1 00:00:00,160 --> 00:00:02,360 All right, Sean, welcome to New York State Podcast. 2 00:00:02,680 --> 00:00:04,720 Glad to have you here. Great to be here. 3 00:00:04,720 --> 00:00:08,440 Thanks. So we worked together several 4 00:00:08,440 --> 00:00:14,440 years ago on what became USDC and I was the lead on the 5 00:00:14,440 --> 00:00:17,840 Coinbase side and we set up the center consortium and you and 6 00:00:17,840 --> 00:00:19,480 Jeremy were the Co founders of Circle. 7 00:00:19,480 --> 00:00:22,440 And I think the Circle Coinbase Center partnership has been 8 00:00:22,440 --> 00:00:26,400 pretty, pretty impressive for the world or last 7-8 years. 9 00:00:27,240 --> 00:00:29,600 Do you want to talk about that give just you know, introduce 10 00:00:29,600 --> 00:00:31,280 people because you know, you're a Co founder of Circle, you've 11 00:00:31,280 --> 00:00:33,160 done a bunch of other things. You can give the give the quick 12 00:00:33,160 --> 00:00:34,960 spiel Sean on Sean. Yeah, yeah. 13 00:00:34,960 --> 00:00:37,080 So Jeremy and I started Circle in 2013. 14 00:00:37,680 --> 00:00:41,200 So it's been, you know, 12 years, which is like 120 years 15 00:00:41,200 --> 00:00:44,280 in crypto or something. Yeah. 16 00:00:44,280 --> 00:00:46,920 And so, you know, we were, we were just obsessed with this 17 00:00:46,920 --> 00:00:51,520 idea about democratizing and sort of decentralizing global 18 00:00:51,520 --> 00:00:55,320 finance and, and you know, sort of have this vision for what 19 00:00:55,320 --> 00:00:58,000 that might look like. And the vision really hasn't 20 00:00:58,000 --> 00:01:00,360 wavered the way that we've tried to execute on the vision. 21 00:01:00,360 --> 00:01:03,880 And obviously it's had a lot of different sort of Criss crossing 22 00:01:03,880 --> 00:01:08,000 paths and probably should have been obvious in 2013 that we 23 00:01:08,000 --> 00:01:10,360 needed something like a stable coin. 24 00:01:10,360 --> 00:01:13,600 But it was actually 2017 when we figured out that that's what we 25 00:01:13,600 --> 00:01:16,200 needed to, that's what we needed in order to build everything 26 00:01:16,200 --> 00:01:21,320 else we were interested in. And, and obviously I was going 27 00:01:21,320 --> 00:01:23,840 to say, obviously you know that part of the story is as you came 28 00:01:23,840 --> 00:01:25,320 into the picture and helped make it happen as well. 29 00:01:25,520 --> 00:01:28,240 Yeah. And I think basically what's 30 00:01:28,240 --> 00:01:32,520 funny about that was that was, I think that was true teamwork. 31 00:01:32,840 --> 00:01:36,960 And what was interesting about stable coins at the time as you 32 00:01:37,080 --> 00:01:41,400 may remember is a lot of people poo pooed them and said, wow, is 33 00:01:41,400 --> 00:01:44,240 your great innovation is putting a dollar on chain. 34 00:01:44,240 --> 00:01:45,680 That's a great innovation, right? 35 00:01:46,160 --> 00:01:49,320 And it was kind of like people not understanding how big a deal 36 00:01:49,320 --> 00:01:52,880 it was to put like, for example, news on online, right? 37 00:01:53,200 --> 00:01:56,840 Or, you know, they, they just didn't understand that, for 38 00:01:56,840 --> 00:02:00,920 example, once it's on chain, you've got programmability and 39 00:02:00,920 --> 00:02:04,040 you've got, you know, anybody can send anybody and you can use 40 00:02:04,040 --> 00:02:06,680 it in smart contracts and you can slice and dice it down to 41 00:02:06,680 --> 00:02:10,400 smaller amounts and you can have devices using machine payments, 42 00:02:10,400 --> 00:02:12,920 as we'll get to with your new startup with the Catina. 43 00:02:13,600 --> 00:02:15,720 And people didn't understand that stuff at the time, but I 44 00:02:15,720 --> 00:02:19,240 think we, we saw that. And you know, what were you 45 00:02:19,240 --> 00:02:20,880 thinking about? You said the vision had changed 46 00:02:20,880 --> 00:02:23,240 for stables over time, but what were what was the number one 47 00:02:23,240 --> 00:02:25,040 thing that you guys were thinking about at the time you 48 00:02:26,120 --> 00:02:28,760 you started work on USCC? Yeah, I mean, it was pretty 49 00:02:28,760 --> 00:02:32,800 simple, honestly. In order to to enable global 50 00:02:32,800 --> 00:02:35,360 payments that didn't posit borders and that we're almost 51 00:02:35,360 --> 00:02:38,080 free, we needed some representation that was, that 52 00:02:38,080 --> 00:02:40,240 was stable and capable of moving over Internet rail. 53 00:02:40,240 --> 00:02:43,720 So the idea was multiple chains, not to not to be religious about 54 00:02:43,720 --> 00:02:47,120 anyone particular chain or piece of infrastructure, but to rely 55 00:02:47,120 --> 00:02:48,720 on block chains to deliver value. 56 00:02:49,160 --> 00:02:52,760 And that in in itself was really a building block so that we can 57 00:02:52,760 --> 00:02:56,840 enable these other use cases. You know, I'd say today stable 58 00:02:56,840 --> 00:02:59,360 coins, you know, largely the use case has been in crypto capital 59 00:02:59,360 --> 00:03:01,240 markets. We haven't yet seen that unlock 60 00:03:01,240 --> 00:03:04,360 of payments, treasury management effects, you know, these other 61 00:03:04,360 --> 00:03:05,640 things. But we're right on the cusp of 62 00:03:05,640 --> 00:03:08,640 it now. And you know, we certainly 63 00:03:08,640 --> 00:03:11,240 needed that building block technically, but also it 64 00:03:11,240 --> 00:03:15,000 couldn't just be a circle coin or a coin base coin. 65 00:03:15,040 --> 00:03:16,200 You know, That's right. That's what I say. 66 00:03:16,440 --> 00:03:20,440 USCC, if you recall, I think we named it so that the C was 67 00:03:20,720 --> 00:03:24,200 Circle and Coinbase and Centre because the Centre consortium, 68 00:03:24,440 --> 00:03:26,520 the consortium meant it was decentralized because we were 69 00:03:26,520 --> 00:03:30,080 partners on it and it also meant that others could in theory 70 00:03:30,080 --> 00:03:31,000 join. That's right. 71 00:03:31,080 --> 00:03:33,440 And that was the idea. We expected there to be 72 00:03:33,440 --> 00:03:37,360 multiple, you know, potentially multiple issuers, all of whom 73 00:03:37,360 --> 00:03:40,400 participated in a consortium, even though they may be 74 00:03:40,400 --> 00:03:42,200 competing at other layers of the stack. 75 00:03:43,280 --> 00:03:46,440 You know, it's the, the idea is just interoperable standards, 76 00:03:46,440 --> 00:03:47,960 right? So like back in the day, 77 00:03:47,960 --> 00:03:50,480 Microsoft had a slightly different version of HTTP than 78 00:03:50,480 --> 00:03:53,240 than that scape. And ultimately, that's 79 00:03:53,240 --> 00:03:55,400 interesting. Well, you know, there was this, 80 00:03:55,400 --> 00:03:58,680 there used to be, it was, it was sort of like, you know, build 81 00:03:58,680 --> 00:04:02,080 your service that ran in IE. You mean you mean do you mean 82 00:04:02,080 --> 00:04:04,480 like blink tags versus marquee like that? 83 00:04:04,480 --> 00:04:05,400 Kind of thing. Exactly. 84 00:04:05,400 --> 00:04:06,760 Yeah, yeah, yeah, yeah, yeah, yeah, yeah. 85 00:04:07,160 --> 00:04:09,440 But ultimately everybody agreed, let's just let's just agree on a 86 00:04:09,440 --> 00:04:12,360 standard unless I argue over different implementations to say 87 00:04:12,360 --> 00:04:16,000 SSL because we can compete at the e-commerce layer so long as 88 00:04:16,000 --> 00:04:18,959 we have this foundation of interoperable standards that no 89 00:04:18,959 --> 00:04:22,920 one vendor really owns. And when it comes to, you know, 90 00:04:22,920 --> 00:04:25,640 delivering dollars or other currencies but stable quints 91 00:04:25,640 --> 00:04:29,240 today, largely people largely want dollars today is what they 92 00:04:29,240 --> 00:04:34,200 mean when they say stables it it can't be circle dollars or 93 00:04:34,600 --> 00:04:35,680 strike point. Based dollars? 94 00:04:35,800 --> 00:04:37,280 Yeah, exactly. Point based dollars, whatever it 95 00:04:37,280 --> 00:04:39,560 is, it's just dollars to people who are using it. 96 00:04:39,560 --> 00:04:43,800 And so and so you know, we sort of saw a need for if dollars 97 00:04:43,800 --> 00:04:46,320 were going to move over Internet rails and over block chains, 98 00:04:46,320 --> 00:04:49,280 then then it needed to move as an interoperable standard. 99 00:04:49,280 --> 00:04:53,000 And so that was the idea is put together a consortium, not sort 100 00:04:53,000 --> 00:04:58,040 of Libra style of 30 people who can't agree on anything, but you 101 00:04:58,040 --> 00:05:02,000 know, a small number of very like minded motivated 102 00:05:02,000 --> 00:05:05,880 participants who could agree on these fundamentals and get it, 103 00:05:06,000 --> 00:05:08,040 you know, get it going. Yeah, you know, I like the Libra 104 00:05:08,040 --> 00:05:11,480 guys and I remember saying to them at the time and I think 105 00:05:11,520 --> 00:05:14,320 David Marcus and Morgan Beller and so on would would agree with 106 00:05:14,320 --> 00:05:16,360 me. Now, I guess it was it was 107 00:05:16,360 --> 00:05:17,600 totally fine. You know, they did their thing 108 00:05:17,600 --> 00:05:19,880 and I I give them props for seeing it through, even though, 109 00:05:19,920 --> 00:05:22,640 you know, it didn't work. Another thing works, but I think 110 00:05:22,640 --> 00:05:26,160 the issue with Libra at the time was a, they were using a basket 111 00:05:26,160 --> 00:05:28,160 of currencies to back their thing. 112 00:05:28,600 --> 00:05:32,640 And the problem with that is you had volatility without upside, 113 00:05:32,800 --> 00:05:34,520 right? You didn't know if it was like 114 00:05:34,520 --> 00:05:35,520 1. Libra is what? 115 00:05:35,520 --> 00:05:39,000 It's $0.87 a day and it's $1.15 tomorrow, right? 116 00:05:39,360 --> 00:05:42,000 And so it wasn't mapped anything that people had psychologically, 117 00:05:42,000 --> 00:05:43,040 right? No, no, no, fixed price. 118 00:05:43,040 --> 00:05:45,320 He had the other hand, they had volatility without upside 119 00:05:45,320 --> 00:05:47,160 because it's never going to be 10X, right? 120 00:05:47,360 --> 00:05:49,240 So it's like worst of both roles in that sense. 121 00:05:49,720 --> 00:05:51,560 And then there are other kinds of things that were like that as 122 00:05:51,560 --> 00:05:54,760 well where it was blockchain. So traditional banks thought it 123 00:05:54,760 --> 00:05:58,000 was risky, but it didn't have any upside. 124 00:05:58,000 --> 00:06:00,320 So crypto people didn't want to back it. 125 00:06:00,320 --> 00:06:03,120 And actually this was something which I think we did right with 126 00:06:03,120 --> 00:06:04,920 USDC. So I'll explain on the Coinbase 127 00:06:04,920 --> 00:06:06,680 and how we were thinking about somewhat similar to how you were 128 00:06:06,680 --> 00:06:10,160 thinking about it. So my view was because at that 129 00:06:10,160 --> 00:06:12,920 time, if you recall, we were in the ZERP era, right? 130 00:06:13,320 --> 00:06:16,120 And so we didn't know. I mean, only with the crystal 131 00:06:16,120 --> 00:06:18,280 ball would you know that they would Jack rates to the moon in 132 00:06:18,280 --> 00:06:22,000 2022, right? That was like 4-4 years out at 133 00:06:22,000 --> 00:06:24,760 that point. And we had been in ZERP for like 134 00:06:25,080 --> 00:06:28,480 almost 10 years, right, since the 2008 financial crisis, 135 00:06:28,480 --> 00:06:32,640 right? And so the USDC, it was not 136 00:06:32,640 --> 00:06:35,720 obvious that it was going to make any money anytime soon. 137 00:06:36,120 --> 00:06:38,760 And the way I conceptualize on the Courtney side is we thought 138 00:06:38,760 --> 00:06:40,880 about it similar to how you thought about it, but we thought 139 00:06:40,880 --> 00:06:44,440 about it as Google login. Like Google did not make money 140 00:06:44,480 --> 00:06:48,640 directly from Google login, but it's sort of projected, you 141 00:06:48,640 --> 00:06:52,880 know, like projected power, projected login over the whole 142 00:06:52,880 --> 00:06:54,040 Internet. And then people would come back 143 00:06:54,040 --> 00:06:56,800 to Google except a decentralized version of Google login where 144 00:06:56,800 --> 00:06:58,240 stable coins would be used everywhere. 145 00:06:58,400 --> 00:07:00,640 And then people would come back and they'd trade on our venues. 146 00:07:00,640 --> 00:07:02,760 They do things they'd basically be part of the crypto economy. 147 00:07:02,760 --> 00:07:05,480 So just like Google logins sort of grows the Internet, we 148 00:07:05,480 --> 00:07:07,320 thought USDC will grow the crypto economy. 149 00:07:07,400 --> 00:07:11,400 That was one kind of process. The second thesis is we, we put 150 00:07:11,400 --> 00:07:15,840 in capital behind it because we didn't think that. 151 00:07:16,600 --> 00:07:19,720 So had it been an ICO for example, there would have been 152 00:07:19,920 --> 00:07:21,440 pressure for a short term return. 153 00:07:21,680 --> 00:07:27,160 So we capitalized it, you know, together and I think that proved 154 00:07:27,160 --> 00:07:31,760 correct because had people gotten or wanted interest back 155 00:07:31,760 --> 00:07:35,520 from it, now today it throws off whatever billion dollars a year. 156 00:07:35,520 --> 00:07:38,840 It's like at 5% rates, 4% rates to 70 billion, it's like two $3 157 00:07:38,840 --> 00:07:40,880 billion a year. Not bad for seven years, right? 158 00:07:41,280 --> 00:07:45,320 But at the time it literally returned like nothing for like 159 00:07:45,320 --> 00:07:47,680 the first four years or something like that because 160 00:07:47,680 --> 00:07:50,880 interest rates were zero. So something like that is, you 161 00:07:50,880 --> 00:07:53,880 know, something where you can't necessarily make money on it 162 00:07:53,880 --> 00:07:56,800 because rates are not, you know, we're not with our control or, 163 00:07:56,800 --> 00:07:59,120 or we, we didn't know. So we had to essentially assume 164 00:07:59,120 --> 00:08:01,560 that it was just a utility thing for the entire ecosystem until 165 00:08:01,560 --> 00:08:03,920 the returns came there. Let me does that comport with 166 00:08:03,920 --> 00:08:05,080 how you you guys thought about it? 167 00:08:05,080 --> 00:08:06,040 Let me know your thoughts on that. 168 00:08:06,200 --> 00:08:10,200 Yeah, You know, the first version of it in 2017 actually 169 00:08:10,840 --> 00:08:13,920 predating Coinbase, we did have a, you know, it's probably still 170 00:08:13,920 --> 00:08:15,560 floating around out there. The first version of the white 171 00:08:15,560 --> 00:08:18,600 paper contemplated the token being released on the network 172 00:08:19,040 --> 00:08:22,920 that was called the Sent. And and so there was sort of 173 00:08:22,920 --> 00:08:26,560 this, this alternative path that we quickly realized this is not 174 00:08:26,560 --> 00:08:29,160 the right way to deliver this. But there was a little bit of a, 175 00:08:29,160 --> 00:08:32,960 a different revenue model tied to it by virtue of, you know, 176 00:08:32,960 --> 00:08:35,440 having a token. And this was, yeah, this was the 177 00:08:35,440 --> 00:08:37,400 ICO, you know, sort of boom and bust. 178 00:08:38,400 --> 00:08:43,240 And, you know, we explored a lot of things to try to deliver it 179 00:08:43,240 --> 00:08:45,560 in that fashion before we did V2. 180 00:08:46,160 --> 00:08:49,560 Yes, we realized that our. Collaboration and then we that's 181 00:08:49,600 --> 00:08:52,280 right. We, we, we had patient capital 182 00:08:52,280 --> 00:08:54,640 on the balance sheet. So we didn't need to we didn't 183 00:08:54,640 --> 00:08:58,640 have IC like nothing against coins, but those investors are 184 00:08:58,640 --> 00:09:01,000 not patient as you know, ICO investors are not patient. 185 00:09:01,320 --> 00:09:04,880 So, you know, and so by having patient capital on the balance 186 00:09:04,880 --> 00:09:07,960 sheet, we could go for years and years without, you know, 187 00:09:08,040 --> 00:09:10,160 worrying about any return in the short term. 188 00:09:10,160 --> 00:09:11,760 And, and I think that that helped a lot. 189 00:09:11,800 --> 00:09:17,640 OK, so now I'll tell you. There before before so you know, 190 00:09:17,640 --> 00:09:20,280 we had sort of decided well, Coinbase is the right partner. 191 00:09:20,280 --> 00:09:24,280 We were also talking to others. I mean, as you know, you know, 192 00:09:24,440 --> 00:09:27,920 and, and assisted with the like my people like Jim Megdal at 193 00:09:27,920 --> 00:09:30,560 Coinbase, we're trying to bring people into the central 194 00:09:30,560 --> 00:09:33,000 consortium. And there were very serious 195 00:09:33,000 --> 00:09:36,480 conversations we had before Coinbase that it didn't work 196 00:09:36,480 --> 00:09:40,560 out, but very large, you know, players when ultimately you, you 197 00:09:40,560 --> 00:09:42,880 leaned in and others at, at Coinbase. 198 00:09:43,160 --> 00:09:46,080 I always wondered internally, did Brian have to be convinced 199 00:09:46,600 --> 00:09:49,400 to to to go this path? It's a great question. 200 00:09:49,400 --> 00:09:51,720 So, so now, now the story can be told right. 201 00:09:51,720 --> 00:09:55,400 So it actually, so again, of course, you know, everything 202 00:09:55,400 --> 00:09:57,200 that happened on the Circle side, I can describe it happened 203 00:09:57,200 --> 00:10:01,760 on the on the Coinbase side. So there was essentially a big 204 00:10:01,760 --> 00:10:05,080 conversation internally, which was build versus buy versus 205 00:10:05,080 --> 00:10:11,600 partner, right? And I wanted to buy would have 206 00:10:11,600 --> 00:10:14,040 been like buy an expensive stable coin or something like 207 00:10:14,040 --> 00:10:16,160 that. And I also felt that wasn't 208 00:10:16,280 --> 00:10:19,560 really in tune with crypto because crypto is about 209 00:10:19,560 --> 00:10:21,280 decentralization and so on and so forth. 210 00:10:21,840 --> 00:10:25,080 And so I wanted to part first of all, partnering would have been 211 00:10:25,200 --> 00:10:28,000 you guys had my, my view on, on what you guys had. 212 00:10:28,000 --> 00:10:30,840 You had an audited smart contract at a time when that was 213 00:10:30,840 --> 00:10:33,120 actually a relatively very scarce thing, right? 214 00:10:33,120 --> 00:10:35,920 Because a trail of bits and what have you had? 215 00:10:36,360 --> 00:10:38,840 You know, they were, they're still, I think they're still 216 00:10:38,840 --> 00:10:41,440 good, but they were backlogged. And you know, there weren't that 217 00:10:41,440 --> 00:10:43,600 many people who knew how to do audited smart contracts at that 218 00:10:43,600 --> 00:10:44,720 time. So I knew that that would take 219 00:10:44,720 --> 00:10:49,400 us months to get to level that I felt that we could put hundreds 220 00:10:49,400 --> 00:10:50,960 of millions or billions of dollars into it, right. 221 00:10:50,960 --> 00:10:53,360 Like an audit of our contract is no, no small thing, right? 222 00:10:53,880 --> 00:10:55,520 There's that. And also, you know, the fact 223 00:10:55,520 --> 00:10:57,920 that you guys were reputable. You'd been around for a long 224 00:10:57,920 --> 00:11:00,920 time. Jeremy had had an exit before 225 00:11:00,920 --> 00:11:03,960 with the his video company. Gosh, I'm forgetting the name of 226 00:11:03,960 --> 00:11:06,200 it, but it was Brightcove. Brightcove. 227 00:11:06,200 --> 00:11:06,880 Exactly. That's right. 228 00:11:06,880 --> 00:11:08,320 Look, streaming stream video company. 229 00:11:08,600 --> 00:11:11,080 And actually we had met you. I don't know if you remember 230 00:11:11,080 --> 00:11:13,880 this when I was partnered Anderson all the way at the very 231 00:11:13,880 --> 00:11:17,280 beginning in 2013 when you guys were starting circle, you know, 232 00:11:17,400 --> 00:11:20,800 so, so we, so I, I, we had game filled on you, right? 233 00:11:20,800 --> 00:11:22,680 And, and obviously I think there's mutual respect. 234 00:11:24,960 --> 00:11:31,720 And then the so my view was a partner B user balance sheet. 235 00:11:32,040 --> 00:11:35,480 And essentially the question was partner versus build, right? 236 00:11:35,800 --> 00:11:38,440 And I felt that partner was the right approach because crypto is 237 00:11:38,520 --> 00:11:41,520 like I'm AI think I'd like to be or we'd like to be win, win 238 00:11:41,520 --> 00:11:43,600 people. And, and, and there's more that 239 00:11:43,600 --> 00:11:46,000 can be done together. And I felt that would be an 240 00:11:46,000 --> 00:11:48,840 organic partnership since you guys were also crypto first and 241 00:11:48,840 --> 00:11:50,880 you've been in the space for a long time and everybody who's in 242 00:11:50,880 --> 00:11:53,920 the class of 2013, we all have our Grays, right? 243 00:11:53,920 --> 00:11:56,120 But that was like the first time that a bunch of venture, you 244 00:11:56,120 --> 00:12:00,080 know, Eric Voorhees of shape shift really got his start back 245 00:12:00,080 --> 00:12:02,720 then, right? I think blockchain at info 246 00:12:02,760 --> 00:12:06,240 really got going. Then, you know, block stream got 247 00:12:06,240 --> 00:12:09,440 funded around that time. So I think quite quite a few 248 00:12:09,440 --> 00:12:11,720 companies, as you may recall, the Coindesk and so on really 249 00:12:11,720 --> 00:12:13,640 got going. So I think some of the really 250 00:12:13,640 --> 00:12:16,040 foundational companies in the space got going at that time. 251 00:12:16,320 --> 00:12:18,440 And what you'd been in the space for the right reasons. 252 00:12:18,440 --> 00:12:19,840 You weren't like a Johnny come lately. 253 00:12:19,840 --> 00:12:21,520 And I mean, it's all obvious stuff, but these are just 254 00:12:21,640 --> 00:12:23,440 explaining what, what our thinking was. 255 00:12:24,520 --> 00:12:27,480 So I felt you'd be a good long term partner for this. 256 00:12:28,200 --> 00:12:30,640 And so that's, that was part of the reasoning on, on our side. 257 00:12:30,640 --> 00:12:32,800 And then what I, what I did, which you, you guys don't care 258 00:12:32,800 --> 00:12:37,560 about, but we, we had just really three or four people on 259 00:12:37,560 --> 00:12:42,480 our side, Jim Mcdole, who you met, Jacob Horn, who's the 260 00:12:42,480 --> 00:12:46,600 product manager, and then Maxim and Mija, who are the two lead 261 00:12:46,600 --> 00:12:48,040 engineers. And then there's several other 262 00:12:48,040 --> 00:12:50,040 people, you know, in design. And so sort of those are, I 263 00:12:50,040 --> 00:12:52,000 think the four principal people. Maybe there's something I'm 264 00:12:52,000 --> 00:12:53,880 forgetting, but I think there's the four principles and the 265 00:12:53,880 --> 00:12:55,520 biggest, one of the hardest things was just clearing 266 00:12:55,520 --> 00:12:57,440 everything off of Maxim and Miha's schedule so they could 267 00:12:57,440 --> 00:13:01,160 just code 24/7 and it block all the HR people and so and so 268 00:13:01,160 --> 00:13:02,560 forth. So no, I just fill out their 269 00:13:02,560 --> 00:13:05,160 forms and we shipped it really fast. 270 00:13:05,920 --> 00:13:07,800 October 23rd, 2018. Go ahead. 271 00:13:08,120 --> 00:13:09,760 Yeah, yeah, I think that's right. 272 00:13:09,760 --> 00:13:11,800 And Jacob, by the way, also, I always felt he was a great 273 00:13:11,800 --> 00:13:15,640 champion for us internally at class and, you know, really, 274 00:13:15,640 --> 00:13:19,040 really enjoyed working with him. We had Joao Reganado, It was PM 275 00:13:19,040 --> 00:13:20,840 on our side, but our heat is really small, too. 276 00:13:21,000 --> 00:13:24,280 It was a handful of people. I mean, the, the tech, obviously 277 00:13:24,280 --> 00:13:27,320 it needed to be secure. It needed to be a strong 278 00:13:27,800 --> 00:13:29,200 actually we we launched on Ethereum. 279 00:13:29,200 --> 00:13:31,680 It's now in like 20 on everything. 280 00:13:32,080 --> 00:13:36,080 It's on everything, but it has to be secure into the, the code 281 00:13:36,080 --> 00:13:39,080 mattered a lot, but there's not a whole lot of tech there. 282 00:13:39,480 --> 00:13:43,400 I think the, the partnerships and the distribution, you know, 283 00:13:43,400 --> 00:13:46,120 obviously trying to, to clear the liquidity mode, which is 284 00:13:46,120 --> 00:13:48,880 still, you know, they're, they're only two, you know, 285 00:13:48,880 --> 00:13:51,400 dollar stables that have any meaningful liquidity right now, 286 00:13:51,520 --> 00:13:54,480 because that liquidity mode is really difficult to clear. 287 00:13:54,640 --> 00:13:58,000 Like all of these things were at least as important as, you know, 288 00:13:58,000 --> 00:13:59,720 the the pure engineering. That's right. 289 00:13:59,760 --> 00:14:04,920 And I think they're the reason on the engineering side, it 290 00:14:04,920 --> 00:14:07,880 wasn't so much, it was hard because it wasn't a lot of code, 291 00:14:08,160 --> 00:14:12,360 but it had to be extremely correct because, you know, smart 292 00:14:12,360 --> 00:14:14,360 cut, you know, re entrancy bugs, there's all kinds of these 293 00:14:14,600 --> 00:14:16,800 things that can get you. There are two early things that 294 00:14:16,800 --> 00:14:18,920 I remember at the time that really important. 295 00:14:19,400 --> 00:14:23,520 First was that a lot of other people had been doing stable 296 00:14:23,520 --> 00:14:26,040 coins backed by what I call plutonium, right? 297 00:14:26,320 --> 00:14:28,600 A very unstable thing over here. And you didn't know what the 298 00:14:28,600 --> 00:14:30,640 value is because it jumped up and down all the time. 299 00:14:30,960 --> 00:14:33,000 And nothing against Heather. They've proven themselves. 300 00:14:33,000 --> 00:14:35,480 They've executed over the last X number of years, right? 301 00:14:36,520 --> 00:14:40,000 But at the time it wasn't known that, you know, they're, that 302 00:14:40,000 --> 00:14:42,320 they're, they had enough paper to back what their outstanding 303 00:14:42,320 --> 00:14:44,160 assets were and so and so forth, right. 304 00:14:44,600 --> 00:14:47,120 And then there are other people, not even Tether, who are doing 305 00:14:47,120 --> 00:14:49,200 algorithmic table coins, some of which, as you know, you know, 306 00:14:49,200 --> 00:14:52,960 blew up later. And so one big thing for us was 307 00:14:52,960 --> 00:14:57,680 using the banking relationships to get $1.00 on chain back by 308 00:14:57,680 --> 00:15:00,480 $1.00 off chain and do it in a, in a dumb. 309 00:15:00,480 --> 00:15:06,080 But like, you know, it's simple is hard and, and good, right? 310 00:15:06,440 --> 00:15:10,240 Like, you know, so the simple thing of just one currency off 311 00:15:10,240 --> 00:15:12,120 chain, 1 currency on chain, that was one big piece. 312 00:15:13,400 --> 00:15:17,480 The second big piece, if if you may recall, is you know, and 313 00:15:17,480 --> 00:15:19,560 this is a big debate with those banks, right? 314 00:15:19,760 --> 00:15:25,360 Was would people have to KYC to send to somebody? 315 00:15:25,680 --> 00:15:29,960 Or would it be blacklist rather than whitelist, right? 316 00:15:29,960 --> 00:15:34,200 Meaning would we maintain so we had to have some compliance type 317 00:15:34,200 --> 00:15:36,000 stuff and so and so for somebody that misused it. 318 00:15:36,560 --> 00:15:41,520 So we eventually agreed that freeze and you know, in, in, in 319 00:15:41,520 --> 00:15:45,120 extreme circumstances, freeze or reverse transactions in, in the 320 00:15:45,120 --> 00:15:49,880 contract was better than having every single address have to pre 321 00:15:49,880 --> 00:15:52,240 verify in KYC beforehand, which would defeat the whole point of 322 00:15:52,240 --> 00:15:55,360 it, right? So once we, I think that was 323 00:15:55,560 --> 00:15:58,560 another key thing is we got enough banking partners to agree 324 00:15:59,280 --> 00:16:03,760 a, OK with a warrant you can freeze or, or seize or, or, or 325 00:16:03,760 --> 00:16:05,960 reverse transactions just like you can with wires. 326 00:16:07,040 --> 00:16:12,160 And B, you can also guard the, the egress and ingress at the 327 00:16:12,160 --> 00:16:14,920 exchanges when USDC is being swapped for USD, right? 328 00:16:15,480 --> 00:16:18,360 And so given that, then you otherwise have fun on chain and 329 00:16:18,360 --> 00:16:21,920 do what you need to do, right? And I think that actually now 330 00:16:21,920 --> 00:16:24,560 opened up the door for, among other things, something you and 331 00:16:24,560 --> 00:16:26,600 I have both thought a lot about kind of your new company, which 332 00:16:26,600 --> 00:16:30,520 is Cortana, right? Because by doing that, since, 333 00:16:31,160 --> 00:16:33,960 you know, if you had to KYC every address, you couldn't have 334 00:16:33,960 --> 00:16:38,880 a program generate 100 new addresses because they'd all 335 00:16:38,880 --> 00:16:41,960 have to be KYC by like the human and some photograph or something 336 00:16:41,960 --> 00:16:43,160 like that. Too much friction on being. 337 00:16:43,200 --> 00:16:44,720 Possible. That'd be impossible, right? 338 00:16:45,040 --> 00:16:48,320 So that brings us to the unlock of machine to machine payments. 339 00:16:48,320 --> 00:16:49,880 So why don't you talk about your new thing, Cortana? 340 00:16:50,000 --> 00:16:51,760 Yeah. So I I also say something on the 341 00:16:51,840 --> 00:16:55,800 the, the, the way that the sort of funds freeze happens. 342 00:16:55,800 --> 00:16:59,960 I was still never honestly, I I always felt like we could have 343 00:17:00,080 --> 00:17:05,400 we could have innovative there a bit more past just address, you 344 00:17:05,400 --> 00:17:08,720 know, management and you and I talked at the time about the 345 00:17:08,720 --> 00:17:10,920 things that were more sophisticated probably would not 346 00:17:10,920 --> 00:17:13,599 have been executable then, but this like governance. 347 00:17:13,599 --> 00:17:17,720 Contracts that like, yeah. Like risk and reputation on 348 00:17:17,720 --> 00:17:19,960 chain to, you know, sort of handle these these sorts of 349 00:17:19,960 --> 00:17:21,839 things. And it turned out that you're 350 00:17:21,839 --> 00:17:23,920 right. The simple approach was, was the 351 00:17:23,920 --> 00:17:27,240 one that was going to work, you know, But I still always felt 352 00:17:27,240 --> 00:17:29,240 like, you know, there's something there. 353 00:17:30,160 --> 00:17:32,560 And it's the, you know, one of the things that always nagged me 354 00:17:32,560 --> 00:17:35,400 about the Internet is there's no real like, like really, you 355 00:17:35,400 --> 00:17:39,120 know, distributed identity layer or people can sort of manage 356 00:17:39,120 --> 00:17:41,040 their own credentials effectively and, and not have 357 00:17:41,040 --> 00:17:43,360 to, to lean on a Google or, or whatever. 358 00:17:44,040 --> 00:17:46,280 And I, I was sort of like the remnants of that. 359 00:17:46,280 --> 00:17:49,800 I wanted to somehow figure out maybe how we could incorporate 360 00:17:49,800 --> 00:17:51,160 it. But at the end of the day, it 361 00:17:51,160 --> 00:17:54,080 was all about not even getting the regulators comfortable, but 362 00:17:54,080 --> 00:17:56,200 our de facto regulators, which are our bank partners. 363 00:17:57,160 --> 00:17:59,600 And this, the simple approach, was the one that was going to 364 00:17:59,600 --> 00:18:02,040 win there for both settlement and Reserve Bank partners. 365 00:18:02,120 --> 00:18:04,320 Yes. And, and I think that basically 366 00:18:05,280 --> 00:18:08,720 you know, like you, so I think today it's, it's possible to 367 00:18:08,720 --> 00:18:13,320 innovate on that, right. But you know, my, my view on a 368 00:18:13,320 --> 00:18:15,480 lot of things like this is sort of like minimum necessary 369 00:18:15,480 --> 00:18:18,760 innovation, right. So no coin, we just did off the 370 00:18:18,760 --> 00:18:22,200 balance sheet, no plutonium, we just did 1 to 1 backing, you 371 00:18:22,200 --> 00:18:26,440 know, USD for USCC, no coin governance. 372 00:18:26,440 --> 00:18:28,760 We just had rude access to the contract. 373 00:18:29,160 --> 00:18:31,480 We had the minimum size of the consortium of the game was just 374 00:18:31,480 --> 00:18:36,720 two and and we just minimized all complexity and was still not 375 00:18:36,720 --> 00:18:39,120 a completely trivial thing to scale and build this thing over 376 00:18:39,120 --> 00:18:41,920 the last few years. But but it was awesome. 377 00:18:42,000 --> 00:18:46,000 And you know, really, you know, it's funny, you do 20 things and 378 00:18:46,280 --> 00:18:48,920 it's interesting as to what things actually turn out to be 379 00:18:48,920 --> 00:18:52,280 really big. You know, it's kind of like for, 380 00:18:52,480 --> 00:18:55,640 you know, they're they're like 50 investments I do did in a 381 00:18:55,640 --> 00:18:56,840 year. And then one of them is like 382 00:18:56,840 --> 00:18:58,800 Ethereum or one of them is salon or something like that. 383 00:18:58,800 --> 00:19:00,800 And then it's like, oh, OK, well, that one, you know, that 384 00:19:00,840 --> 00:19:03,680 goes really big, right? And it it's interesting because 385 00:19:04,320 --> 00:19:10,920 a year later or two years later USTC was useful, but it is it is 386 00:19:10,960 --> 00:19:15,080 after I think Zerp ended that it really went boom like this and 387 00:19:15,080 --> 00:19:17,080 became as material as it is now, right. 388 00:19:17,320 --> 00:19:18,520 And it's funny. So I'll tell you another 389 00:19:18,520 --> 00:19:22,600 discussion that not public, but I think, you know, there was a 390 00:19:22,600 --> 00:19:26,120 discussion internally as to how much to out, you know, like 391 00:19:26,120 --> 00:19:28,280 resources to allocate towards USDC. 392 00:19:29,120 --> 00:19:32,880 And unusually see, most of the time you know, good advice of a 393 00:19:32,880 --> 00:19:39,200 company is like ignore macro. But in this case, you couldn't 394 00:19:39,200 --> 00:19:42,320 ignore macro because 100% decision was what is the Fed 395 00:19:42,320 --> 00:19:47,440 going to do, right? Because it's like, is it going 396 00:19:47,440 --> 00:19:49,600 to increase rates or is it going to keep them down or is it going 397 00:19:49,600 --> 00:19:51,880 to go wiggle like this or what? What is it going to do? 398 00:19:52,280 --> 00:19:55,400 And like, it felt actually it was much more Wall Street than 399 00:19:55,400 --> 00:19:59,520 I'm used to in tech because completely 100% of the decision 400 00:19:59,520 --> 00:20:01,720 was your mental model of the Fed. 401 00:20:02,040 --> 00:20:03,720 I I don't know if you have any thoughts on that, then we can go 402 00:20:03,720 --> 00:20:05,560 to Cantina over there. Yeah, yeah, yeah, yeah, yeah. 403 00:20:05,560 --> 00:20:07,400 No, that's true. I mean, I think that my 404 00:20:07,400 --> 00:20:12,280 recollection is that even after we had, we had decided on most 405 00:20:12,280 --> 00:20:15,400 of the structural pieces in terms of governance and and 406 00:20:15,400 --> 00:20:20,200 attack and we had, you know, lined up the banking partners. 407 00:20:20,400 --> 00:20:22,160 And this was also Speaking of banking partners. 408 00:20:22,160 --> 00:20:27,080 This was also just barely after the era when there's only really 409 00:20:27,080 --> 00:20:29,800 one bank in the world that would bank Coinbase or Circle. 410 00:20:30,160 --> 00:20:31,200 Yeah. So the making partner piece, 411 00:20:31,200 --> 00:20:34,880 which you mentioned was it was a huge negotiation and discussion 412 00:20:34,880 --> 00:20:37,440 because things are different now. 413 00:20:37,960 --> 00:20:40,360 You know, we have, you know, Bony Melon and BlackRock and you 414 00:20:40,360 --> 00:20:46,440 know, you know, but it was not like that then a short time ago. 415 00:20:46,440 --> 00:20:47,280 So. So yeah. 416 00:20:47,280 --> 00:20:51,200 But my recollection is the the the most, I would say difficult 417 00:20:51,200 --> 00:20:53,680 but time consuming hardest pieces to put together were the 418 00:20:53,680 --> 00:20:57,480 economics. It was the macro situation, a 419 00:20:57,480 --> 00:21:00,000 sort of analysis, but also micro. 420 00:21:00,000 --> 00:21:03,800 Is this what what are the what should the economics be if we 421 00:21:03,800 --> 00:21:07,360 have a consortium not of 100 people, but of a small number 422 00:21:07,360 --> 00:21:10,280 starting with two, What what should what does that look like? 423 00:21:10,280 --> 00:21:12,960 And there's a lot of discussion. It's been revised, you know, 424 00:21:13,200 --> 00:21:15,600 since since our time, you know, focused on that. 425 00:21:16,040 --> 00:21:19,000 But but yeah, absolutely. I think that that was the big 426 00:21:19,000 --> 00:21:20,880 puzzle. And and even then I remember we 427 00:21:20,880 --> 00:21:23,360 had people talking about, you know, maybe can you do JPY 428 00:21:23,360 --> 00:21:26,440 stable coin or like negative interest rates or this doesn't 429 00:21:26,760 --> 00:21:28,080 you? Know all that is now happening, 430 00:21:28,960 --> 00:21:32,520 Yeah. Absolutely, yeah, Yeah, so, so 431 00:21:32,520 --> 00:21:34,800 that's right. So, so you asked about Capena 432 00:21:34,800 --> 00:21:37,240 and I keep going back to USDC because it's fun to chat with 433 00:21:37,240 --> 00:21:38,960 you about it. But you know, one of the one of 434 00:21:38,960 --> 00:21:42,760 the nice things about about machine native money is that 435 00:21:42,760 --> 00:21:45,440 it's it's very reliable for machines to use it. 436 00:21:46,360 --> 00:21:49,200 So, so when we have these orchestrated workflows called a 437 00:21:49,200 --> 00:21:51,280 mate and no one can agree on what an agent is, but sort of 438 00:21:51,920 --> 00:21:54,600 workflows with alolim intelligence involved in making 439 00:21:54,600 --> 00:21:59,080 decisions, then as those workflows become economic 440 00:21:59,080 --> 00:22:01,800 actors, they need to be able to transact. 441 00:22:01,800 --> 00:22:03,040 They need to be able to hold money. 442 00:22:03,040 --> 00:22:04,280 They need to be able to send money. 443 00:22:04,280 --> 00:22:06,280 They need to be able to make, you know, treasury effects 444 00:22:06,400 --> 00:22:10,760 decisions, potentially with, you know, all within guardrails and 445 00:22:10,920 --> 00:22:12,320 humans in the loop where necessary. 446 00:22:12,800 --> 00:22:17,200 But simply tacking AI intelligence on to the edges of 447 00:22:17,200 --> 00:22:22,000 say a credit card flow is just aside from the economics, it is 448 00:22:22,000 --> 00:22:24,120 not reliable. It's really difficult to make 449 00:22:24,120 --> 00:22:27,480 those even the just the simple tool calls very reliable. 450 00:22:28,440 --> 00:22:32,080 And, and so you, when you take a 5 to 7 to 9 party system and you 451 00:22:32,080 --> 00:22:35,640 add a couple more parties on the edges of it, then, you know, 452 00:22:35,640 --> 00:22:39,120 it's different than taking a, a stable coin, you know, point to 453 00:22:39,120 --> 00:22:42,960 point transaction and, and you know, gentech work flows were 454 00:22:42,960 --> 00:22:44,840 very good at signing cryptographic messages. 455 00:22:45,720 --> 00:22:49,120 And, and so, and so, and then all of a sudden, you know, you 456 00:22:49,120 --> 00:22:51,720 sort of unlock the capability of transacting in fractional 457 00:22:51,720 --> 00:22:54,680 amounts in like streaming payments in real time. 458 00:22:54,680 --> 00:22:57,120 All these things we've been talking about for 20 years that 459 00:22:57,120 --> 00:22:58,880 have really never been economically possible or 460 00:22:58,880 --> 00:23:02,480 technically possible until we have the combination of machines 461 00:23:02,480 --> 00:23:05,760 becoming economic actors and stable coins. 462 00:23:05,800 --> 00:23:07,880 I don't know if we keep calling them stable coins, but it's a, 463 00:23:07,960 --> 00:23:12,120 you know, money. Yes, that's right. 464 00:23:12,120 --> 00:23:15,680 Well, so because this is funny because, you know, 10 years ago, 465 00:23:15,840 --> 00:23:19,480 I also took a crack at the machine payable web and machine 466 00:23:19,480 --> 00:23:24,320 micro payments and 402. And the issue was that Bitcoin 467 00:23:24,320 --> 00:23:28,640 at the time, see Gavin Anderson, if you recall, had published a 468 00:23:28,640 --> 00:23:31,520 road map for scaling to huge numbers of transactions of big 469 00:23:31,520 --> 00:23:33,680 blocks and so on. And then the Bitcoin civil war 470 00:23:33,680 --> 00:23:36,920 happened and became small blocks and fine, but that just 471 00:23:36,920 --> 00:23:39,560 basically meant that you couldn't scale on Bitcoin in the 472 00:23:39,560 --> 00:23:41,160 way that, you know, lots of transactions. 473 00:23:41,160 --> 00:23:43,000 So everything had to move to other chains. 474 00:23:43,440 --> 00:23:47,400 And so, and, and frankly, you know, it's funny now today in 475 00:23:47,400 --> 00:23:52,760 retrospect, now that we have Solana and Bass and we have USDC 476 00:23:53,240 --> 00:23:55,800 and we had to have Ethereum and we had to have Bitcoin and we 477 00:23:55,800 --> 00:24:00,040 also had to like win this gigantic political battle, right 478 00:24:00,440 --> 00:24:04,760 after all of that. Now finally a lots of the pent 479 00:24:04,760 --> 00:24:07,120 up innovation and also we have to be able to get on phones, 480 00:24:07,240 --> 00:24:09,880 right, because you have to win this political battle because 481 00:24:10,160 --> 00:24:12,800 this is something, it's a really interesting thing, which you 482 00:24:12,800 --> 00:24:16,200 know, and I know so many people for some years, like where are 483 00:24:16,200 --> 00:24:19,080 the crypto apps? Why aren't they on phones? 484 00:24:19,840 --> 00:24:23,280 And the reason is that Apple especially, and to a lesser 485 00:24:23,280 --> 00:24:26,760 extent, for real extent Android would nerf all kinds of apps 486 00:24:26,760 --> 00:24:30,800 that use crypto on phones. Because there was some like 487 00:24:30,800 --> 00:24:34,240 Apple payment, you know, the IAP thing, in app purchases thing, 488 00:24:34,240 --> 00:24:36,320 all this kind of stuff where they've now actually lost some 489 00:24:36,320 --> 00:24:38,600 cases on that, you know, and they're sort of being forced to 490 00:24:38,600 --> 00:24:40,440 do that. And to be clear, look, Apple 491 00:24:40,600 --> 00:24:44,040 does a lot of stuff well, but they're sort of genetically anti 492 00:24:44,040 --> 00:24:46,360 crypto. And so there's a chicken and egg 493 00:24:46,360 --> 00:24:47,880 thing where you just couldn't get. 494 00:24:47,880 --> 00:24:50,240 And that's why crypto is actually developed as much more 495 00:24:50,240 --> 00:24:52,080 of a web phenomenon than a mobile phenomenon. 496 00:24:52,080 --> 00:24:54,840 We have to be choke point in a certain way where we're only web 497 00:24:54,840 --> 00:24:57,080 rather than mobile, and we have to be choke pointed through the 498 00:24:57,080 --> 00:25:00,240 banks in a certain way. And like you couldn't do equity 499 00:25:00,240 --> 00:25:01,840 issuance, you can only do meme coins. 500 00:25:01,960 --> 00:25:04,840 So this I think bank list also had a very similar observation. 501 00:25:04,840 --> 00:25:07,240 Like there's all this pent up spring of innovation go Boeing 502 00:25:07,240 --> 00:25:11,200 like this over the next 10 years that that that machine payable 503 00:25:11,560 --> 00:25:12,920 machine payments are part of. Go ahead. 504 00:25:13,600 --> 00:25:17,120 Yeah, yeah, absolutely. I mean, my thesis is that in the 505 00:25:17,120 --> 00:25:22,760 future, the only actors that we will trust with our money and 506 00:25:22,760 --> 00:25:26,000 our assets and the only actors that will be capable of of 507 00:25:26,000 --> 00:25:29,960 generating competitive returns will be agentic. 508 00:25:30,480 --> 00:25:32,680 Interesting. And and and I. 509 00:25:32,720 --> 00:25:35,280 Only play the agentic or do you mean delegates of yourself? 510 00:25:36,000 --> 00:25:37,440 So this isn't, that's a good question. 511 00:25:37,720 --> 00:25:40,920 So right now there's a little bit of a of a battle, I would 512 00:25:40,920 --> 00:25:45,360 say, between, you know, those who believe that agents will 513 00:25:45,360 --> 00:25:48,280 always be, should be mapped to tasks that always get mapped to 514 00:25:48,280 --> 00:25:50,480 some user to be, you know, business or consumer or 515 00:25:50,480 --> 00:25:52,680 whatever. And those, those that believe 516 00:25:52,840 --> 00:25:56,120 that may be true, but we'll also see semi autonomous or 517 00:25:56,120 --> 00:25:58,280 autonomous agents that actually do have their own identity and 518 00:25:58,280 --> 00:26:01,520 act on their own on, on not just a planning loop that can truly 519 00:26:01,520 --> 00:26:05,440 act on their own behalf. And so there's debates on on 520 00:26:05,440 --> 00:26:08,840 both sides of those things, but I, I think, you know, either 521 00:26:09,000 --> 00:26:14,600 either way, you know, in terms of being able to make good 522 00:26:14,600 --> 00:26:19,920 decisions rapidly enough with, with money, we will just not 523 00:26:19,920 --> 00:26:21,360 want to trust anything other than. 524 00:26:21,360 --> 00:26:23,880 I mean, it is hard to imagine now because people's anecdotal 525 00:26:23,880 --> 00:26:26,840 experiences are that, you know, Claude or ChatGPT can't give 526 00:26:26,840 --> 00:26:30,120 them a, a, you know, recipe for a brownie correctly because of 527 00:26:30,120 --> 00:26:34,320 the, the sort of hallucinations. But it, but it is, I still have 528 00:26:34,320 --> 00:26:37,480 strong conviction that it is the truth that we will see AI actors 529 00:26:37,480 --> 00:26:39,840 that will be the most effective economic participants the 530 00:26:39,840 --> 00:26:41,960 world's ever seen. And then we will want to, we use 531 00:26:41,960 --> 00:26:44,960 those, but we will have no need to ever execute a transaction 532 00:26:44,960 --> 00:26:46,520 ourselves. We will always be doing it 533 00:26:46,520 --> 00:26:50,440 through some form of, you know, personal or or other AI. 534 00:26:51,400 --> 00:26:53,800 Interesting. You don't think you'll ever need 535 00:26:53,800 --> 00:26:55,240 to execute a transaction yourself? 536 00:26:55,680 --> 00:26:57,840 That's interesting. That's a strong form version. 537 00:26:59,600 --> 00:27:02,560 So where do you? When that happens when that 538 00:27:02,560 --> 00:27:04,880 happens is a is a question. I always find that when the 539 00:27:04,880 --> 00:27:08,040 hardest question for me to answer in this so I have 540 00:27:08,040 --> 00:27:11,920 conviction around the what is it take three years to touch 541 00:27:11,920 --> 00:27:15,360 certain domains before others? Or does it take two months? 542 00:27:15,360 --> 00:27:17,840 Does it take 20 years? It's really difficult to say. 543 00:27:18,320 --> 00:27:22,240 So 12 years ago, actually me and actually the Winklevoss's and 544 00:27:22,240 --> 00:27:25,320 Neville, there's some panel and I remember thinking about 545 00:27:25,320 --> 00:27:28,480 machine payments then. And my kind of example of 546 00:27:28,480 --> 00:27:31,360 something Bitcoin could do that, you know, other currencies like 547 00:27:31,360 --> 00:27:34,880 Fiat couldn't do is if you had two self driving cars on the 548 00:27:34,880 --> 00:27:38,960 highway, one of them could pay the other to pass, for example, 549 00:27:39,000 --> 00:27:42,480 right? And now that's just kind of a 550 00:27:42,480 --> 00:27:46,040 made-up example, but the general concept is machines could 551 00:27:46,040 --> 00:27:49,120 negotiate prioritization amongst themselves or what have you, 552 00:27:49,120 --> 00:27:51,120 right? In practice, by the way, self 553 00:27:51,120 --> 00:27:52,960 driving cars might just drive fast enough that they could 554 00:27:52,960 --> 00:27:56,880 just, you know, boom, snap, snap onto each other and and and go, 555 00:27:56,880 --> 00:27:59,200 right. But there's probably something 556 00:27:59,200 --> 00:28:03,200 to that of machines being able to negotiate back and forth. 557 00:28:03,400 --> 00:28:08,160 You know, for example, you could have a fully, you know, right 558 00:28:08,160 --> 00:28:11,480 now, Waymo and and Tesla robotaxis just to drop off and 559 00:28:11,480 --> 00:28:13,000 pick up. Finally, we have truly fully 560 00:28:13,000 --> 00:28:16,760 self driving end to end cars. But you could have something 561 00:28:16,760 --> 00:28:20,360 load food into it and then unload the food on their side, 562 00:28:20,360 --> 00:28:24,160 like sidewalk robots exist, right, and drive thrus exist. 563 00:28:24,360 --> 00:28:27,920 And you could have something where like the car pulls up, it 564 00:28:27,920 --> 00:28:30,080 pays the robot, it it gets the food. 565 00:28:30,080 --> 00:28:31,200 You know what I mean? Right. 566 00:28:31,440 --> 00:28:32,960 So that's an example of machine to machine. 567 00:28:33,160 --> 00:28:35,760 I don't know if now the thing about that is that's actually in 568 00:28:35,760 --> 00:28:37,920 the physical real world, which is always hard because there's 569 00:28:37,920 --> 00:28:40,120 other kinds of issues that Wi-Fi drop out, the Bluetooth drop. 570 00:28:40,360 --> 00:28:42,440 So I wanted to know where you thought the first two or three 571 00:28:42,440 --> 00:28:46,000 applications of machine payments would be for a. 572 00:28:46,000 --> 00:28:49,160 Catina Yeah. So I'll see what we're seeing 573 00:28:49,160 --> 00:28:51,400 today. What we're not seeing is true 574 00:28:51,400 --> 00:28:53,520 agent to agent or machine to machine payments. 575 00:28:54,680 --> 00:28:58,360 We're seeing AI workflows pay for access for, to resources, 576 00:28:58,600 --> 00:29:01,760 content, data, paying for, you know, access to APIs. 577 00:29:01,880 --> 00:29:04,800 There may be another agent on the other side of an API, but 578 00:29:04,800 --> 00:29:06,960 the interface is not truly agent to agent yet. 579 00:29:07,760 --> 00:29:09,560 And I think there are a lot of reasons for that that have 580 00:29:09,560 --> 00:29:10,920 nothing to do with the payment side. 581 00:29:11,600 --> 00:29:13,720 You know, there's no DNS for agents that we haven't really 582 00:29:13,720 --> 00:29:15,520 figured out the right way to handle discovery. 583 00:29:18,040 --> 00:29:20,040 You know, if you're, if you, if you're, if your agent 584 00:29:20,040 --> 00:29:22,160 interfaces. Do talk about text fields, I 585 00:29:22,160 --> 00:29:24,680 guess. But yeah, not really, because 586 00:29:24,920 --> 00:29:26,600 agents don't have persistent domains either. 587 00:29:26,600 --> 00:29:28,720 They can't, Yeah. I mean, they kind of do, but not 588 00:29:28,720 --> 00:29:30,320 really. There are proposals for like 589 00:29:30,320 --> 00:29:32,080 registry concepts. So Google has, for you know, 590 00:29:32,080 --> 00:29:35,120 proposed something for a registry concept attached to 591 00:29:35,120 --> 00:29:36,560 their A to a framework. Maybe. 592 00:29:36,640 --> 00:29:38,880 Maybe there isn't. Really like a standard for it. 593 00:29:39,280 --> 00:29:43,040 Maybe you could use ENS or SNS, the Solana name system for doing 594 00:29:43,040 --> 00:29:44,840 that. Yeah, because that's a little 595 00:29:44,840 --> 00:29:46,520 more flexible than DNS. Yeah, you can I. 596 00:29:46,520 --> 00:29:48,760 Think we may see a little bit of fragmented approach, whether it 597 00:29:48,760 --> 00:29:50,920 be, you know, multiple things that will need to be supported. 598 00:29:51,080 --> 00:29:53,520 But in the meantime, it's sort of like, you know, if you and I 599 00:29:53,520 --> 00:29:56,600 go build a website, we get a domain and we just deploy it or, 600 00:29:56,800 --> 00:29:59,200 you know, there are lots of ways for it to be discovered in the 601 00:29:59,200 --> 00:30:00,320 world. And there aren't really there's 602 00:30:00,320 --> 00:30:03,280 not really that equivalent for agents working outside of their 603 00:30:03,280 --> 00:30:05,280 ecosystems. And even when people build 604 00:30:05,560 --> 00:30:07,840 agents, it's usually in a particular framework or in a 605 00:30:07,840 --> 00:30:10,560 particular ecosystem. It's and it talks to other 606 00:30:10,560 --> 00:30:14,600 agents in that same ecosystem, but it doesn't really span the 607 00:30:14,600 --> 00:30:16,240 ecosystem. So if you build an agent that 608 00:30:16,240 --> 00:30:21,240 you deploy, say you can build it and say you know a framework 609 00:30:21,240 --> 00:30:23,720 like Lane graph. How does it make it self 610 00:30:23,720 --> 00:30:27,120 discoverable outside of like a chat interface to an agent 611 00:30:27,120 --> 00:30:28,760 deployed in like Salesforce or something? 612 00:30:28,760 --> 00:30:33,000 So, so you know, there are smart contracts that are something, 613 00:30:33,000 --> 00:30:36,760 something dot ETH, right. And so that, I mean, basically 614 00:30:36,760 --> 00:30:39,880 the smart contract and the agent are the same thing potentially 615 00:30:39,880 --> 00:30:45,000 and you just have an ENS name or an SNS name and that is that is 616 00:30:45,000 --> 00:30:47,240 the registry. But but I like your framing of 617 00:30:47,240 --> 00:30:51,600 it, which is smart contract registries and agent registries 618 00:30:51,600 --> 00:30:54,400 should be the same thing. And that there should be like, I 619 00:30:54,400 --> 00:30:56,960 think that itself is a business in its own right to just 620 00:30:56,960 --> 00:30:58,680 register and discover agents, you know? 621 00:30:58,880 --> 00:31:02,600 Yeah, I think that there's probably a, there's some form 622 00:31:02,600 --> 00:31:05,640 of, you know, Verisign for agents company that wants to be 623 00:31:05,640 --> 00:31:07,920 built. And just verification on them 624 00:31:07,920 --> 00:31:09,720 too, yeah, because they could just take your funds and 625 00:31:09,720 --> 00:31:11,880 something, yeah. Exactly because the real and 626 00:31:11,880 --> 00:31:14,000 this is, you know, part of the thing part of the the work that 627 00:31:14,000 --> 00:31:16,640 we have done at container so far is just this this foundational 628 00:31:16,640 --> 00:31:19,160 layer of trying to address trust and policy. 629 00:31:19,160 --> 00:31:21,600 The the issue that people have with agent took work flows is 630 00:31:22,400 --> 00:31:25,800 not yet things related to price sensitivity or even largely data 631 00:31:25,800 --> 00:31:27,840 privacy issues. It's just reliability. 632 00:31:28,160 --> 00:31:31,400 How can you trust these things? Who's who's vouching for them 633 00:31:31,440 --> 00:31:33,880 or, or when they, when they exceed their guardrails, who's 634 00:31:33,880 --> 00:31:38,240 liable, you know, all these sorts of guardrail policy 635 00:31:38,240 --> 00:31:42,040 reliability, really trust issues and so. 636 00:31:42,240 --> 00:31:44,120 This is why, yeah, exactly. That's the thing. 637 00:31:44,160 --> 00:31:50,280 The AI like, you know, I recognize that it's in like in 638 00:31:50,280 --> 00:31:52,400 some sense a through of the Gartner Hype cycle. 639 00:31:52,440 --> 00:31:54,800 Now it's weird because both there's tons of money going into 640 00:31:54,800 --> 00:32:01,120 it, but also there's less energy than there was a year or two ago 641 00:32:01,120 --> 00:32:03,240 in the sense of Oh my God, it's going to kill everybody and so 642 00:32:03,240 --> 00:32:03,800 and so forth. Right. 643 00:32:04,120 --> 00:32:11,760 But my view is, at least right now, as toy 25, AI requires a 644 00:32:11,760 --> 00:32:14,520 lot of supervision to get anything non trivial done. 645 00:32:14,600 --> 00:32:16,600 And the smarter you are, the smarter the AI is. 646 00:32:16,880 --> 00:32:19,200 And it doesn't do it end to end. It doesn't middle to middle 647 00:32:20,040 --> 00:32:23,920 right because the prompting and the verifying have to be done by 648 00:32:23,920 --> 00:32:26,760 humans. And so it's great generating 649 00:32:26,760 --> 00:32:31,640 reams and reams of text, but often a lot of that is filler. 650 00:32:32,560 --> 00:32:35,120 And I can always tell when someone's used AI to do 651 00:32:35,120 --> 00:32:37,800 something and even when you're using it. 652 00:32:37,880 --> 00:32:41,040 Like I think the biggest thing for me whenever we use AI 653 00:32:41,040 --> 00:32:45,880 internally is there's actually only relatively few kinds of 654 00:32:45,880 --> 00:32:51,160 tasks which can truly tolerate non deterministic error prone 655 00:32:51,160 --> 00:32:56,400 output. Images and video can, but back 656 00:32:56,400 --> 00:32:59,040 in code can't. Like front end code is more 657 00:32:59,040 --> 00:33:02,320 tolerant of those issues because we have our GPU's in our eyes 658 00:33:02,320 --> 00:33:05,720 and we can instantly see it and verify that the widget is off. 659 00:33:06,040 --> 00:33:09,400 But back end code is much more subtle and it's like much harder 660 00:33:09,480 --> 00:33:11,920 to determine that it's off without like really going line 661 00:33:11,920 --> 00:33:14,840 by line, you know, right. And and it's possible there's 662 00:33:14,840 --> 00:33:17,080 some like visualization mechanism, you know, like a 663 00:33:17,080 --> 00:33:20,480 Fourier transform or like an audio spectrogram can turn 664 00:33:20,560 --> 00:33:22,400 something that's not visual into something visual. 665 00:33:22,640 --> 00:33:25,360 So it's possible we might be able to turn some of these other 666 00:33:25,360 --> 00:33:27,680 data structures like back end code into things that we can 667 00:33:27,680 --> 00:33:30,400 just visually debug by eye, you know, like a state machine or 668 00:33:30,400 --> 00:33:32,520 something like that, right. So it might be might be 669 00:33:32,520 --> 00:33:35,640 something along those lines or it's in a domain with smart 670 00:33:35,640 --> 00:33:37,680 contracts, for example. As you know, formal verification 671 00:33:37,680 --> 00:33:41,360 finally became useful because these were such compact programs 672 00:33:41,360 --> 00:33:43,520 and of such high value that all the compute for formal 673 00:33:43,520 --> 00:33:45,000 verification actually became valuable. 674 00:33:45,000 --> 00:33:47,320 Right. So maybe maybe that's the answer 675 00:33:47,320 --> 00:33:50,880 is, you know, we, we have an only generate code that we can 676 00:33:50,880 --> 00:33:53,160 do formal verification on or something along those lines. 677 00:33:54,680 --> 00:34:00,080 But the reason I ask is where, you know, shopping seems to me 678 00:34:00,080 --> 00:34:03,880 to be like the simple kind of thing of get me a good plane 679 00:34:03,880 --> 00:34:07,200 ticket that satisfies my requirements. 680 00:34:08,239 --> 00:34:12,520 Like from point A to point B that doesn't cost more than X 681 00:34:13,280 --> 00:34:16,080 seems to me like a good place to start because that actually does 682 00:34:16,080 --> 00:34:18,199 take a surprising amount of time to go and book tickets. 683 00:34:18,199 --> 00:34:20,120 It's like actually kind of a pain still, right? 684 00:34:20,440 --> 00:34:22,840 Oh, you're putting your name, blah, blah, blah, Fill in all 685 00:34:22,840 --> 00:34:25,320 this stuff on every new, you know, airline site. 686 00:34:25,840 --> 00:34:30,120 So that's like one thing, right? And I think the Max budget stops 687 00:34:30,120 --> 00:34:32,560 it from totally screwing up too much, right? 688 00:34:32,560 --> 00:34:36,280 And maybe you just assign it to get food or a book you like or 689 00:34:36,280 --> 00:34:37,040 something like that. I don't know. 690 00:34:37,040 --> 00:34:39,800 Maybe you have some thoughts on what it would buy for you first 691 00:34:39,800 --> 00:34:44,199 or what you do with it first. Yeah, I so my belief is that 692 00:34:45,080 --> 00:34:47,400 right now, I mean things could change by the time this 693 00:34:47,400 --> 00:34:49,760 conversation is done. This space is very so quickly, 694 00:34:49,920 --> 00:34:52,840 right. But right now I think consumer 695 00:34:52,840 --> 00:34:55,880 retail shopping experiences will be one of the hardest nuts to 696 00:34:55,880 --> 00:34:59,520 crack in terms of a, in terms of a full shopping flow. 697 00:34:59,680 --> 00:35:03,960 So I think, you know, I can see it replacing search, which is 698 00:35:03,960 --> 00:35:05,960 sort of already happened to content on the web in some way. 699 00:35:06,360 --> 00:35:11,280 But the full flow, I think that'll be, I think we'll see B 700 00:35:11,280 --> 00:35:15,560 to B or even B to B to B sort of agentic payment solutions happen 701 00:35:15,560 --> 00:35:19,160 before we see some of the consumer use cases, but. 702 00:35:20,560 --> 00:35:22,800 How about EC2 then compute auctions, right? 703 00:35:22,800 --> 00:35:26,200 So Amazon has auctions of compute because you've got real 704 00:35:26,200 --> 00:35:28,600 time pricing as demand goes up and down, right? 705 00:35:28,600 --> 00:35:30,760 And you can get reserved instances versus real time. 706 00:35:30,760 --> 00:35:33,840 Maybe somebody's already doing that, but that's an example of a 707 00:35:34,040 --> 00:35:37,280 machine bidding on a resource used by a machine for an 708 00:35:37,280 --> 00:35:40,520 algorithm where it's it's like a truly a machine economy. 709 00:35:41,000 --> 00:35:43,640 I don't know, maybe you? Did yeah and, and, and use cases 710 00:35:43,640 --> 00:35:46,960 like, you know, supply supply chains or another one where a 711 00:35:46,960 --> 00:35:49,480 buyer agent and a seller agent can potentially negotiate. 712 00:35:49,480 --> 00:35:53,440 So if you, I have 12 approved vendors for this, you know, this 713 00:35:53,440 --> 00:35:56,720 piece that I need to acquire the, this sort of discovery, 714 00:35:56,720 --> 00:36:00,280 negotiation and also handling, you know, things like the 715 00:36:00,280 --> 00:36:04,320 fulfillment could be handled by agents on both sides and, and, 716 00:36:04,320 --> 00:36:07,960 and in that case, the, the, the data may be known, but in many 717 00:36:07,960 --> 00:36:10,240 cases, the data types themselves, the schemas may 718 00:36:10,240 --> 00:36:12,560 vary. And so you have this need to 719 00:36:12,560 --> 00:36:17,200 parse, you know, not necessarily unstructured, but not always, 720 00:36:18,360 --> 00:36:21,040 you know, schema compliance structure data, which elements 721 00:36:21,040 --> 00:36:25,040 can be quite good at. But to make that reliable, you, 722 00:36:25,080 --> 00:36:29,120 you, you know, what we found so far to make that reliable, It's 723 00:36:29,120 --> 00:36:31,600 not the AI that is actually executing the task for 724 00:36:31,600 --> 00:36:33,280 interpreting the data and beginning to handle the 725 00:36:33,280 --> 00:36:36,800 negotiation that requires human, you know, subject matter expert 726 00:36:36,800 --> 00:36:38,680 alignment. It's actually the evaluating 727 00:36:39,560 --> 00:36:42,440 sort of elements as jury sort of, you know, evaluators that 728 00:36:42,440 --> 00:36:45,000 are other AIS just evaluating the output of those tasks 729 00:36:45,320 --> 00:36:48,440 executors. It's that that layer that 730 00:36:48,440 --> 00:36:50,960 requires human subject matter expert alignment in order to 731 00:36:50,960 --> 00:36:54,840 make the the sort of upstream task like much more reliable and 732 00:36:54,840 --> 00:36:57,000 effective. So, so actually, you know, just 733 00:36:57,000 --> 00:36:59,880 poking on that for a little bit. The only thing about supply 734 00:36:59,880 --> 00:37:06,280 chain is, you know, I actually ordered a lot of, you know, I, I 735 00:37:06,280 --> 00:37:08,600 still, I still building network school. 736 00:37:08,600 --> 00:37:11,520 So I actually, I'm constantly looking at bill of materials and 737 00:37:11,520 --> 00:37:13,200 and so on and so forth for various physical things. 738 00:37:14,640 --> 00:37:19,920 That is a manual in, in my view, a fairly manual process because 739 00:37:20,160 --> 00:37:23,120 you're talking to the vendors you are. 740 00:37:23,520 --> 00:37:25,680 I think an agent can help you with getting quotes. 741 00:37:25,680 --> 00:37:28,360 Maybe that can be helpful, right, If it's hitting a bunch 742 00:37:28,360 --> 00:37:31,160 of telegrams and Whatsapps and, and so that that actually could 743 00:37:31,160 --> 00:37:33,000 be quite helpful. Like I have a spreadsheet and 744 00:37:33,000 --> 00:37:35,800 I'm like, agent, get me, get me quotes on this stuff where it's 745 00:37:35,800 --> 00:37:38,840 like only partially listed a little bit like OTC markets, you 746 00:37:38,840 --> 00:37:40,880 know, where some stuff isn't listed that that's maybe 747 00:37:40,880 --> 00:37:43,840 helpful. But a lot of those are like 748 00:37:43,840 --> 00:37:46,560 large enough. And I also don't need it so fast 749 00:37:46,960 --> 00:37:49,080 that I wouldn't review the purchase order before I hit 750 00:37:49,080 --> 00:37:52,800 submit on it, you know? But that example would like EC2 751 00:37:52,800 --> 00:37:56,280 compute is something where you might literally put that in a 752 00:37:56,280 --> 00:37:58,680 subroutine of get me the best price on this, which people 753 00:37:58,680 --> 00:38:01,040 already do for reserved instances and so on. 754 00:38:01,040 --> 00:38:05,600 You know, so like that's kind of machine resources like getting 755 00:38:05,600 --> 00:38:08,720 storage, getting compute that feels like where it's like kind 756 00:38:08,720 --> 00:38:11,640 of native, you know, or I don't know, maybe go ahead. 757 00:38:12,400 --> 00:38:17,280 Yeah, I think that, you know, as the the sort of business model 758 00:38:17,280 --> 00:38:20,200 of the web transforms from from, you know, search and advertising 759 00:38:20,640 --> 00:38:25,040 dominated into, you know, sort of data curation, this sort of 760 00:38:25,040 --> 00:38:28,280 data curation flywheel that is beginning to emerge, which is 761 00:38:30,200 --> 00:38:32,560 intelligence, whether it's agents or Ellen Foundation, 762 00:38:32,560 --> 00:38:35,560 people providing agentic intelligence need data. 763 00:38:36,360 --> 00:38:38,560 That data needs to be curated ultimately by humans, can be 764 00:38:38,560 --> 00:38:40,960 synthetic data, but it ultimately needs to be curated 765 00:38:41,280 --> 00:38:44,080 and sort of aligned by humans and humans need to be paid for 766 00:38:44,080 --> 00:38:47,200 that job. So that creates a little bit 767 00:38:47,200 --> 00:38:50,280 more of a data content service marketplace that has not 768 00:38:50,560 --> 00:38:52,760 existed. And the services that are 769 00:38:52,760 --> 00:38:56,600 helping to provide that data to the intelligence providers will 770 00:38:56,600 --> 00:39:00,160 pay humans for that job. And so the, the sort of task of 771 00:39:00,160 --> 00:39:04,480 a human begins to be curate these, this data content, these 772 00:39:04,480 --> 00:39:06,560 services, so that the intelligence ultimately is 773 00:39:06,560 --> 00:39:09,440 better for us. We can be more productive and we 774 00:39:09,440 --> 00:39:11,160 get paid for, for doing that task. 775 00:39:11,160 --> 00:39:13,360 And then, and then we pay for the intelligence in turn. 776 00:39:13,720 --> 00:39:16,480 And, and so whether it's machine resources for things like 777 00:39:16,480 --> 00:39:19,480 compute or whether it's particular content or data or, 778 00:39:19,520 --> 00:39:22,840 you know, services that can be quantified in like a data schema 779 00:39:22,840 --> 00:39:27,200 or something, then that becomes a place where agents are passing 780 00:39:27,200 --> 00:39:30,760 money, making payment decisions, but also actually executing, you 781 00:39:30,760 --> 00:39:36,000 know, a binary payment task. So the thing where we where we 782 00:39:36,000 --> 00:39:38,560 sort of move, you know, a little bit up the stack. 783 00:39:38,760 --> 00:39:40,960 So we're still missing some primitives, by the way, to make 784 00:39:40,960 --> 00:39:43,800 that reliable. Some of them are related to this 785 00:39:44,720 --> 00:39:47,480 potential certification or identity layer that's necessary 786 00:39:47,480 --> 00:39:48,760 for agents. There are many people who are 787 00:39:48,760 --> 00:39:51,440 working on different versions of that, but it's kind of notion of 788 00:39:51,640 --> 00:39:54,080 agentic identity and handling authentication effectively. 789 00:39:54,480 --> 00:39:56,760 Even in the shopping example, if I'm do, if I'm shopping in 790 00:39:56,760 --> 00:40:00,480 Amazon and I'm not just using a chatbot, I'm not on Amazon.com, 791 00:40:01,120 --> 00:40:03,560 but I'm just, I'm interacting with the Amazon agent in some 792 00:40:03,560 --> 00:40:05,880 other way, how do I be sure it's Amazon? 793 00:40:06,600 --> 00:40:09,240 There isn't sort of this. Like, yeah, the authentication, 794 00:40:09,640 --> 00:40:13,400 the verification, yeah, it's like whose whose dog is this? 795 00:40:14,000 --> 00:40:16,400 Whose dog is this? And then, and then, and then, 796 00:40:17,040 --> 00:40:19,000 you know, what are the new security and risk factors that 797 00:40:19,000 --> 00:40:23,200 emerge when AI actors are the ones who are executing these 798 00:40:23,200 --> 00:40:27,280 kinds of transactions or an agent on my behalf and an Amazon 799 00:40:27,520 --> 00:40:30,320 agent on its behalf? What new risk factors emerge? 800 00:40:30,320 --> 00:40:32,280 This is why ultimately we're, we decided to create a new 801 00:40:32,280 --> 00:40:36,520 financial institution from the ground up using AI is because, 802 00:40:37,040 --> 00:40:39,600 and we know this from circle, you know this from, from, you 803 00:40:39,600 --> 00:40:43,440 know, building infrastructure to manage risk as well. 804 00:40:44,480 --> 00:40:47,480 You know, classic finance and banking risk infrastructure is 805 00:40:47,480 --> 00:40:49,400 designed to make sure no bots can ever use it. 806 00:40:50,040 --> 00:40:52,120 You need to be a KYC or KYV, an individual. 807 00:40:52,120 --> 00:40:54,040 What we actually need is completely the other way around. 808 00:40:54,240 --> 00:40:57,720 Let's assume the only actors who will be using this system will 809 00:40:57,720 --> 00:41:02,480 be bots or agents And, and how do we let the good ones in? 810 00:41:02,680 --> 00:41:04,200 How do we keep the bad actors out? 811 00:41:04,440 --> 00:41:06,680 How do we give them rules and policies so that they, they 812 00:41:06,680 --> 00:41:09,720 can't exceed? How do they how do we tell them 813 00:41:09,720 --> 00:41:13,560 to escalate to us as needed when they do exceed those policies? 814 00:41:13,600 --> 00:41:16,440 How do we have insurance or sort of liability protections for 815 00:41:16,440 --> 00:41:17,600 this? There's all these like these 816 00:41:17,600 --> 00:41:21,320 problems have not been cracked. And so ultimately we decided. 817 00:41:21,320 --> 00:41:23,560 Well, I mean, self-directed cars will be the first place where 818 00:41:23,560 --> 00:41:25,520 this stuff is happening now in some ways. 819 00:41:25,520 --> 00:41:26,680 Liability, all that kind of stuff. 820 00:41:26,680 --> 00:41:28,040 Who's at fault? All that stuff. 821 00:41:28,040 --> 00:41:30,120 Who's at fault exactly? And there is a path obviously 822 00:41:30,120 --> 00:41:32,560 to, to have sort of, yeah, we're mostly at level 2. 823 00:41:32,640 --> 00:41:36,240 I think there may be one Level 3 self driving in the US now. 824 00:41:36,240 --> 00:41:37,640 But anyway, we're mostly at level 2. 825 00:41:37,800 --> 00:41:40,680 There's sort of a path to get, you know, Level 3 self driving. 826 00:41:41,240 --> 00:41:43,800 There isn't really a similar path for agents executing 827 00:41:43,800 --> 00:41:47,200 financial transactions or or you know, aside from commerce doing 828 00:41:47,200 --> 00:41:50,520 things like handling treasury management a really large sums 829 00:41:50,520 --> 00:41:52,440 of money. And that is where we're headed. 830 00:41:53,800 --> 00:41:57,440 Yes, Well, I think you might want to, I mean, obviously 831 00:41:57,440 --> 00:42:00,720 agents and machine payments in crypto where an agent is trading 832 00:42:00,720 --> 00:42:04,200 on crypto bots, that's like a place where you start with some 833 00:42:04,200 --> 00:42:05,880 sometimes small, sometimes large amounts of money. 834 00:42:05,880 --> 00:42:09,160 So that like people have been building some intuitions on this 835 00:42:09,160 --> 00:42:11,160 for a while, right? But but I think you're right 836 00:42:11,160 --> 00:42:13,760 that we could do, we could do a lot more. 837 00:42:14,840 --> 00:42:16,160 OK. I wanted, I wanted to, unless 838 00:42:17,680 --> 00:42:19,440 you had something else on this, I wanted to switch gears a 839 00:42:19,440 --> 00:42:21,800 little bit and just ask about some different topics, if that's 840 00:42:21,800 --> 00:42:22,240 all right. OK. 841 00:42:22,320 --> 00:42:24,080 Go for it So. OK. 842 00:42:24,080 --> 00:42:27,800 So what, what I've been thinking a lot about what I'm working on 843 00:42:27,920 --> 00:42:31,560 is physical world crypto in a sense, right? 844 00:42:31,760 --> 00:42:34,480 Because you have, I think after cryptocurrency there's different 845 00:42:34,480 --> 00:42:37,080 directions you can take it. But I think crypto community is 846 00:42:37,080 --> 00:42:41,000 a big part of it because we got all these conferences and a 847 00:42:41,000 --> 00:42:43,440 conference, in my view, is actually like a it's an 848 00:42:43,440 --> 00:42:48,160 important subroutine for our space because we're so digital. 849 00:42:48,640 --> 00:42:50,720 And the conferences are where actually we all kind of come 850 00:42:50,720 --> 00:42:54,080 together in the physical world. And so we've got this star 851 00:42:54,080 --> 00:42:56,560 Society here off the coast of Singapore. 852 00:42:56,680 --> 00:42:57,920 We're just taking over an island. 853 00:42:58,120 --> 00:42:59,880 We've got thousands of people from around the world. 854 00:42:59,880 --> 00:43:01,280 It's actually really cool. You should come and visit. 855 00:43:02,200 --> 00:43:05,840 So I wanted to know what, you know, what do you think about 856 00:43:06,000 --> 00:43:10,280 Star Societies Network states? We're we're showing on that. 857 00:43:10,280 --> 00:43:11,280 Have you given any thought to that? 858 00:43:11,480 --> 00:43:12,800 Yeah. I mean, I think we're already in 859 00:43:12,800 --> 00:43:17,640 a case where many of the things that even 20 years ago were done 860 00:43:18,240 --> 00:43:21,320 in a confined sort of Geo space, they're online now. 861 00:43:21,320 --> 00:43:24,720 And so we, we've already, whether we want to formally sort 862 00:43:24,720 --> 00:43:27,440 of give a name to it or not, we have informally formed our own 863 00:43:27,440 --> 00:43:28,880 communities and our own societies. 864 00:43:29,720 --> 00:43:33,320 And we're in multiple ones of those with different identities 865 00:43:33,320 --> 00:43:35,120 that we present to those at different times. 866 00:43:35,120 --> 00:43:38,960 Who'd be, you know, sort of a gamer society at 1 moment and 867 00:43:38,960 --> 00:43:42,240 then something that is, you know, related to academia and 868 00:43:42,240 --> 00:43:46,320 the next moment or, you know, family management in the next, 869 00:43:46,320 --> 00:43:48,800 whatever it is, we're already in these multiple online societies, 870 00:43:48,800 --> 00:43:53,280 much more than, you know, then we participate in the physical 871 00:43:53,280 --> 00:43:55,920 space that we're in. So the next step beyond that is, 872 00:43:55,920 --> 00:43:59,160 well, if we're already in these other spaces, then transforming 873 00:43:59,160 --> 00:44:01,080 that to the physical space that we actually want to be. 874 00:44:01,920 --> 00:44:04,760 And, you know, how did, how do we think about, you know, 875 00:44:04,760 --> 00:44:08,400 forming actual real world corollaries to where we're 876 00:44:08,400 --> 00:44:13,720 already spending our mental and, you know, time online today? 877 00:44:13,720 --> 00:44:16,880 And so, you know, that's, that's the sort of transformation, I 878 00:44:16,880 --> 00:44:20,120 think I think of a largely from the view of, you know, we're 879 00:44:20,120 --> 00:44:23,320 creating a new kind of global hyper personalized bank that is 880 00:44:23,320 --> 00:44:26,560 run entirely by a is, is a level of private banking that most 881 00:44:26,560 --> 00:44:27,920 people would never have access to. 882 00:44:28,280 --> 00:44:34,040 And that is enabled by stable coins, the Internet, AI, you 883 00:44:34,040 --> 00:44:36,640 know, all of these things sort of coming together in, in in the 884 00:44:36,640 --> 00:44:41,760 middle of this Venn diagram. And that sort of bank is that is 885 00:44:41,760 --> 00:44:45,960 the new hyper personalized bank for people who are forming like 886 00:44:45,960 --> 00:44:49,000 physically in this, in their, in their space, A version of the 887 00:44:49,000 --> 00:44:51,920 society that they're already joining and already have joined 888 00:44:51,920 --> 00:44:53,840 online. Interesting, yes. 889 00:44:54,000 --> 00:44:57,640 And so let's say you, what is the Shawn Neville, you know, 890 00:44:57,640 --> 00:45:00,040 people ask what kind of company would you start? 891 00:45:00,040 --> 00:45:02,040 And you have started a company which is, you know, Cortana and 892 00:45:02,040 --> 00:45:04,640 you did circle before that. But I have a different question, 893 00:45:04,640 --> 00:45:07,480 which is, and actually we also start a currency, which is who 894 00:45:07,480 --> 00:45:09,240 started USDC. So we started companies who 895 00:45:09,240 --> 00:45:11,960 started currencies. I have, I have a question on the 896 00:45:11,960 --> 00:45:16,040 third kind of thing, which is not an intranet company or an 897 00:45:16,040 --> 00:45:17,640 intranet currency, but an intranet community. 898 00:45:18,200 --> 00:45:21,000 If Sean was starting the community, if you're starting, 899 00:45:21,040 --> 00:45:23,720 you know, something, what would that community be about the 900 00:45:23,720 --> 00:45:25,120 theme of it? Is there something you'd like to 901 00:45:25,120 --> 00:45:28,640 see in the world? For example, Aladdin immersion, 902 00:45:28,920 --> 00:45:31,720 right? Or you know, keto or something 903 00:45:31,720 --> 00:45:34,040 like that, right? Keto, kosher. 904 00:45:34,080 --> 00:45:37,920 OK, there's just examples, toy examples I've got, but you must 905 00:45:37,920 --> 00:45:39,480 have something you'd like to see in the world. 906 00:45:39,480 --> 00:45:42,080 What would be your ideal community if you could fork 907 00:45:42,080 --> 00:45:43,920 Shawn and clone yourself and build something? 908 00:45:44,720 --> 00:45:48,800 Yeah. I mean, for me it's about hyper 909 00:45:48,800 --> 00:45:52,680 personalization, transparency and trust that doesn't require 910 00:45:52,680 --> 00:45:56,320 relying on other human beings. And it's all coded, you know, so 911 00:45:56,520 --> 00:45:59,480 one of the things that drew me to, you know, cryptocurrency in 912 00:45:59,480 --> 00:46:02,880 the 1st place and Bitcoin specifically in, you know, in 913 00:46:02,880 --> 00:46:05,720 2011. And, and it was this idea that 914 00:46:05,960 --> 00:46:10,280 don't really need to rely on fallible humans or the 915 00:46:10,520 --> 00:46:13,120 governance structures that they create that are sort of enforced 916 00:46:13,120 --> 00:46:16,760 by courts. If you can code those things 917 00:46:17,680 --> 00:46:20,120 crypto and then code those things cryptographically and 918 00:46:20,120 --> 00:46:22,760 software. And so, you know, trust enforced 919 00:46:22,760 --> 00:46:26,000 by software is, is just, it's a fundamental theme. 920 00:46:26,000 --> 00:46:29,880 So one of the reasons I have such optimism about, although I 921 00:46:29,880 --> 00:46:31,480 think you're absolutely right, we're in a little bit of the 922 00:46:31,480 --> 00:46:34,080 trough of disillusionment in terms of, you know, people's 923 00:46:34,080 --> 00:46:38,360 views, views of and this switched because nine months ago 924 00:46:38,560 --> 00:46:40,680 people were really at least. They're beyond. 925 00:46:40,760 --> 00:46:43,320 IOAI, yeah, but you couldn't mention stable coins to them 926 00:46:43,440 --> 00:46:46,640 because especially in AI engineer communities, yeah, 927 00:46:46,720 --> 00:46:49,520 yeah, they're they're they crypto is even more toxic than 928 00:46:50,080 --> 00:46:51,680 it's another. Funny, you know what it is. 929 00:46:52,120 --> 00:46:54,240 Actually, it's funny. Savior going to say, I was going 930 00:46:54,240 --> 00:46:56,920 to say, go ahead, give a meme, you know, finish what you're 931 00:46:56,960 --> 00:46:57,880 saying. I was going to say something. 932 00:46:58,200 --> 00:47:00,560 Well, I was going to say that and it used to be that we would 933 00:47:00,560 --> 00:47:02,680 bend over backwards, not to mention the fact that our agents 934 00:47:02,680 --> 00:47:06,080 were using stable coins because, you know, people and I think of 935 00:47:06,080 --> 00:47:08,600 circle in a good way, is probably the most boring company 936 00:47:08,600 --> 00:47:12,000 in crypto because we have it's a dollar, it's a dollar, it's a 937 00:47:12,000 --> 00:47:13,560 dollar. We went into the front door, we 938 00:47:13,560 --> 00:47:14,880 got, we know all these sorts of things. 939 00:47:15,600 --> 00:47:18,080 And yet people would be afraid to shake my hand because I was a 940 00:47:18,080 --> 00:47:20,800 crypto, maybe a, you know, Ponzi schemer or something. 941 00:47:21,440 --> 00:47:24,840 And it's completely put the other way now where people have 942 00:47:24,840 --> 00:47:27,440 a lot of interest in stable coins, which over the last 943 00:47:27,440 --> 00:47:30,320 several months, because of the genius act and other things have 944 00:47:30,320 --> 00:47:32,320 really achieved escape velocity in the mainstream. 945 00:47:32,320 --> 00:47:35,440 I mean, it was inconceivable 6-7 years ago that Stripe would 946 00:47:35,440 --> 00:47:39,440 create a payment chain, you know, it's using and the Visa 947 00:47:39,440 --> 00:47:42,280 would be leaning in and, and you know, and so on and so forth. 948 00:47:42,600 --> 00:47:45,200 But it has flipped over the last few months where now people have 949 00:47:45,200 --> 00:47:48,480 a lot of interest in, in hearing about various forms of uses of 950 00:47:48,480 --> 00:47:50,920 stable coins, not even just as a currency, but as programmable 951 00:47:50,920 --> 00:47:53,280 money and as, as sort of a platform to build on. 952 00:47:53,560 --> 00:47:56,360 They really don't want to hear about the AI component as much. 953 00:47:56,560 --> 00:47:58,000 We're right in the middle of both of those things. 954 00:47:58,280 --> 00:48:00,480 But it is absolutely flipped the other direction. 955 00:48:00,840 --> 00:48:05,120 It's it's like first it's AI and then crypto and now it's like 956 00:48:05,280 --> 00:48:08,280 stable coins and a little bit, you know, it's just funny you 957 00:48:08,280 --> 00:48:12,560 have to say, yeah, but it's there's there's one thing about 958 00:48:12,560 --> 00:48:14,880 that, you know, the meme, which was the guy getting, you know, 959 00:48:14,880 --> 00:48:17,680 hanged and he says like first time to their guy, you know, 960 00:48:17,680 --> 00:48:20,800 like from it's, it's I will put them on the screen. 961 00:48:21,040 --> 00:48:26,440 But so I thought about that because, you know, right around 962 00:48:26,440 --> 00:48:30,160 the time when AI was really going vertical in late 2022 and 963 00:48:30,160 --> 00:48:35,280 FTX was happening, right? So many people were like, AI is 964 00:48:35,280 --> 00:48:39,040 the real innovation. The crypto stuff is all bogus. 965 00:48:39,640 --> 00:48:41,720 So many people were saying something like that, right? 966 00:48:42,280 --> 00:48:44,960 And you know, of course AI is a real innovation. 967 00:48:44,960 --> 00:48:48,160 But what's what's interesting today and huge obviously, but 968 00:48:48,160 --> 00:48:50,800 it's interesting today is I think in the fullness of time, 969 00:48:51,360 --> 00:48:55,520 we see a couple of things. First is AI is actually for many 970 00:48:55,520 --> 00:49:01,840 people in some contexts equivalent to fake AI scam. 971 00:49:02,080 --> 00:49:04,680 Is that AI or is it real? You'll often hear people say 972 00:49:04,680 --> 00:49:07,920 that, right? So it actually has in some ways 973 00:49:07,920 --> 00:49:10,160 the same. It's got pulled into some of the 974 00:49:10,160 --> 00:49:12,480 same issues that crypto has. Where is it a crypto scam? 975 00:49:12,480 --> 00:49:13,520 Is an AI scam? Is that right? 976 00:49:13,760 --> 00:49:17,440 So a lot of the AI guys, like lots of artists and so on, are 977 00:49:17,440 --> 00:49:19,320 super mad about AI. And so it's OK, That's one 978 00:49:19,320 --> 00:49:20,920 piece. The other thing that's 979 00:49:20,920 --> 00:49:24,640 interesting is in many ways crypto is what AI can't do. 980 00:49:26,760 --> 00:49:28,920 That's a deep point. Like I say, because AI is 981 00:49:28,920 --> 00:49:30,600 probabilistic and crypto is deterministic. 982 00:49:31,120 --> 00:49:33,520 AI can solve partial differential equations, but it 983 00:49:33,520 --> 00:49:35,040 can't solve cryptographic equations. 984 00:49:35,640 --> 00:49:39,200 And so because it cannot, you know, invert a hash function, a 985 00:49:39,200 --> 00:49:42,680 cryptographic hash function, there's certain kinds of NP hard 986 00:49:42,680 --> 00:49:44,000 problems. Obviously it can't, it literally 987 00:49:44,000 --> 00:49:45,480 can't solve them without us changing. 988 00:49:45,520 --> 00:49:48,600 But we know about computer science, you know, develop maybe 989 00:49:48,600 --> 00:49:50,360 P equals NP, but we don't know that. 990 00:49:52,200 --> 00:49:55,000 So without breakthroughs in theoretical computer science, 991 00:49:55,200 --> 00:49:57,600 like AI cannot solve cryptographic equations. 992 00:49:58,240 --> 00:50:02,560 And so that that means crypto is like a hard wall that can 993 00:50:02,600 --> 00:50:07,120 constrain and bound AI, right? It can say I have a dollar and 994 00:50:07,120 --> 00:50:11,440 the other AI or the human can say, OK, send it to me on chain. 995 00:50:11,440 --> 00:50:14,160 And if it can't do that, then it's actually fake. 996 00:50:14,160 --> 00:50:17,120 It's just emitting words. Your crypto is the actions and 997 00:50:17,120 --> 00:50:19,440 AI is the words in many ways. And the actions speak louder 998 00:50:19,440 --> 00:50:23,040 than words, right? So in a very deep sense, crypto 999 00:50:23,040 --> 00:50:27,240 and AI are complementary technologies that you know, you 1000 00:50:27,440 --> 00:50:30,080 like what you're what you're doing is one example of I think 1001 00:50:30,080 --> 00:50:33,760 a useful synthesis of them with an AI agent that spends crypto, 1002 00:50:33,880 --> 00:50:35,400 right? But I think there's others as 1003 00:50:35,400 --> 00:50:36,520 well. When you start thinking of them 1004 00:50:36,520 --> 00:50:38,760 as sort of peanut butter and Jelly, like dual technologies, 1005 00:50:39,040 --> 00:50:40,400 dual to each other. Let me know your thoughts. 1006 00:50:40,480 --> 00:50:41,600 Yeah, no, I think that's right on. 1007 00:50:42,400 --> 00:50:44,320 And you know the exact form that it takes. 1008 00:50:44,320 --> 00:50:46,440 No one. I, I, I, this is like his face 1009 00:50:46,440 --> 00:50:48,760 where I think I heard somebody mentioned it's really difficult 1010 00:50:48,760 --> 00:50:52,800 to play chess strategically in AI because, you know, you just. 1011 00:50:52,800 --> 00:50:55,000 Have to kind of go fast. Everything changes too fast. 1012 00:50:55,000 --> 00:50:56,880 The thing is to just keep marching the ponds down the 1013 00:50:56,880 --> 00:50:59,080 board right now. And so people are trying 1014 00:50:59,080 --> 00:51:00,240 different, different versions of that. 1015 00:51:00,240 --> 00:51:03,920 But they're great companies that are, you know, experimenting 1016 00:51:03,920 --> 00:51:08,240 with merging the two in ways that solve real world problems. 1017 00:51:08,320 --> 00:51:11,160 And there are many companies that large and small startups 1018 00:51:11,160 --> 00:51:13,960 and comments that have sort of realized, yeah, these these two 1019 00:51:13,960 --> 00:51:15,680 do fit. You know, there is a little bit 1020 00:51:15,680 --> 00:51:20,480 of a, there's the Crypto X AI piece, which I think has been a 1021 00:51:20,480 --> 00:51:23,560 little maybe a little over hyped, but there's reality there 1022 00:51:23,560 --> 00:51:26,320 in combining these two technologies at the same time, 1023 00:51:26,520 --> 00:51:27,240 so. That's right. 1024 00:51:27,440 --> 00:51:31,520 And in a sense, you know, this is another kind of parallel 1025 00:51:31,520 --> 00:51:35,040 financial system, which is it's the machine economy that is 1026 00:51:35,040 --> 00:51:37,920 booting up right now. Yeah. 1027 00:51:37,920 --> 00:51:39,800 Incentive perspective. Yeah, right. 1028 00:51:39,800 --> 00:51:42,960 So it's like another whole plane of where things are, you know, 1029 00:51:42,960 --> 00:51:45,200 trading back and forth. Absolutely. 1030 00:51:45,200 --> 00:51:47,960 I mean, I think ultimately Balji, I think it's going to 1031 00:51:47,960 --> 00:51:50,760 change everything. I mean, I, I know, you know, the 1032 00:51:50,760 --> 00:51:53,520 closest and the kind of analogy that we use are, well, there are 1033 00:51:53,520 --> 00:51:56,360 so many, you know, people who first experienced the Internet 1034 00:51:56,360 --> 00:51:58,440 on their devices that we sort of skipped the whole laptop error. 1035 00:51:58,440 --> 00:52:01,040 And then maybe there'll be other people who suddenly begin to 1036 00:52:01,040 --> 00:52:03,280 experience the Internet only through, you know, agentic 1037 00:52:03,280 --> 00:52:05,400 interfaces. And I don't think that's a great 1038 00:52:05,400 --> 00:52:07,480 analogy because I think the change is even bigger. 1039 00:52:08,520 --> 00:52:12,760 Are we still using web browsers to to interact with the Internet 1040 00:52:12,760 --> 00:52:15,800 if our interface is always going to be some series of agents? 1041 00:52:16,200 --> 00:52:19,480 We're not trying to crack the UX problem here, but these are the 1042 00:52:19,480 --> 00:52:21,600 kind of questions, you know that emerge. 1043 00:52:21,960 --> 00:52:26,280 I mean, replet is genuinely replet and chat CBT in codecs 1044 00:52:26,280 --> 00:52:31,560 and cloud code are genuinely new interfaces for, I mean, because 1045 00:52:31,560 --> 00:52:33,560 they can accept probabilistic input. 1046 00:52:34,520 --> 00:52:36,880 I like a lot of it, right? As opposed to, you know, it's 1047 00:52:36,880 --> 00:52:39,320 the exact opposite of how you type into a search engine. 1048 00:52:39,320 --> 00:52:42,120 A search engine you sort of invert and you find the least 1049 00:52:42,120 --> 00:52:44,400 frequent keywords in your head and you give the fewest 1050 00:52:44,400 --> 00:52:46,800 characters and boom, you've got a search query. 1051 00:52:47,280 --> 00:52:51,840 And with an agent, you just like you write a huge amount of text 1052 00:52:52,040 --> 00:52:55,320 and you're as detailed as possible and or necessarily a 1053 00:52:55,320 --> 00:52:58,480 huge amount, but like you can write quite a lot and it'll take 1054 00:52:58,480 --> 00:53:02,720 all of that and your vocabulary actually constrains the machine 1055 00:53:02,720 --> 00:53:04,880 and it's just completely different, you know, interface 1056 00:53:04,880 --> 00:53:08,360 in that sense. I mean, one other thought I had 1057 00:53:08,360 --> 00:53:12,440 for you, which is we, we didn't discuss this too much yet, but 1058 00:53:12,760 --> 00:53:16,840 there's a lot more robots of different form factors, not just 1059 00:53:16,840 --> 00:53:20,680 humanoid robots, but like robot dogs and drones and all kinds of 1060 00:53:20,680 --> 00:53:23,520 things coming out of China that you know, robot locks and so on 1061 00:53:23,520 --> 00:53:26,160 and so forth, right? That is another potential area 1062 00:53:26,160 --> 00:53:29,960 for machine to machine. For example, smart locks, like 1063 00:53:30,400 --> 00:53:35,800 here's an example, there's public you could have, for 1064 00:53:35,800 --> 00:53:39,160 example, in Singapore, there are these shipping containers that 1065 00:53:39,160 --> 00:53:42,560 have gyms like like fitness centers. 1066 00:53:42,560 --> 00:53:43,880 They're just in a shipping container. 1067 00:53:44,120 --> 00:53:47,040 OK. So you could in theory just go 1068 00:53:47,040 --> 00:53:51,720 up to it and go zip, zip like this and do a machine to machine 1069 00:53:51,720 --> 00:53:55,480 payment and open the the smart lock and have it debited, right? 1070 00:53:55,880 --> 00:53:58,640 That seems to me to be a pretty good application. 1071 00:53:58,680 --> 00:54:03,280 I mean, you can do it with Apple Pay, but it's lower cost if you 1072 00:54:03,280 --> 00:54:05,840 do it with machine to machine and it's cool, right? 1073 00:54:06,440 --> 00:54:09,920 And and you might be able to do things like that where you know 1074 00:54:09,920 --> 00:54:13,360 you've got a robot dog and it's sitting down and then you go 1075 00:54:13,760 --> 00:54:16,400 like this and it jumps up and it starts running around and doing 1076 00:54:16,400 --> 00:54:18,480 something right. I feel you might be able to do 1077 00:54:18,480 --> 00:54:20,360 some cool demos there. Let me know if you have. 1078 00:54:20,360 --> 00:54:22,360 Any yeah, yeah, yeah, yeah. It's a very cool idea. 1079 00:54:22,520 --> 00:54:25,200 I'm transforming this to the physical is just is is a great 1080 00:54:25,200 --> 00:54:28,200 space it's creative space to be thinking about regardless. 1081 00:54:28,480 --> 00:54:30,920 I do think in those kinds of examples, like you said, yes, 1082 00:54:30,920 --> 00:54:33,280 you can do this with Apple, but I think it's going to be very 1083 00:54:33,280 --> 00:54:35,360 important not to have vendor lock in in the next version of 1084 00:54:35,360 --> 00:54:39,240 this sort of agentic web. And, and you know, I, I think 1085 00:54:39,720 --> 00:54:41,920 that's one of the obviously is one of the things that kind of 1086 00:54:42,200 --> 00:54:45,760 the web was an amazing sort of culmination of many, many 1087 00:54:45,760 --> 00:54:47,400 partnerships and technologies and so on. 1088 00:54:47,400 --> 00:54:49,840 But we had also steered in certain direction that was not 1089 00:54:49,840 --> 00:54:52,400 the healthiest and, you know, did not. 1090 00:54:53,280 --> 00:54:54,680 And vendors. Exactly. 1091 00:54:54,760 --> 00:54:57,920 And like 5 companies that matter in terms of controlling 1092 00:54:57,920 --> 00:55:00,080 identity, you know, for, for instance. 1093 00:55:00,600 --> 00:55:03,240 And so the next version of this, when we, we are using agentic 1094 00:55:03,240 --> 00:55:06,400 interfaces, it ought to be much more broadly distributed. 1095 00:55:07,040 --> 00:55:09,320 And that's the way that it becomes more prosperous for more 1096 00:55:09,320 --> 00:55:13,360 people and, and sort of avoiding this, you know, yes, you could. 1097 00:55:13,360 --> 00:55:15,320 I trust Apple. I don't want to have to trust 1098 00:55:15,320 --> 00:55:17,000 that. And I, you know, obviously 1099 00:55:17,080 --> 00:55:18,840 respect Apple. I don't want to have to trust 1100 00:55:18,840 --> 00:55:21,720 them to get access to that gym. Yeah. 1101 00:55:22,040 --> 00:55:25,600 And, and the thing what's interesting is in some ways the 1102 00:55:25,600 --> 00:55:29,760 society where everything is, is computational trust actually 1103 00:55:29,760 --> 00:55:32,520 then loops around and becomes a high trust society again in some 1104 00:55:32,520 --> 00:55:36,640 circumstances, because you know, that person can't defraud you if 1105 00:55:36,640 --> 00:55:40,120 you checked it, right? So that means they don't have an 1106 00:55:40,120 --> 00:55:42,600 incentive to try, even try it. You know what I mean? 1107 00:55:42,760 --> 00:55:46,360 Right. So anyway, I, I think there's 1108 00:55:46,360 --> 00:55:47,880 something to that which is interesting. 1109 00:55:47,880 --> 00:55:51,360 Like if that their guy knows he can't get one over on somebody 1110 00:55:51,360 --> 00:55:54,200 because they'll check it. It doesn't even try, you know, 1111 00:55:54,200 --> 00:55:57,000 there's like a superior force, which is, you know, the the 1112 00:55:57,000 --> 00:56:00,880 blockchain like above them that is enforcing rule of code, you 1113 00:56:00,880 --> 00:56:05,880 know, between them, you know, so OK, cool. 1114 00:56:06,360 --> 00:56:09,680 I I enjoyed this. And if there's AI guess people 1115 00:56:09,680 --> 00:56:13,520 can go to Catina labs right now. It's join the team and 1116 00:56:13,680 --> 00:56:16,720 catinala-bs.com. Anything else you want to? 1117 00:56:17,320 --> 00:56:19,360 Yeah, I don't know, I'd say we've been, we've been pretty 1118 00:56:19,360 --> 00:56:21,440 open about some of the foundational sort of agent, you 1119 00:56:21,440 --> 00:56:24,280 know, identity and policy rules efforts that we're contributing 1120 00:56:24,280 --> 00:56:26,480 just open source code, been pretty quiet about the 1121 00:56:26,720 --> 00:56:29,840 commercial product offering. And so we we're hard work 1122 00:56:29,840 --> 00:56:31,720 building, we're really excited about it. 1123 00:56:31,720 --> 00:56:34,640 We're taking a big swing leverages all, all these 1124 00:56:34,640 --> 00:56:36,600 technologies. And so I'd say people are 1125 00:56:36,600 --> 00:56:40,920 interested from, you know, a partnership perspective, 1126 00:56:41,280 --> 00:56:44,400 learning more, working with us, potentially happy to have 1127 00:56:44,400 --> 00:56:46,440 conversations. Awesome, Great. 1128 00:56:46,640 --> 00:56:49,560 All right, Thank you very much, Sean, and thanks. 1129 00:56:49,560 --> 00:56:50,280 Pleasure. Great.