe example they could cause transactions that go to some startup company they consider a threat to their business model to fail some of the time, causing damage to the entity by making them seem flakey/unreliable, or costing them money in having to provide other payment options, or ask customers for payment again. Or they could fuck with payments to big businesses in certain countries in an attempt to manipulate exchange rates in a way as to maximize profits as reported is USD, etc. Alternatively it is very possible that in some cases VISA may decide to act in ways that don't maximize profits. This may sound surprising, but it is extremely common. First of all, companies operating on incomplete information will sometimes make decisions they think will maximize profits, but that actually wont. Or you get plenty of CEOs that decide to pull off dumb stunts to try to boost this quarter's numbers to better match analysts predictions, at the cost of reduced future profit. Companies are often sufficiently opaque that such things are not always immediately obvious. This is especially problematic for something like VISA, which gets to arbitrarily set the network rules. It is not like your business can just say, ok then VISA, I'll send transactions for your cards over the MasterCard network instead. Having a system that gets processed by a large number of different mutually distrustful entities helps avoid all this. It is unlikely that all or even most of them will view your startup as a threat. There will most likely be at least some that really are trying to maximize profits. And among those with slightly different goals (which are usually still somewhat of an approximation of maximizing profits), it is unlikely that a large majority, will deviate from expected behavior in exactly the same way (if they do, it would effectively be some of that 50+% collusion, even if it was accidentally rather than deliberate). The rules for the network cannot be arbitrarily changed on the whims of a single central entity. For somebody running a business, who does not particularly trust that some American run Card network wont fuck with their transactions, a crypto based payment network could legitimately look far nicer. They can more reliably predict the behavior of the network with respect to a transaction they send/receive. The network does not have a lot of arbitrary rules, and those rules don't change very often. Foreign governments you don't trust are less likely to be able to meddle in your transactions, etc. And to be clear, I'm no crypto shill. Realistically it gets used for wild speculation, scams, illegal or frowned upon transactions. There are some (but big picture still not all that many) legitimate entity to entity transactions not tied to somebody using crypto as speculative investments. I see that "Smart contracts" (which are neither smart nor contracts) while interesting are still deeply flawed, and realistically have limited uses not better performed by centralized systems. Etc. And all of that is not even considering the environmental caused by the use of work.

RC_ITR 4y ago

>That is really the whole idea, to make it so that the economically rational thing to do is the correct thing. Why is that not the case for Visa or any other centralized entity? You think there's a world in which they incorrectly process transactions and end up wealthier? Like you can keep re-explaining how proof-of-work functions (and trust me I get that), but why is it better than centralization?

jsmith45 4y ago
Yes VISA 100% without a doubt could make more money by processing transactions incorrectly in some cases. Foe example they could cause transactions that go to some startup company they consider a threat to their business model to fail some of the time, causing damage to the entity by making them seem flakey/unreliable, or costing them money in having to provide other payment options, or ask customers for payment again. Or they could fuck with payments to big businesses in certain countries in an attempt to manipulate exchange rates in a way as to maximize profits as reported is USD, etc. Alternatively it is very possible that in some cases VISA may decide to act in ways that don't maximize profits. This may sound surprising, but it is extremely common. First of all, companies operating on incomplete information will sometimes make decisions they think will maximize profits, but that actually wont. Or you get plenty of CEOs that decide to pull off dumb stunts to try to boost this quarter's numbers to better match analysts predictions, at the cost of reduced future profit. Companies are often sufficiently opaque that such things are not always immediately obvious. This is especially problematic for something like VISA, which gets to arbitrarily set the network rules. It is not like your business can just say, ok then VISA, I'll send transactions for your cards over the MasterCard network instead. Having a system that gets processed by a large number of different mutually distrustful entities helps avoid all this. It is unlikely that all or even most of them will view your startup as a threat. There will most likely be at least some that really are trying to maximize profits. And among those with slightly different goals (which are usually still somewhat of an approximation of maximizing profits), it is unlikely that a large majority, will deviate from expected behavior in exactly the same way (if they do, it would effectively be some of that 50+% collusion, even if it was accidentally rather than deliberate). The rules for the network cannot be arbitrarily changed on the whims of a single central entity. For somebody running a business, who does not particularly trust that some American run Card network wont fuck with their transactions, a crypto based payment network could legitimately look far nicer. They can more reliably predict the behavior of the network with respect to a transaction they send/receive. The network does not have a lot of arbitrary rules, and those rules don't change very often. Foreign governments you don't trust are less likely to be able to meddle in your transactions, etc. And to be clear, I'm no crypto shill. Realistically it gets used for wild speculation, scams, illegal or frowned upon transactions. There are some (but big picture still not all that many) legitimate entity to entity transactions not tied to somebody using crypto as speculative investments. I see that "Smart contracts" (which are neither smart nor contracts) while interesting are still deeply flawed, and realistically have limited uses not better performed by centralized systems. Etc. And all of that is not even considering the environmental caused by the use of work.