1 00:00:00,000 --> 00:00:02,000 And do you call it shards? 2 00:00:02,000 --> 00:00:08,000 I have completely scrubbed the word shards from my app and like marketing materials. 3 00:00:08,000 --> 00:00:09,000 Yeah. 4 00:00:09,000 --> 00:00:12,000 And it, almost all of the code. 5 00:00:12,000 --> 00:00:14,000 So I'm happy to use that term on the podcast. 6 00:00:14,000 --> 00:00:20,000 What I found is that trying to have the consistent metaphor of keys and vaults. 7 00:00:20,000 --> 00:00:27,000 So I call them keys in the app, which is also confusing for those of us who know who are used to having like scuttlebutt keys. 8 00:00:27,000 --> 00:00:28,000 Super confusing. 9 00:00:28,000 --> 00:00:29,000 Stuff like that. 10 00:00:29,000 --> 00:00:44,000 But for people, I don't know, for the people I was testing with, the metaphor of needing multiple keys to open a vault landed a lot better than the metaphor of like bringing these shards together to reassemble a crystal or something. 11 00:00:44,000 --> 00:00:45,000 Yeah. 12 00:00:45,000 --> 00:00:46,000 Yeah. 13 00:00:46,000 --> 00:00:47,000 No, that makes sense. 14 00:00:47,000 --> 00:00:53,000 I'm not convinced that I have, you know, like the perfect language and metaphor and everything. 15 00:00:53,000 --> 00:00:55,000 Well, maybe that can be a shout out. 16 00:00:55,000 --> 00:01:00,000 We can ask for suggestions from people who listened to this episode. 17 00:01:00,000 --> 00:01:01,000 Okay, cool. 18 00:01:01,000 --> 00:01:03,000 I might actually keep this part. 19 00:01:03,000 --> 00:01:04,000 Okay. 20 00:01:04,000 --> 00:01:07,000 Because I think that's interesting. 21 00:01:07,000 --> 00:01:13,000 But before we get too much deep into interesting stuff, let's just start. 22 00:01:13,000 --> 00:01:14,000 You ready? 23 00:01:14,000 --> 00:01:15,000 Okay. 24 00:01:15,000 --> 00:01:16,000 I'm ready. 25 00:01:23,000 --> 00:01:45,000 Welcome to today's episode of solar cast. 26 00:01:45,000 --> 00:01:48,000 And today we're joined by Matt Lawrence. 27 00:01:48,000 --> 00:02:00,000 And Matt Lawrence has been working on a project that is an evolution of a fan favorite, one could say, back in the day it was called dark crystal. 28 00:02:00,000 --> 00:02:17,000 And it became famous for being a distributed or decentralized way to get back access to your account if you lost your private key without having the need for a centralized admin of sorts. 29 00:02:17,000 --> 00:02:27,000 And Matt has evolved this concept quite far in his project called Horcrux. 30 00:02:27,000 --> 00:02:34,000 And today we're going to hear more about the status and the evolution so far. 31 00:02:34,000 --> 00:02:38,000 Now without further ado, let's dive in. 32 00:02:38,000 --> 00:02:47,000 So good to have you here, Matt. 33 00:02:47,000 --> 00:02:56,000 And we recently met for the first time ever, even though we've known each other for longer at the D web camp. 34 00:02:56,000 --> 00:03:00,000 And that was this summer in Berlin. 35 00:03:00,000 --> 00:03:02,000 Yeah, it was a great time. 36 00:03:02,000 --> 00:03:05,000 It was my first time in Berlin, first time in Germany. 37 00:03:05,000 --> 00:03:11,000 And yeah, first time meeting several people that I only knew on the internet. 38 00:03:11,000 --> 00:03:21,000 And it's quite interesting because your project, Horcrux, is kind of an evolution of a project that I guess you heard about from the internet. 39 00:03:21,000 --> 00:03:30,000 But like, did you, were you physically in touch or were you part of the crew that was like the predecessor with like dark crystal? 40 00:03:30,000 --> 00:03:32,000 Or how has that evolved? 41 00:03:32,000 --> 00:03:35,000 Yeah, so I was not part of the crew. 42 00:03:35,000 --> 00:03:44,000 I was just a fan of the dark crystal project back in the day when they were working on Scuttlebutt. 43 00:03:44,000 --> 00:03:50,000 They had some pieces integrated into the patch bay client. 44 00:03:50,000 --> 00:03:57,000 And I think that's where I first learned about this nerdy algorithm called Shamir Secret Sharing. 45 00:03:57,000 --> 00:04:08,000 The idea of social backup, you know, like having your friends essentially be responsible for, or be the source of truth for which, which key on the network is yours. 46 00:04:08,000 --> 00:04:20,000 But essentially who you are on the internet, you know, not having a corporation or just a single private key be the, like, the final source on that was really compelling. 47 00:04:20,000 --> 00:04:25,000 So you're diving straight in. And for those who have heard of dark crystal, this might be familiar. 48 00:04:25,000 --> 00:04:33,000 But for those who have not, if we take two steps back and look at like, why did it matter? 49 00:04:33,000 --> 00:04:39,000 Or what was it that inspired you about dark crystal originally and what is that you're carrying with you into Horcrux? 50 00:04:39,000 --> 00:04:41,000 Yeah, that's a great question. 51 00:04:42,000 --> 00:04:49,000 I mean, I feel like I trace a lot of my current work back to where I was in university. 52 00:04:49,000 --> 00:04:57,000 You know, I think I had just switched to computer science as my major and the Edward Snowden revelations came out. 53 00:04:57,000 --> 00:04:59,000 And that was really like a fork in the road for me. 54 00:04:59,000 --> 00:05:04,000 I think I got into computers and technology and programming because it was fun. 55 00:05:04,000 --> 00:05:08,000 It felt powerful and a way to express myself creatively. 56 00:05:08,000 --> 00:05:13,000 But that was like a, the turning point for me of like, oh, these are not just fun tools. 57 00:05:13,000 --> 00:05:18,000 You know, those were the early days of like Twitter and Facebook and what did it feel like back then? 58 00:05:18,000 --> 00:05:23,000 To me as a, you know, as a young man, it just, it just felt fun. 59 00:05:23,000 --> 00:05:29,000 Like it was fun to learn from other people that I found interesting. 60 00:05:29,000 --> 00:05:51,000 You know, back then I was all into iOS development and all these indie iOS developers, you know, kind of living this, this dream of like working on some little app that you're passionate about and finding, paying customers in this new app store ecosystem that didn't, it didn't make a lot of sense to people back then that people would pay money for apps. 61 00:05:52,000 --> 00:06:00,000 But they were and people were making livings. That was kind of like my, yeah, my entry point into, into the career anyway. 62 00:06:00,000 --> 00:06:09,000 Yeah. But then you kind of diverged from the path of app developer in the OS environment, so to speak. 63 00:06:09,000 --> 00:06:15,000 And took a turn towards peer to peer. 64 00:06:15,000 --> 00:06:18,000 What brought you there and eventually towards start crystal? 65 00:06:18,000 --> 00:06:30,000 Yeah. So that, that moment led me down a path of, of, of feeling some responsibility of like, oh, like I, I need to be investigating these alternatives. 66 00:06:30,000 --> 00:06:45,000 Like I need to be using them because like how, how, how, how are, you know, how is my grandma ever going to like get off of Facebook unless I first get off of Facebook and like find something, find the better thing or. 67 00:06:45,000 --> 00:06:58,000 Yeah. And so, yeah, that's what led me on to these alternative networks of, you know, tent and it was app.net and eventually I found scuttlebutt and. 68 00:06:58,000 --> 00:07:14,000 And yeah, that's, that's where I encountered dark crystal and honestly a whole subculture that I was, you know, a solar punk and even just like a lot of more progressive values in general that I really. 69 00:07:14,000 --> 00:07:16,000 Hadn't been exposed to in that way. 70 00:07:16,000 --> 00:07:17,000 Yeah. 71 00:07:17,000 --> 00:07:27,000 Yeah. So the, these ideas that, you know, that corporations should not be, should not be the ones having the final say about who gets to participate in the internet. 72 00:07:27,000 --> 00:07:30,000 What happens to your data? How is it used? 73 00:07:30,000 --> 00:07:40,000 And to today where now we have a whole different scale of information and essentially behavior control that has developed out of these big tech companies. 74 00:07:40,000 --> 00:07:49,000 I think that's for a lot of people is what has led to this generation of P2P or self sovereign identity networks and protocols. 75 00:07:49,000 --> 00:07:58,000 Consistently the biggest hurdle on these protocols for most people is key management, right? Because instead of having a email address and a password, 76 00:07:58,000 --> 00:08:01,000 you can reset anytime you forget it. 77 00:08:01,000 --> 00:08:06,000 You essentially have this private key, which is like a password that no one can reset for you. 78 00:08:06,000 --> 00:08:16,000 And you are expected to keep it extremely safe and yeah, this just doesn't work for normal people, right? It never really has. 79 00:08:16,000 --> 00:08:18,000 And so this whole space I've. 80 00:08:18,000 --> 00:08:35,000 It's also the shift going from having like an account where you sign up with your email and you just press, I forgot my password and like the amount of times I've registered to get like an account and then immediately within two minutes, like lost my password. 81 00:08:36,000 --> 00:08:38,000 It's like a lot. 82 00:08:38,000 --> 00:08:48,000 And I mean, it's not as easy to lose your private key necessarily, but I mean, since it's stored on your computer locally. 83 00:08:48,000 --> 00:08:52,000 But still, if you lose it, you lose access to everything. 84 00:08:52,000 --> 00:08:55,000 And as you say, there's no central point of access. 85 00:08:55,000 --> 00:08:59,000 There's no admin you can write that's like, oh yeah, retrieve your key for you. 86 00:08:59,000 --> 00:09:04,000 But like before we dive into how to retrieve your key, why have that system at all? 87 00:09:04,000 --> 00:09:10,000 Why why have a key based identity system digitally rather than like logins? 88 00:09:10,000 --> 00:09:12,000 Oh yeah, that's that's a great question. 89 00:09:12,000 --> 00:09:20,000 I mean, I think there's some consensus now that like representing people with keys is actually also not great. 90 00:09:20,000 --> 00:09:30,000 Just like making making some corporation the authority on your identity on who you are and what you can do. 91 00:09:30,000 --> 00:09:37,000 It's also, there are also many problems with representing one person as one key. 92 00:09:37,000 --> 00:09:49,000 But I think the reason that it's that it's taken off and the reason that a lot of these networks use it despite knowing that we need to do better is because of the cryptography, like the math that exists around. 93 00:09:49,000 --> 00:10:09,000 A public key cryptography allows you to prove, essentially give it some data and you can prove to others that it's from you, which means it opens up possibilities for all these other types of systems where the infrastructure is kind of separate from the integrity of the data. 94 00:10:10,000 --> 00:10:18,000 And I think what that means for normal people is that like, is that we don't have to trust these particular like servers or companies or anything. 95 00:10:18,000 --> 00:10:26,000 So like on Secure Scuttle, but you can you can pass, you know, messages are passed around over a gossip protocol. 96 00:10:26,000 --> 00:10:33,000 So I might get your messages from someone else and I'm still able to verify that they came from me. 97 00:10:33,000 --> 00:10:34,000 Exactly. 98 00:10:34,000 --> 00:10:35,000 Yeah. 99 00:10:36,000 --> 00:10:51,000 So why, because you mentioned that there's like critique and that there's now like a sort of the direction conversations are going is that like key pairs are not good for it entities. 100 00:10:51,000 --> 00:10:56,000 And you said that that critique exists, but not exactly what it is. 101 00:10:56,000 --> 00:10:58,000 I'm really curious now. 102 00:10:58,000 --> 00:11:02,000 What is the current critique towards that system? 103 00:11:02,000 --> 00:11:05,000 I'm not the best person to speak on this. 104 00:11:05,000 --> 00:11:06,000 I wish I were. 105 00:11:06,000 --> 00:11:16,000 I know I've read some really great articles by Philip Sheldrake on this and a lot of people in the Medigov community, I think, are making a big push right now towards experiments with alternatives. 106 00:11:16,000 --> 00:11:23,000 I'm sure there are others too, but I think part of the critique is that keys are too reducible. 107 00:11:24,000 --> 00:11:32,000 There's also, I think, a lot of energy behind people having different identities or like different contexts in which we present ourselves. 108 00:11:32,000 --> 00:11:35,000 Like we do this socially all the time in person. 109 00:11:35,000 --> 00:11:43,000 On the internet, it's really unnatural for kind of like this one identifier to follow you around to all these other, all these different systems and contexts. 110 00:11:43,000 --> 00:11:49,000 And so even as humans, like we don't just assume one identity. 111 00:11:49,000 --> 00:11:52,000 And so having one key is problematic in that way as well. 112 00:11:52,000 --> 00:11:59,000 Like having an identity, like a password manager kind of situation or distributed identity. 113 00:11:59,000 --> 00:12:02,000 That's possible and I think some systems do this. 114 00:12:02,000 --> 00:12:10,000 Part of the reason why I think more of our decentralized protocols haven't taken this on is that the cryptography for it is really hard. 115 00:12:10,000 --> 00:12:14,000 And then the UX on top of the cryptography. 116 00:12:14,000 --> 00:12:20,000 Like how do you even talk to users about multiple keys in your app and how you manage them? 117 00:12:20,000 --> 00:12:26,000 To my knowledge, there are no systems being used by more than 100 people that have actually done. 118 00:12:26,000 --> 00:12:36,000 The thing that we all want I think which is a system of strong rotation and having morality of identities that are in some sense loosely tied to a human. 119 00:12:36,000 --> 00:12:44,000 Do you think that's because the use case patterns just haven't really like fallen into place because humans aren't used to using these systems? 120 00:12:44,000 --> 00:12:52,000 Or because financial incentives aren't there to make decentralized systems flourish? 121 00:12:52,000 --> 00:12:54,000 Or is it a multitude of them? 122 00:12:54,000 --> 00:12:57,000 I think it's probably both, you know. 123 00:12:57,000 --> 00:13:08,000 And I guess one of the critiques that you mentioned earlier on was the aspect of like what do you do when you lose your key and how risky that is. 124 00:13:08,000 --> 00:13:16,000 And I guess this is exactly one of the main challenge that Horcrux seeks to solve. 125 00:13:16,000 --> 00:13:17,000 Yeah, yeah. 126 00:13:17,000 --> 00:13:30,000 Yeah, Horcrux is coming at this problem from the direction of like how could we incorporate more social into these kind of anti-social, anti-human, strict key systems. 127 00:13:30,000 --> 00:13:33,000 So it's not a silver bullet. 128 00:13:33,000 --> 00:13:47,000 It does not solve all these problems with keys, but it certainly is solving some problems in a way that I think is consistent with the values that a lot of us share. 129 00:13:47,000 --> 00:14:05,000 Solving social trust with social solutions rather than solving social trust with mathematical solutions, which oftentimes in cryptography scenes, like because one's dealing with keys and security, there's so many different attempts at solving a social issue with a mathematical solution. 130 00:14:05,000 --> 00:14:09,000 And that's something really beautiful that I think about this core idea. 131 00:14:09,000 --> 00:14:17,000 But before we dive into talking about it too much because there's a bunch of different side aspects here, could you just like give us the rundown? 132 00:14:17,000 --> 00:14:18,000 What is it? 133 00:14:18,000 --> 00:14:29,000 Yeah, so Horcrux is an app that helps you back up sensitive data to the devices of your friends and family or whoever you choose. 134 00:14:29,000 --> 00:14:31,000 There's some cryptography there. 135 00:14:31,000 --> 00:14:36,000 There's a lot more that we'll get into, but at a root level, it's a backup tool. 136 00:14:36,000 --> 00:14:41,000 Back up to what pictures, for how far does it extend? 137 00:14:41,000 --> 00:14:44,000 How far can we draw this abstraction? 138 00:14:44,000 --> 00:14:47,000 I mean, I think I like to say sensitive data. 139 00:14:47,000 --> 00:14:51,000 So the things that I think of primarily are keys. 140 00:14:51,000 --> 00:14:54,000 Some keys, not every key belongs in the system. 141 00:14:54,000 --> 00:15:01,000 I've really been interested in the use case of digital wills and inheritance plans, like a... 142 00:15:01,000 --> 00:15:10,000 What happens to this data or what happens to the digital parts of me when I'm gone? 143 00:15:10,000 --> 00:15:12,000 And like, who's gonna do what? 144 00:15:12,000 --> 00:15:13,000 You know, like there's... 145 00:15:13,000 --> 00:15:14,000 Yeah, I think. 146 00:15:14,000 --> 00:15:22,000 So there are also like, things like passwords, two-factor authentication codes and like backup codes that I think fit well into this. 147 00:15:22,000 --> 00:15:29,000 But really, conceptually and technically, you can have a vault of arbitrary size. 148 00:15:30,000 --> 00:15:35,000 WorkRx doesn't support arbitrary size yet, but... 149 00:15:35,000 --> 00:15:39,000 Yeah, but this is really cool because this is completely changing my interpretation of it. 150 00:15:39,000 --> 00:15:47,000 Because Dark Crystal, back in the day, the original project that both you and I knew of was specifically for keys, right? 151 00:15:47,000 --> 00:15:49,000 As we've been talking about up until now. 152 00:15:49,000 --> 00:15:53,000 But you're actually extending it to be really... 153 00:15:53,000 --> 00:15:56,000 What was the term you used? 154 00:15:56,000 --> 00:15:57,000 Sensitive data. 155 00:15:57,000 --> 00:16:00,000 Yeah, sensitive data. That's really cool. 156 00:16:00,000 --> 00:16:08,000 And it also makes more sense with the earlier discussion that I've decided to keep in before we started the interview. 157 00:16:08,000 --> 00:16:14,000 Which, where you talked about like the use case of the terms, like shards was very common. 158 00:16:14,000 --> 00:16:18,000 Way of talking about like the pieces of a private key. 159 00:16:18,000 --> 00:16:25,000 And now you said that you're calling it multiple keys, which of course, as you mentioned, has double connotation for people who are in cryptography. 160 00:16:25,000 --> 00:16:33,000 Is that that's kind of what you mean then, that you can kind of unlock a vault if we stick to the metaphor. 161 00:16:33,000 --> 00:16:36,000 Like, how does it work? Like, let's say, okay, let's walk through a case. 162 00:16:36,000 --> 00:16:42,000 Yeah, so I'm gonna walk through like my primary use case, to be honest. 163 00:16:42,000 --> 00:16:49,000 I have a group of friends and we have all had this shared problem for a while, which is that we all are home labbers, right? 164 00:16:49,000 --> 00:16:55,000 We have computers servers running at our house and we like to self host services. 165 00:16:55,000 --> 00:17:01,000 Like, I have my password manager on my home server, you know, I have all of our family photos backed up there. 166 00:17:01,000 --> 00:17:09,000 So we have kind of, we've mutually agreed, even before I started working on Horcrux, we had a mutual agreement that like, 167 00:17:09,000 --> 00:17:11,000 hey, anything happens to one of us. 168 00:17:11,000 --> 00:17:15,000 The others of us will help wind down, help each other's families and partners like, 169 00:17:15,000 --> 00:17:20,000 tear down these servers, move the services to places that make sense and take care of that. 170 00:17:20,000 --> 00:17:25,000 And so we're always like, yeah, it would be great if we had a tool to do this. 171 00:17:25,000 --> 00:17:30,000 Because we don't just want to get, you know, we could, like, I would trust them. 172 00:17:30,000 --> 00:17:35,000 I would give them the, you know, the root password to my Linux server. 173 00:17:35,000 --> 00:17:38,000 But they don't really want, want it. 174 00:17:38,000 --> 00:17:41,000 Like, they don't want it just sitting in their password manager. 175 00:17:41,000 --> 00:17:44,000 Nope. 176 00:17:45,000 --> 00:17:51,000 And so, and they don't want it just like on, yeah, like, I don't want to just put it in my safe or something, you know, 177 00:17:51,000 --> 00:17:54,000 bury it in the backyard on a, on a piece of paper. 178 00:17:54,000 --> 00:17:56,000 So, yeah. 179 00:17:56,000 --> 00:18:00,000 So the way this would work in Horcrux is that I download the app. 180 00:18:00,000 --> 00:18:04,000 I create a vault, I give it a name and then I, there's a text field, right? 181 00:18:04,000 --> 00:18:07,000 You know, that I can put in arbitrary text. 182 00:18:07,000 --> 00:18:11,000 So I might say, like, hey, yeah, my server is at this IP address. 183 00:18:11,000 --> 00:18:13,000 Here's, you know, it's, it's running. 184 00:18:13,000 --> 00:18:15,000 Here's where it's physically located in my house. 185 00:18:15,000 --> 00:18:19,000 Here's the, the root user name and password. 186 00:18:19,000 --> 00:18:24,000 And then on the server, I've got a, you know, maybe I got a document that lays out what's running here and what, 187 00:18:24,000 --> 00:18:30,000 what I think you should do with these services and where you should move this data or where the data is. 188 00:18:30,000 --> 00:18:35,000 And then Horcrux walks you through the process of inviting what I call stewards. 189 00:18:35,000 --> 00:18:41,000 So stewards are the people who are going to hold the keys to your vault and, and let you open it. 190 00:18:41,000 --> 00:18:48,000 So you get to decide how many keys to make and how many keys are required to open the vault. 191 00:18:48,000 --> 00:18:51,000 And then you give one key to each steward. 192 00:18:51,000 --> 00:18:58,000 So in my case, you know, I would, there are like three stewards that I would add and I'd probably set my threshold at two. 193 00:18:58,000 --> 00:19:02,000 So it, this vault takes two keys to open and I'd hold a key myself. 194 00:19:02,000 --> 00:19:09,000 So I make four keys. Horcrux gives you a little, you know, like invitation link that you send to your friends. 195 00:19:09,000 --> 00:19:17,000 They download Horcrux. They open it up and they get asked like, Hey, you know, Matt's asking you to steward to become a steward of his vault. 196 00:19:17,000 --> 00:19:24,000 Here's what that means. Do you agree? And if you say yes, then there's a little handshake between my own and yours. 197 00:19:24,000 --> 00:19:30,000 And, and then once all the stewards have joined, the data is divided up. 198 00:19:30,000 --> 00:19:36,000 So you can see the key is encrypted and the key is divided into four pieces. You can think of it. 199 00:19:36,000 --> 00:19:44,000 But in this special way, using some cool math where only two are required to reassemble the whole key and decrypt the vault. 200 00:19:44,000 --> 00:19:55,000 And then those keys are pushed out onto these relay servers and your stewards download them onto their devices, along with the encrypted vault payload. 201 00:19:55,000 --> 00:20:09,000 And then you're, then you're backed up. So then you're in a steady state. And when the time comes to recover, there's, there's a recovery, you know, wizard and process whereby somebody, any steward actually can initiate recovery, 202 00:20:09,000 --> 00:20:23,000 which I think this is another difference from, from Dark Crystal, but someone initiates recovery and then all this, every steward gets a push notification that says, Hey, this person's trying to open the vault. 203 00:20:23,000 --> 00:20:36,000 So you're requesting your key, including the main owner gets a push notification in case they're not like, and there's a section in there. You can, as the owner, you get to write instructions that everyone sees. 204 00:20:36,000 --> 00:20:52,000 So, you know, you might put something like before you, you know, open the vault, like, please get on a video call together, like with all the stewards or, you know, or, you know, meet, or, you know, check with my partner and like make sure that this is legit or something. 205 00:20:52,000 --> 00:21:11,000 Or, you know, only open if I am dead. So things like that. And so then each steward has to make that decision. You know, this is, you're trusting these people to, to honor your wishes in some sense, although you get some redundancy there, right? 206 00:21:12,000 --> 00:21:22,000 Like, if two people want to collude against you, but you set your threshold to three, then you're still safe. 207 00:21:22,000 --> 00:21:32,000 This is so great. Yeah, when, whenever start that recovery session gets all the keys back, then, then they're able to decrypt, decrypt the contents. 208 00:21:32,000 --> 00:21:42,000 First of all, immediate thing that pops into my mind is there was actually an issue. A heist was used that I talked about it with Kate, who was on the podcast earlier. 209 00:21:42,000 --> 00:21:50,000 And it was a heist in like Southeast Asia of like 50 million. I think it was Singaporean dollars or something. 210 00:21:51,000 --> 00:22:08,000 And it was done via a video call where a bunch of people had to confirm that this transfer was supposed to be done, but every person in the video call was a deep fake of the executives of the company. 211 00:22:08,000 --> 00:22:13,000 And I'm thinking like already there, like, Horcrux solves that issue. 212 00:22:14,000 --> 00:22:35,000 Like, this is already an issue of like, for example, for large transfers of money that has not been solved by even banks or like multinational corporations yet Horcrux is solving it in a way that improves the already existing social protocols at place. 213 00:22:36,000 --> 00:22:43,000 Because I guess even if you have the deep fakes, you would not have access to everyone's apps at the same time as having the deep fakes. 214 00:22:43,000 --> 00:22:45,000 For example, that's true. 215 00:22:45,000 --> 00:22:58,000 Steward impersonation can be a problem with Horcrux, but yeah, I think by itself, like in this sort of situation where the confirmation was consent on a video call, you're right. 216 00:22:59,000 --> 00:23:10,000 Like they would have also needed access to enough of the keys to the vault and the vault data as well as the consent of a bunch of parties. 217 00:23:10,000 --> 00:23:27,000 And yeah, as I've worked on the project, it's interesting you mentioned banks because somebody recently told me that like they knew of banks in France that actually do manage some of their cryptographic keys using Shamiya's secret sharing, which is the underlying algorithm. 218 00:23:27,000 --> 00:23:28,000 Yeah. 219 00:23:28,000 --> 00:23:31,000 So I need, I want to learn more about that, obviously. 220 00:23:31,000 --> 00:23:34,000 I haven't done my homework on that yet. 221 00:23:34,000 --> 00:23:46,000 But yeah, it's a, it's a common problem, I think, and it's only going to get more important maybe as we, as we continue marching into the information age with AI and. 222 00:23:46,000 --> 00:23:52,000 Yeah, all the impersonation methods that are unlocked there. 223 00:23:52,000 --> 00:23:53,000 Yay! 224 00:23:53,000 --> 00:23:55,000 Another reason to love AI. 225 00:23:55,000 --> 00:23:56,000 Sorry. 226 00:23:56,000 --> 00:23:57,000 I'm sorry. 227 00:23:57,000 --> 00:23:58,000 Sarcasm, of course. 228 00:23:58,000 --> 00:24:07,000 But the other thing that I said that I did not want to lose track of in my own brain was, oh yeah. 229 00:24:07,000 --> 00:24:19,000 So when we were at DWebCamp, we had a conversation because I was super thrilled because it's been a reoccurring conversation throughout my years of being in the peer to peer spaces or peer for peer spaces. 230 00:24:20,000 --> 00:24:28,000 And that, the challenge of like, what do you do when you lose your private key and, and Dark Crystal kept being brought up. 231 00:24:28,000 --> 00:24:48,000 And the reason I was told back in the day by someone, and I don't know where this started, it was from the grapevines, you know, was, I told you this at DWeb as well, was that the reason Dark Crystal stopped being used or stopped being developed was that they couldn't figure out the issue 232 00:24:48,000 --> 00:24:51,000 of how to get the keys back to the people who lost the keys. 233 00:24:51,000 --> 00:25:03,000 Because what it was then was like, you had a network that was most likely peer to peer and you wanted to send the keys back over the network that you had lost the keys to. 234 00:25:03,000 --> 00:25:07,000 So then if you have lost the keys, you can't get access to the shards anymore. 235 00:25:07,000 --> 00:25:12,000 And then how do you solve that without giving also emails or whatever. 236 00:25:13,000 --> 00:25:33,000 So how A, how have you solved that and B, you told me something really interesting at DWeb Camp last, which was that there's actually a whole other issue which has to do with a sense of responsibility and not trust, but yeah, you can fill us in on that one. 237 00:25:33,000 --> 00:25:44,000 But first off, how do you access the keys or the secrets or the sensitive data via the app? 238 00:25:44,000 --> 00:25:49,000 Like, do you log in normally? Like email, how does that work? 239 00:25:49,000 --> 00:26:02,000 I think to answer your first question, I have totally sidesteped the problem that Dark Crystal was having, or rather I think I'm solving a slightly different use case, where I have not tried to solve the problem. 240 00:26:02,000 --> 00:26:16,000 So I think to solve the problem of how do you, yeah, then get the key back to the original owner when the only method of communication of secure communication that you have is the protocol itself. 241 00:26:16,000 --> 00:26:19,000 Or the key that you're trying to restore. 242 00:26:19,000 --> 00:26:24,000 So I assume that you have out of band communication, like the invitation links. 243 00:26:24,000 --> 00:26:32,000 So you know, they're safe to like just text message, even unencrypted email to your stewards. 244 00:26:32,000 --> 00:26:41,000 So you leave it kind of agnostic. People can choose to communicate however they want outside of like the application itself. 245 00:26:41,000 --> 00:26:53,000 That's right. Yeah. And so, you know, that's a strength and a weakness, right? Like in some ways you're, then if you choose insecure methods of side channel communication, then you're decreasing the security of the overworld. 246 00:26:53,000 --> 00:27:06,000 But that's what I think I didn't see away. Like it would have been easy for me to add encrypted messaging into the app because because I've built on the nostril protocol, which which has support for that. 247 00:27:06,000 --> 00:27:16,000 But I chose not to for that same reason. If you're backing up your key that you use for the app, then yeah, you're presented with kind of this intractable problem. 248 00:27:16,000 --> 00:27:23,000 Is this the two generals problem? You can't authenticate the other person until you have a secure method of communication or something. 249 00:27:23,000 --> 00:27:30,000 I may be totally butchering that. I'm looking it up. The two generals problem is a computer science.exferium. 250 00:27:30,000 --> 00:27:39,000 Two generals trying to agree on this exact time to attack when they can only communicate by sending messages through an unreliable enemy held area. 251 00:27:39,000 --> 00:27:45,000 It seems like it. There's more to it. But something like that. Because messages and their confirmation is going to be good. 252 00:27:45,000 --> 00:27:55,000 Something like that. Cool. So maybe this dovetails into your into your other question, which is, you know, how is your how do you get into Horcrux? 253 00:27:55,000 --> 00:28:08,000 And the answer is with a key, of course. I already mentioned I'm using the nostril protocol. The nostril protocol uses uses public private keys. 254 00:28:08,000 --> 00:28:26,000 And I don't talk a lot about key management. So this is, I guess you could say an experiment that I'm doing where when you join the app, especially if you're a steward, like I do offer you the option to log in. 255 00:28:26,000 --> 00:28:36,000 I call them accounts even, but to log in with an existing account or create a new one. If you tap login, then it says some words about nostril. 256 00:28:36,000 --> 00:28:51,000 But I've tried not to make the protocol like part of the UX as much as I can avoid it. And so if you create a new account, it actually will then offer as your first Horcrux vault to back up your private key in a vault. 257 00:28:51,000 --> 00:29:05,000 And it walks you through walks you through that process. So can you then save your private key? Do you get prompted to like save your private key for your Horcrux account as soon as you create it? 258 00:29:05,000 --> 00:29:15,000 Like, do you know what I mean? Like, how does that work? Like if you know, I guess that then we're back at the two generals problem. If that's what it is. 259 00:29:15,000 --> 00:29:29,000 Yes, so the current flow is not ideal. And I know what I want to build and haven't had the time to build this flow nicely, like a nice UX into the app. 260 00:29:29,000 --> 00:29:41,000 So right now I think when you onboard, it asks you pretty early on, like, why are you here essentially? Like, are you here as a steward? Like, did somebody invite you? Are you here to create a vault? 261 00:29:41,000 --> 00:29:58,000 There's an option for, like, I lost my key to Horcrux. And what it currently does is it takes you to like a wall of text. That basically explains like if you backed up your key in Horcrux, you need to have a steward restore it. 262 00:29:58,000 --> 00:30:12,000 And then somehow securely send you the key. I'm thinking back on a conversation I had recently. I was talking to ADZ about Pyjapanda and more like the general concepts. 263 00:30:12,000 --> 00:30:21,000 And a nostril was mentioned there and specifically like the data types of nostril, which was different than a lot of other protocols. 264 00:30:22,000 --> 00:30:42,000 But I think my question is more general. Why did you choose the nostril protocol and how tied to the nostril protocol is it? Like, it sounds like it's quite agnostic. You could use it for pretty much sharding any, or not sharding. I'm using old terminology. 265 00:30:43,000 --> 00:30:48,000 We're pretty much distributing keys towards any sort of vault or a secret. 266 00:30:48,000 --> 00:31:11,000 Yeah. The main reason I chose nostril is because, like, sitting back and thinking about this as a tool, especially the digital will inheritance planning use case that's been my primary focus is one of the main features of this kind of system is that it should still work. 267 00:31:11,000 --> 00:31:25,000 In like 50 or 100 years, which is kind of a silly thing to think about because obviously technology is going to look much different, I'm sure, in 50 to 100 years. 268 00:31:25,000 --> 00:31:38,000 But if we were going to have some software that you could still run, and maybe more importantly, some data that you could still read and interact with in that sort of timeframe, it should be really simple. 269 00:31:38,000 --> 00:31:46,000 It should use as many standard, widely used conventions and protocols as possible. 270 00:31:46,000 --> 00:32:01,000 And that's sort of what nostril is. It's like, I think the founder called it the simplest decentralized protocol that could work, or censorship resistant protocol that could work or something like that. 271 00:32:02,000 --> 00:32:15,000 But, you know, nostril is essentially for those who understand these terms, it's JSON arbitrary, more or less arbitrary JSON signed by keys. 272 00:32:15,000 --> 00:32:21,000 And they're like, the keys format is the same as Bitcoin, which is like pretty popular. 273 00:32:22,000 --> 00:32:27,000 And JSON is like the most popular data format of the internet. 274 00:32:27,000 --> 00:32:41,000 And then the transmission of these signed JSON blobs are these really simple relay servers that are just, you know, store and forward servers. 275 00:32:41,000 --> 00:32:43,000 It's just a plain server on the internet. 276 00:32:43,000 --> 00:32:51,000 There's a small query language for like getting the events that you're interested in that you can implement in an afternoon. 277 00:32:51,000 --> 00:32:54,000 Like you could write your own relay server. 278 00:32:54,000 --> 00:33:03,000 And so, yeah, these pieces seemed appropriate for this type of tool. 279 00:33:03,000 --> 00:33:08,000 But to answer the second part of your question, it's very, it's pretty agnostic. 280 00:33:08,000 --> 00:33:16,000 Like, so the Horcrux protocol you could say, because Horcrux is open source, it's free. 281 00:33:16,000 --> 00:33:30,000 There's a draft sub spec, essentially, that details how the data model and the data transmission work, so that somebody else could write their own app or for Corecrux. 282 00:33:30,000 --> 00:33:34,000 And it would still be fully compatible with Horcrux. 283 00:33:34,000 --> 00:33:38,000 Like I could even own a vault and you could be a steward and you could be using a separate app. 284 00:33:38,000 --> 00:33:40,000 Like if we spoke the same language, right? 285 00:33:40,000 --> 00:33:56,000 So the Horcrux sub protocol is really also, I've tried to keep it a simple, stateless, not 100% stateless, but stateless system of exchanging data. 286 00:33:56,000 --> 00:34:06,000 And if I wanted to exchange that data over Iro or P2Panda or Willow, like it wouldn't be that hard. 287 00:34:06,000 --> 00:34:14,000 It's really the transport layer is completely separate from the data layer. 288 00:34:14,000 --> 00:34:25,000 And in fact, that's something I've been thinking a lot about because Horcrux is not really a local first app exactly because by default, 289 00:34:25,000 --> 00:34:31,000 it's going to take your data and put it on the nostril relay servers that you choose. 290 00:34:31,000 --> 00:34:34,000 And it's going to sit there for a while. 291 00:34:34,000 --> 00:34:45,000 And at the end of the day, like that's how all these P2P protocols work is like even Iro, you know, you're going through other people's computers. 292 00:34:45,000 --> 00:34:57,000 The point is that the data and the metadata as much as possible is hidden from those infrastructure providers passing the data to the people you actually want to get it to. 293 00:34:57,000 --> 00:35:07,000 But Horcrux is local first in the sense that if you get a signed piece of JSON into the app, it doesn't really matter how it got there. 294 00:35:08,000 --> 00:35:18,000 It is not relying on any like there's no privileged third party server or like piece of infrastructure that you need. 295 00:35:18,000 --> 00:35:20,000 Does that make sense? 296 00:35:20,000 --> 00:35:22,000 Yeah, I think so. 297 00:35:22,000 --> 00:35:24,000 There's one follow up here. 298 00:35:24,000 --> 00:35:30,000 I think I remember was it the data types that nostril uses? 299 00:35:30,000 --> 00:35:32,000 Is that like grow only sets? 300 00:35:33,000 --> 00:35:38,000 No, nostril doesn't use grow only sets. 301 00:35:38,000 --> 00:35:51,000 In fact, so the data model, the JSON is similar to very similar to the scuttlebutt data format, but it's missing the sequence number. 302 00:35:52,000 --> 00:36:13,000 So instead of all these signed JSON blobs being in a chain, they're just free, free radicals floating around the interwebs, which is kind of nostril strength and it's core weakness. 303 00:36:14,000 --> 00:36:15,000 Yeah. 304 00:36:15,000 --> 00:36:23,000 So I have yet to look into nostril so much because I heard all these bad rumors about it. 305 00:36:23,000 --> 00:36:30,000 But I'll ask you more into it after this interview or if this conversation. 306 00:36:30,000 --> 00:36:34,000 But I'm curious to hear a bit. 307 00:36:34,000 --> 00:36:39,000 I'm happy to talk about it on the partners in fact. 308 00:36:40,000 --> 00:36:53,000 Maybe I should because that's the community that has developed around, especially the public. 309 00:36:53,000 --> 00:36:59,000 You know, nostril is an acronym for notes and other stuff transmitted by relays. 310 00:36:59,000 --> 00:37:04,000 And notes is the name for the data type. 311 00:37:04,000 --> 00:37:07,000 It's essentially like a tweet, like a micro block. 312 00:37:07,000 --> 00:37:14,000 A public and public text that you've published that you want everyone to see. 313 00:37:14,000 --> 00:37:25,000 And so, yeah, so there's a community and a culture that's heavily populated by Bitcoin maximalists, which was not a term I was. 314 00:37:25,000 --> 00:37:29,000 I knew before I got on nostril. 315 00:37:29,000 --> 00:37:34,000 Is this anonymous with crypto rose? 316 00:37:34,000 --> 00:37:40,000 So they would be very upset if you called them crypto bros. 317 00:37:40,000 --> 00:37:49,000 Because I think what it means to be a Bitcoin maximalist is like Bitcoin is the only good cryptocurrency. 318 00:37:49,000 --> 00:37:54,000 Everything else is like some worse fork of Bitcoin. 319 00:37:54,000 --> 00:38:05,000 And Bitcoin is like the primary tool through which we can achieve the sorts of social revolutions that we need. 320 00:38:05,000 --> 00:38:14,000 Or that they want to see in the world, which when I say social revolutions, probably the things that come to mind are not. 321 00:38:15,000 --> 00:38:20,000 I'm hasn't yet unconvinced. 322 00:38:20,000 --> 00:38:26,000 But maybe I'm going to see the big picture. 323 00:38:26,000 --> 00:38:29,000 I am with you there. 324 00:38:29,000 --> 00:38:36,000 And so nostril is a community is problematic and has been problematic. 325 00:38:37,000 --> 00:38:47,000 It's widely acknowledged now even by the core community that their Bitcoin maximalism is an issue for nostril. 326 00:38:47,000 --> 00:38:51,000 Because the technology I think is very interesting. 327 00:38:51,000 --> 00:38:56,000 In many ways, it's the JavaScript of nerds. 328 00:38:56,000 --> 00:38:59,000 It's the JavaScript of decentralized protocols. 329 00:38:59,000 --> 00:39:01,000 JavaScript is a horrible programming language. 330 00:39:01,000 --> 00:39:03,000 And everyone knows it. 331 00:39:03,000 --> 00:39:10,000 And it's like, was made by someone, I guess the initial version was made in like a day or two or something. 332 00:39:10,000 --> 00:39:11,000 And it shows. 333 00:39:11,000 --> 00:39:16,000 And yet it's the most popular language in the world. 334 00:39:16,000 --> 00:39:28,000 So nostril is similar by glossing over a lot of the thornier issues that a lot of the more mature and strict protocols have. 335 00:39:28,000 --> 00:39:35,000 And it's allowed a high degree of experimentation and flexibility that I think is interesting. 336 00:39:35,000 --> 00:39:42,000 You know, I'm not here to say that like this is the best decentralized protocol or that it's going to take over the world. 337 00:39:42,000 --> 00:39:53,000 But I do think like it's, you know, for projects like Horcrux, I honestly think it's like the best protocol out there right now. 338 00:39:53,000 --> 00:39:56,000 I'm not going to debate you on that one. 339 00:39:56,000 --> 00:40:01,000 And I think you know best as the one who's developing the protocol. 340 00:40:01,000 --> 00:40:08,000 And if someone wants to debate you on that one, how do they reach out and get in touch? 341 00:40:08,000 --> 00:40:09,000 Yes. 342 00:40:09,000 --> 00:40:15,000 The easiest way is probably my personal website, mattlarens.com. 343 00:40:15,000 --> 00:40:22,000 And I have links there to my other, you know, my nostril profile, my mastodon, my blue sky. 344 00:40:22,000 --> 00:40:29,000 Is there a space like we're talking about this or is it like in the repo of Horcrux? 345 00:40:29,000 --> 00:40:36,000 I have not set up like a community space for this yet. No. No, I would like to though. 346 00:40:36,000 --> 00:40:42,000 That's a good idea. So I've just released 1.0 like a few weeks ago. 347 00:40:42,000 --> 00:40:44,000 And... 348 00:40:44,000 --> 00:40:46,000 Exciting! 349 00:40:46,000 --> 00:40:48,000 Yeah, congratulations. 350 00:40:48,000 --> 00:40:54,000 Thank you. It's been a huge milestone in a lot of ways. It feels like the beginning. 351 00:40:54,000 --> 00:41:04,000 Like the end of phase one. But there's so much work still to be done to make this... 352 00:41:04,000 --> 00:41:06,000 At the stage you'd like it to see? 353 00:41:06,000 --> 00:41:07,000 I want to say. 354 00:41:07,000 --> 00:41:13,000 Yeah. To help it live up to what it could be. 355 00:41:14,000 --> 00:41:23,000 Yeah. I mean, it sounds like it can be a lot. And it sounds like there's so many different now, like, implementation options for it. 356 00:41:23,000 --> 00:41:28,000 Which is sometimes a blessing and a curse, like knowing what to focus on. 357 00:41:28,000 --> 00:41:37,000 But it sounds like you've also focused in on like wills and things of that nature, which is a really interesting use case as well. 358 00:41:37,000 --> 00:41:40,000 And something that's unresolved as up to you yet. 359 00:41:40,000 --> 00:41:49,000 Yeah, I think one way that I think about it is it's a social layer for a bunch of systems that didn't have one before, right? 360 00:41:49,000 --> 00:41:59,000 Yeah. I'm super excited and keen to follow the project and the progress of workworks. 361 00:41:59,000 --> 00:42:05,000 So if you do end up creating a channel, I'll want to lurk. 362 00:42:05,000 --> 00:42:09,000 I don't know if I'll sign up on the nostril one, but... 363 00:42:09,000 --> 00:42:12,000 Exactly. Where would you like the channel to be? 364 00:42:15,000 --> 00:42:19,000 Tell me, where do we even gather these days? 365 00:42:19,000 --> 00:42:25,000 That's actually... I think that's one of the reasons I started this podcast, actually. 366 00:42:25,000 --> 00:42:30,000 Because we stopped having a space together in the same way. 367 00:42:30,000 --> 00:42:34,000 Like, there's a lot of different gathering points, I think, right now. 368 00:42:34,000 --> 00:42:41,000 But something that was as much of a hub as Scuttlebutt was back then, I haven't encountered. 369 00:42:41,000 --> 00:42:48,000 But I mean, there's a lot of group chats on DeltaChat that I'm seeing, and DeltaChat is moving in the peer-to-peer direction. 370 00:42:48,000 --> 00:42:52,000 Bonfire is pretty cool, but that's more federated. 371 00:42:52,000 --> 00:42:58,000 But there also, I heard talks about moving in a peer-to-peer direction. 372 00:42:58,000 --> 00:43:03,000 Let's see. I mean, Mastodon in general is a big space. 373 00:43:03,000 --> 00:43:10,000 And now there's a spacious, which we tried out, which might become something for more like... 374 00:43:10,000 --> 00:43:18,000 More akin to maybe Discord or like Asynchronous Voice Chat. 375 00:43:18,000 --> 00:43:21,000 Let's run over peer-to-panda. 376 00:43:21,000 --> 00:43:25,000 But yeah, honestly, I think there's a gap to be filled. 377 00:43:25,000 --> 00:43:27,000 I've been missing it myself. 378 00:43:27,000 --> 00:43:30,000 And I think in general, people are missing it right now. 379 00:43:30,000 --> 00:43:33,000 But that's for the peer-to-peer enthusiasts. 380 00:43:33,000 --> 00:43:41,000 I've been noticing that in general, like, people who are coming from like Facebook world or Instagram world 381 00:43:41,000 --> 00:43:48,000 are feeling similarly about Mastodon, as maybe people felt about like Scuttlebutt. 382 00:43:48,000 --> 00:43:54,000 But yeah, the prepper in me does not see it as enough to just be federated. 383 00:43:54,000 --> 00:43:58,000 Yeah. 384 00:43:58,000 --> 00:44:02,000 What do you think? 385 00:44:02,000 --> 00:44:06,000 Is that the sense... Have you got the same sense as me there? 386 00:44:06,000 --> 00:44:09,000 Yeah, 100%. 387 00:44:09,000 --> 00:44:14,000 But we'll wait. I feel like the protocol developments are going so fast right now. 388 00:44:14,000 --> 00:44:21,000 And things are happening at such a speed that I don't think is going to be long until there's... 389 00:44:21,000 --> 00:44:24,000 There's a new space for us to gather. 390 00:44:24,000 --> 00:44:27,000 If you hear something, let me know. 391 00:44:27,000 --> 00:44:29,000 I'll join in. 392 00:44:29,000 --> 00:44:32,000 I'll be listening here to find out. 393 00:44:32,000 --> 00:44:35,000 Alright, alright. 394 00:44:35,000 --> 00:44:37,000 We'll see, we'll see. 395 00:44:37,000 --> 00:44:40,000 But thank you so much for joining that. 396 00:44:40,000 --> 00:44:44,000 And it was so lovely meeting you in person for the first time. 397 00:44:44,000 --> 00:44:46,000 Yeah. 398 00:44:46,000 --> 00:44:52,000 And I really... Where do people follow what's been happening? 399 00:44:52,000 --> 00:44:56,000 If there's not a space to gather and talk about it, where do people follow? Updates. 400 00:44:56,000 --> 00:45:03,000 I've been posting updates on Blue Sky, Mastodon, and Nostor in Scuttlebutt. 401 00:45:03,000 --> 00:45:06,000 So find me on your preferred... 402 00:45:06,000 --> 00:45:12,000 Yeah, the major project updates are going on all four of those. 403 00:45:12,000 --> 00:45:15,000 Awesome! 404 00:45:15,000 --> 00:45:20,000 There's plenty of spaces to keep up to date then. 405 00:45:20,000 --> 00:45:24,000 And last but not least, and this is a bit of a boring question to end on, 406 00:45:24,000 --> 00:45:29,000 but I want to ask it because it's a tradition that I always ask about this, 407 00:45:29,000 --> 00:45:32,000 because it's a challenge everyone in the open source spaces dealing with, 408 00:45:32,000 --> 00:45:35,000 how are you funding this development? 409 00:45:35,000 --> 00:45:38,000 Yes. Great question. 410 00:45:38,000 --> 00:45:41,000 I received a grant a year ago. 411 00:45:41,000 --> 00:45:48,000 We were just finished actually from a foundation called OpenSats.org. 412 00:45:48,000 --> 00:45:56,000 And they are a private foundation that is mostly interested in funding open source Bitcoin tech, 413 00:45:56,000 --> 00:45:59,000 which I don't have. 414 00:45:59,000 --> 00:46:06,000 Horcrux has nothing to do with Bitcoin, but they have a like Nostor-focused fund. 415 00:46:06,000 --> 00:46:09,000 And that's who funded phase one. 416 00:46:10,000 --> 00:46:14,000 I'm looking for more funding right now actually. 417 00:46:14,000 --> 00:46:16,000 I'm planning to reapply with them. 418 00:46:16,000 --> 00:46:18,000 And I'm also looking for collaborators. 419 00:46:18,000 --> 00:46:21,000 Like, I don't want this to be just my thing. 420 00:46:21,000 --> 00:46:27,000 And I wish that I could have done phase one with a larger group of collaborators. 421 00:46:27,000 --> 00:46:32,000 But as I'm sure you know and have discussed, like, 422 00:46:32,000 --> 00:46:38,000 the shape of the funding absolutely influences the shape of these projects. 423 00:46:38,000 --> 00:46:45,000 And yeah, I'll also stay tuned for when we get that figured out. 424 00:46:45,000 --> 00:46:51,000 I heard some very interesting about some very interesting experiments at DWebCamp this year. 425 00:46:51,000 --> 00:46:52,000 Yeah. 426 00:46:52,000 --> 00:46:55,000 You know, people are always trying things. 427 00:46:55,000 --> 00:47:05,000 So I'm looking forward to the day when we can find ways to pool our resources and distribute them. 428 00:47:05,000 --> 00:47:06,000 Yeah. 429 00:47:06,000 --> 00:47:08,000 That more aligns with our values. 430 00:47:08,000 --> 00:47:11,000 Yeah, definitely. 431 00:47:11,000 --> 00:47:13,000 Very cool, very cool indeed. 432 00:47:13,000 --> 00:47:16,000 And that will be a conversation for the future. 433 00:47:16,000 --> 00:47:18,000 I don't know when and I don't know where. 434 00:47:18,000 --> 00:47:23,000 But until then, have a great one, Matt. 435 00:47:23,000 --> 00:47:25,000 And talk to you later. 436 00:47:25,000 --> 00:47:26,000 You too. 437 00:47:26,000 --> 00:47:27,000 Bye. 438 00:47:35,000 --> 00:47:38,000 So you can ask.