Post-Quantum Multi-Factor Wallet Designed for Humans

Transcript

[music] [music] All right. See, there awesome. Well, really uh happy to be here. I think DevCon um this year is especially great uh because all the people who don't actually care about this stuff went to AI uh and we can now focus on all the actual problems of crypto that need to get solved. So I'm pretty uh pretty happy. Love AI. It's amazing. But uh uh but still I think there's like there's just so much to do.

It's like um it doesn't matter. AI is not going to help. We need we need more than AI. Uh so I wanted to do a quick talk on postquantum multiffactor wallet design. Um I'm going to make a an admission. Uh I'm not like a Justin Drake, you know. I uh actually postquantum stuff not my cup of tea. Uh I think it's important. Um but I actually think there's a bigger problem that you'll see in a second. So uh we have heard from uh well now actual experts uh that we need to migrate off of um you know ECDSA and and our sort of standard uh standard cryptography and we need to kind of move to to postquantum

cryptography and and um and for that I defer you to people who actually uh you know have dedicated their lives to that. Um but uh none nonetheless it's it's happening very quickly. Um I I'm in this little Telegram group. I shouldn't be, but I am as like a little passive observer. And uh it's quite scary. It's like a lot of uh a lot of really smart people um saying words I don't understand using claude and uh seemingly breaking everything that I've come to love about cryptography uh and applied cryptography [snorts] and it's happening really really fast. So um you [clears throat] know so for this like you know the initial sort of reaction is uh is kind of like this uh you know uh

which I think was probably one of the best memes [laughter] when when uh that Justin Drake post came out. Uh pretty much sums it up you know uh it's like like we might actually all have to move to AI you know if you know if we didn't have any options but we actually have great options and um you know the the postquantum discussion has been going on for a long time. So, uh, the US government, everything else has already designed lots of different options. Um, they're not totally ideal, but they're there. And, um, we have tons of options for doing wallets, for doing crypto. And as Ethereum kind of moves towards this and moves towards, you know, lean lean VM and and um, and Stark

Stark proving everything, you know, we we we've got we've got good options for for building wallets. Um, I made a small uh, demonstration wallet. You can check this out at um bunker.supplies. And uh this is actually a um a it's to reduce your exposure. So I wouldn't say it's fully postquantum in and of itself um on the on the get-go, but uh the idea is you have a single proxy address and then you have rotating keys underneath. So kind of it's not it's not a stealth address because it's bonded to a proxy, but it's it's a rotating um owner. And the idea is every transaction you rotate to a new owner in your um in your derivation path. And um this nice little mechanism is actually quite I wouldn't

say post postquantum proof but it's um a lot better to shield your yourself with you know tools and things that we already have lying around that are fully production. So you're literally just, you know, you have a single single proxy for your wallet. That's your single address, but underneath the owner is continuing to rotate to a new owner that the public key is not exposed. [snorts] Um, so this little design is great. I also, just for fun, uh, couldn't stop there. So, I also added a um a two of two multiffactor design with this that you can check out that also tags on a unodudited version of DSA4 or an unodudited sphinx plus C10 um quantum signature as well. So you have a two of

two such that you have the rotating key is one uh which is production and then you have this unproduction unodudited yolo I audited it with deepseek and claude tagged on but it's still safe because generally either uh we're in postquantum so that's better than the ECDSA key or um we're not in postquantum and then you're protected by the rotating key in production element. So uh do definitely check out this uh this demo. Uh it's all open source so you can you can do it. Mainly this is just largely a demonstration of what postquantum could could sort of look like for us. Um and and how that that kind of design can work. But there's

actually a much bigger problem uh than postquantum in our uh in our space. Uh and the much bigger threat is um is basically yourself. [laughter] Uh so the the biggest enemy to um self-cussying your own funds is just you. Um you are your own worst security nightmare. And um you're basically going to [ __ ] it up and probably lose a lot of money. And if you have been in crypto for a while, pretty much ask everybody, they've all lost money self-custodying their own funds because human beings are really bad at this. So when we talk about self-custodying um key loss, bad backups passwords coercion device

theft. So some of these are not you, but some of these are you're setting yourself up for them. And um we have as humans a really hard time uh to actually solve this. So a very quick uh idea and I'm running out of time so I'll just be quick is basically to use uh a wallet where you know the factors are going to get lost and you use typical factors that human beings are actually good to manage. So laptop phone with pass keys and recovery >> [snorts] >> uh via like an email spending caps transfer delays and recovery paths are sort of the the major way to to get around this. And the idea is that with a multiffactor design you need to plan for a two of three loss. So, you actually need to plan for losing all of your

factors. And a trick that no one uses yet that's very exciting is to simply use a daily withdrawal on all the factors. When you do this, it means that any one factor can slowly withdraw all the funds. And this protects against the biggest threat, which is you losing many of your keys and then needing to get all your money back. So, having a recovery mech recovery mechanism like this is very important. And lastly, as well, um, in the design, using things like 24-hour hold windows and slowing down both an attacker and just slowing down the process is quite important. Um but similarly as well something that hasn't been implemented well yet is cancel and evacuate strategies largely around hey

um you know I've kind of [ __ ] up all my factors and now I need to uh recover or in the event of like an attacker that's stealing say your laptop and your phone now I need to shut things down. So I need to lock it down and go back to a recovery setup. So basically um these core mechanisms most people don't use them most people don't even know about them most people don't even have them and these are just basic human policies that uh every self-custody um uh kind of owner should have over their crypto. Um and the core idea is that uh while we protect um sort of the uh the the signing layer um via

cryptography and postquantum cryptography the policy layer is actually the thing that is completely underexplored and basically not many people are really tapped into yet but it's like super super basic. So policy layer just as important. So anyway, Justin Drake and people are going to take care of this multiffactor and policies are actually protecting you largely from yourself because uh in the end uh you will [ __ ] it up and probably lose all your money. So having a design like this protects you against yourself and against attackers as well. Happy to answer questions outside because I think I blew through my uh [music]