Oddbean

Not many people have the ability or the knowledge to audit all the code they use. At what point do you stop auditing and start trusting? How do we define that point? Surely it differs for everyone. For instance, I don't know C++. I have not audited bitcoin core in depth. Nor many other apps running on my laptop.

There’s a difference in trusting code versus trusting others with your bitcoin. You don’t need to know how a car works in order for you to drive a car. However, it would be naive to trust your car keys to a complete stranger.

I have said Fedimint could be a good idea for certain situations especially families or small communities. But to scale to 7 billion people I am very skeptical. You need something more user friendly and easy to use thats also trust minimised as possible.

I think fedimint bros think that they'll be the one runnung the fedimint, so they never consider the downsides. I wouldn't use a fedimint run by my family or friends because I prefer not to have money involved in those relationships. Most people feel this way. May take years of pain to realize that a dynamic federation kept sybill resistant and annonymous through proof of work is a good scaling solution but it will happen. Let's hope the self delusion of fedimint bros doesn't retard this realization for so long that another coin gets more network effect.

When you say "most people feel this way", what part of the world are you referring to? What downsides are you referring to? please elaborate. There are plenty of trade offs in all things, but if you're more specific, I can better address the concern. In your last paragraph, are you referring to another "coin" aside from bitcoin? It reads like you are but you're unspecific. Also, the ad hominem attack doesn't help convince anyone. Give me real arguments. What coin are we waiting for network effect to increase? The market has been pretty clear on what it values the most at the moment.

Thanks for the thoughtful reply, I'll respond to each paragraph: 1) Not to any particular part of the world, to most people generally. It's simple to see that this true, if you have money, you keep it in a safe place only you have access to or in a bank, not with your brother. I think this was true even in the days of gold and silver. 2) The main downside is that the custodians might collude and run away with your money. The other downsides are the politics involved in chosing the custodians, the day to day annoyances of running the mint (even if the software is great there will always be tech issues that will have to be dealt with), the politics of chosing a fair fee structure to compensate the custodians, etc. In addition, having your friends and family serve as custodians might be better in terms of decentralization but it also introduces drama into relationships that you wpuldn't otherwise have. 3) It could be any "coin", the point is that it is a mistake to think thay Bitcoin is invincible. Ethereum has a lot of network effect and if they get a scaling solution with less downsides it might be preferred by the billions of people that will need to be onboarded in the coming decades. More worringly, cbdc-esque "coins" will come with the threat of state violence to reinforce their network effect, of course they have their own downsides (lol) but the fewer downsides our scaling solutions have the better.

1. And now most of the world trusts banks with their "money" (currency really). I don't think this paradigm will go away and making people comfortable with the personal responsibility of holding all of your wealth yourself will take a long time. In the meantime, custodians are going to exist. Giving people a privacy preserving trade off for a custodian seems a good way to move toward a better future, while we educate the masses on self-custody and personal responsibility. 2. Agree with all of this. Trade offs always. There is going to be complexity with any technology until there's not. It used to be super complex and require lots of people to run mainframes, and now we have super computer in our pockets. The more time passed, the more abstracted away from electrons on a motherboard we get. Even 10 years ago, managing a single bitcoin wallet was a super technical task, but today I can download an app to my phone and hold the keys to some coin. 3. I get your point here and don't disagree. This is why I dedicate my free time to working on a solution that I think will help adoption of bitcoin. The "coin" with the best tools will capture more economic share. I hope you're building your ideal future as well. (Personally, I think ETH will be what they build the CBDCs on. All the code and infrastructure is there for the State to scoop up.) likewise, thanks to you for the reply.

Glad we agree on so much but you're missing the point. Friends and family are bad custodians. Bitcoin miners are better custodians (dynamic and annonymous so very hard for governments to pressure, incentive aligned with Bitcoin, etc. ) Privacy on a chaumian mint is comparable to a zcash sidechain but the zcash sidechain also protects users from the custodians ibflating the supply. So the sidechain is strictly better. We have millions of dollars and tons of smart young people (fedimint bros) working on a worse tool and telling their followers that an opt-in softfork that enables a zcash drivechain is "dangerous" because "shitcoins on bitcoin" or "changes incentives". This is a problem. That being said, I don't think working on fedimints is bad or anything, we should have as many solutions as we possibly can and let the best one win.

I am. But I thought Hawala was for payments. Not a bank like fedimint. More akin to lightning. It makes sense to me that having friends or family as intermediaries for a short duration could work since afterwards the relationship goes back to being unencumbered with financial concerns.

Why can't ecash be for payments? I personally wouldn't want large amounts of long-lived amounts stored in a fedimint. The former reducing overall rug risk and the latter reduces the chance of inflationary practices (aside from running a client that validates the version of fedimint run by guardians). Fedimint doesn't explicitly have to be a "bank", although, admittedly, that's been a advertised use case from louder voices than mine own. Laying out an example: I'm visiting LA. I somewhat know people in the area, maybe I'm visiting them. There's a smaller network of merchants that participate in a fedimint. Similar to handling cash, they don't keep large amounts in the drawer (in the fedimint). Just enough to do business of the day. When i arrive, I can peg-in (on-chain or with LN) to the fedimint and now have ecash to spend. Each merchant and myself don't have to mange LN nodes and the complexity inherent to it. Another: Remove the ecash/BTC/LN modules and insert a password manger or any other arbitrary data. Since the data is validated via consensus among fedimint nodes, the risk and access is distributed. It enables the syncing and validity across disaggregate nodes. I personally would love to auto-sync my bitwarden instance between home and office servers. I understand there are solutions to that with some automation, but fedimint could offer this out of the box for the less technically inclined. Myriad of solutions and fedimint could be one. Definitely not the end all be all, but i have to pick something to focus on with my limited time. 🫠

Exactly my original point. You, a technically competent fedimint enthusiast, don't see any problems because you can deposit onchain or over ln, do your business and get out, doesn't matter who the custodian is because most of your meaningful transactions happen non-custodially. As we've already agreed, some form of custody will be necessary for most people in thw future (impossible for billions of people to own their own utxos). Thw question then becomes what does the ideal custodian look like? Is it your uncle Jim and neighbor Johnny or is it a sybill resistant, annonymous, dynamic collection of parties deeply invested in the succees of Bitcoin as a money?

Fair point on the technically competent piece but I point back to wifi. That seemed unobtainable to many for a number of years and is now ubiquitous. You now longer have to get some external card to slide into a slot on your laptop to get wifi. It just works ™️ You point presupposes that humans won’t learn and adapt to using new technology over time. I disagree. The motivation to learn the tech or to abstract the tech to simpler UX exists and grows with time. Look at bitcoin itself. You used to have to run CLI commands to interact with the network. Now…

The more important point, is that not everyone can own a utxo. So custodians will be necessary, not for you, but for small people. You think they'd be better off keeping their money in the custody of a sister and some neighbors or in the custody of a sybill resistant, annonymous, incentive aligned, Bitcoin miners?