I’m mostly in the midst of the refi/regen cluster of the crypto community, so I’m constantly surrounded by all the good stuff that people try to do with crypto. I’m mostly thinking of the Funding the Commons conference series, projects by Protocol Labs, projects I’m a bit less familiar with like Ixo and Commons Stack, and the general cluster of things that Vitalik Buterin might get excited about. It’s all not as impact-focused as we are, but it’s good people.
I’ve been musing whether we could come up with something like “blockchain but without the money” to keep away the scammers. What’ll remain will probably not have a $1t market cap, but it’ll be web3 stripped down to its most valuable core elements. (Sadly, I have no idea, at this point, how “blockchain but without the money” would work.)
ACX/SSC has a history of writings on coordination failures, to the point that it has established the metaphor of “Moloch” for collective prisoners’ dilemmas that are stuck in defect-defect equilibria. Smart contracts let you test out and rapidly iterate new coordination mechanisms. Maybe the field of mechanism design is cursed and there are no better coordination mechanisms to be found than the ones that we have. Maybe seemingly valuable mechanisms like Pol.is or quadratic voting or my impact markets will turn out to be flawed. But at the moment I’m optimistic that the EV of experimentation with new coordination mechanisms is high, and a lot of that is happening in various crypto communities.
There’s probably also value in breaking up monopolies to foster innovation. A lot of this experimentation with coordination mechanisms and whatnot would work just fine without blockchains (e.g., Pol.is). But I trust PayPal and Stripe to not rug-pull me because they’re big and old; they wouldn’t have gotten big and old if they rug-pulled people. (And they don’t care about rug-pulling me specifically because they’re big and I’m small.) But it takes a lot of time to get big and old. If we want to foster quick experimentation with coordination mechanisms that require trust, we need a quicker way to establish trust – such as to put all the smart contracts on a public blockchain and have them audited. All sorts of things can still go wrong, but it’s a start, and it makes the trust advantage of big, old companies a bit less expensive to acquire.
There is that exchange-like platform that I’m building. Building something like Kraken from scratch is difficult because the security is so tricky and important to get right (unless you’re callous about it). For a lot of standard use cases you can use Stripe or PayPal instead of building them from scratch, and you don’t need to think about the security. But for all the rest of use cases, you have no other choice than to code it yourself, get the licenses, hire the security experts, etc.… Unless there is a provably secure blockchain with audited software built on top of it that you can just use. (Cardano is the Motte here but I can make do with the level of security that Ethereum or Solana provide.) Defi will probably be somewhat more mature by the time of the next crypto summer and will make it a lot more efficient to implement secure coordination mechanisms. You can still make mistakes when you start writing your own smart contracts (though there is audited template code and the SPL for a lot of stuff), but it’s less code, so fewer mistakes.
Those are the side of crypto that I’m most exposed to.
Preventing s-risks from AI, which can result from coordination failures, is the highest priority in my mind. So the general field of innovation of coordination mechanisms seems plausibly extremely valuable to me, at a high level. (Plausibly even more valuable than enabling people to conduct financial transactions without having to fight with the corrupt or inefficient systems in their countries.)
Whether all the concrete stuff that is being trailed in the various crypto communities is at all relevant to s-risks in the end is difficult to answer. But I’m only trying to defend the thesis here that all the innovation that is happening in the blockchain space is highly valuable, not that it can compete with (s-risk-focused) AI safety. It’s much too unfocused for that, and I don’t yet know the particular blockchain projects that might be more focused.
That said, I don’t know whether I agree with your conclusion. Usually people go like “Here’s a cool solution for incentivizing carbon capture,” and the response is, “Crypto is a scam. Blocking you.” But just now I saw an EA tweet something and someone replied, “EA is a scam. Blocking you.” That’s a new one to me. So I’m afraid the public perception of crypto might be infecting EA. I feel a lot more protective of EA than of crypto, so I kind of want to shield it. Then again that goes dangerously far in the direction of doing things because of their “optics,” which is also cringe.
One solution could be to just wait a few more years to allow defi solutions to mature to the point where the usability rivals cefi. Maybe then we’ll be able to draw on all the blockchain innovation and collaborate with companies in the space without looking distinctively cryptoesque while doing so. The best typeface is the one that the reader doesn’t notice, so I hope defi will be like Linux Libertine soon.
Or put differently: Using an acoustic coupler in a phone booth to post on the usenet might’ve come off as a distinctly nerdy thing to do. So if you don’t want to appear nerdy, that’d be something to avoid. But commenting on TikTok today is much less distinctively nerdy. So maybe we can have our EA cake and eat it too once defi becomes more like TikTok.
You can either: (1) take a smart contract platform and remove the currency and financial applications. (2) take a smart contract platform and replace the native currency by a fully fiat backed collatoral. You keep the money, but at least the issuance is state-controlled.
I’m mostly in the midst of the refi/regen cluster of the crypto community, so I’m constantly surrounded by all the good stuff that people try to do with crypto. I’m mostly thinking of the Funding the Commons conference series, projects by Protocol Labs, projects I’m a bit less familiar with like Ixo and Commons Stack, and the general cluster of things that Vitalik Buterin might get excited about. It’s all not as impact-focused as we are, but it’s good people.
I’ve been musing whether we could come up with something like “blockchain but without the money” to keep away the scammers. What’ll remain will probably not have a $1t market cap, but it’ll be web3 stripped down to its most valuable core elements. (Sadly, I have no idea, at this point, how “blockchain but without the money” would work.)
ACX/SSC has a history of writings on coordination failures, to the point that it has established the metaphor of “Moloch” for collective prisoners’ dilemmas that are stuck in defect-defect equilibria. Smart contracts let you test out and rapidly iterate new coordination mechanisms. Maybe the field of mechanism design is cursed and there are no better coordination mechanisms to be found than the ones that we have. Maybe seemingly valuable mechanisms like Pol.is or quadratic voting or my impact markets will turn out to be flawed. But at the moment I’m optimistic that the EV of experimentation with new coordination mechanisms is high, and a lot of that is happening in various crypto communities.
There’s probably also value in breaking up monopolies to foster innovation. A lot of this experimentation with coordination mechanisms and whatnot would work just fine without blockchains (e.g., Pol.is). But I trust PayPal and Stripe to not rug-pull me because they’re big and old; they wouldn’t have gotten big and old if they rug-pulled people. (And they don’t care about rug-pulling me specifically because they’re big and I’m small.) But it takes a lot of time to get big and old. If we want to foster quick experimentation with coordination mechanisms that require trust, we need a quicker way to establish trust – such as to put all the smart contracts on a public blockchain and have them audited. All sorts of things can still go wrong, but it’s a start, and it makes the trust advantage of big, old companies a bit less expensive to acquire.
There is that exchange-like platform that I’m building. Building something like Kraken from scratch is difficult because the security is so tricky and important to get right (unless you’re callous about it). For a lot of standard use cases you can use Stripe or PayPal instead of building them from scratch, and you don’t need to think about the security. But for all the rest of use cases, you have no other choice than to code it yourself, get the licenses, hire the security experts, etc.… Unless there is a provably secure blockchain with audited software built on top of it that you can just use. (Cardano is the Motte here but I can make do with the level of security that Ethereum or Solana provide.) Defi will probably be somewhat more mature by the time of the next crypto summer and will make it a lot more efficient to implement secure coordination mechanisms. You can still make mistakes when you start writing your own smart contracts (though there is audited template code and the SPL for a lot of stuff), but it’s less code, so fewer mistakes.
Those are the side of crypto that I’m most exposed to.
Preventing s-risks from AI, which can result from coordination failures, is the highest priority in my mind. So the general field of innovation of coordination mechanisms seems plausibly extremely valuable to me, at a high level. (Plausibly even more valuable than enabling people to conduct financial transactions without having to fight with the corrupt or inefficient systems in their countries.)
Whether all the concrete stuff that is being trailed in the various crypto communities is at all relevant to s-risks in the end is difficult to answer. But I’m only trying to defend the thesis here that all the innovation that is happening in the blockchain space is highly valuable, not that it can compete with (s-risk-focused) AI safety. It’s much too unfocused for that, and I don’t yet know the particular blockchain projects that might be more focused.
That said, I don’t know whether I agree with your conclusion. Usually people go like “Here’s a cool solution for incentivizing carbon capture,” and the response is, “Crypto is a scam. Blocking you.” But just now I saw an EA tweet something and someone replied, “EA is a scam. Blocking you.” That’s a new one to me. So I’m afraid the public perception of crypto might be infecting EA. I feel a lot more protective of EA than of crypto, so I kind of want to shield it. Then again that goes dangerously far in the direction of doing things because of their “optics,” which is also cringe.
One solution could be to just wait a few more years to allow defi solutions to mature to the point where the usability rivals cefi. Maybe then we’ll be able to draw on all the blockchain innovation and collaborate with companies in the space without looking distinctively cryptoesque while doing so. The best typeface is the one that the reader doesn’t notice, so I hope defi will be like Linux Libertine soon.
Or put differently: Using an acoustic coupler in a phone booth to post on the usenet might’ve come off as a distinctly nerdy thing to do. So if you don’t want to appear nerdy, that’d be something to avoid. But commenting on TikTok today is much less distinctively nerdy. So maybe we can have our EA cake and eat it too once defi becomes more like TikTok.
You can either:
(1) take a smart contract platform and remove the currency and financial applications.
(2) take a smart contract platform and replace the native currency by a fully fiat backed collatoral. You keep the money, but at least the issuance is state-controlled.
What remains is a permissioned blockchain system (hyperledger, multichain) where you can still reach consensus on state. Multichain has a lot of great articles on this: https://www.multichain.com/blog/2016/03/blockchains-vs-centralized-databases/