1 00:00:01,760 --> 00:00:03,779 Join us as we gather around the hedge, 2 00:00:03,839 --> 00:00:05,219 where we dig into technology, 3 00:00:05,519 --> 00:00:08,240 business, and culture with the finest minds in 4 00:00:08,240 --> 00:00:09,220 computer networking. 5 00:00:20,934 --> 00:00:23,335 Good morning, Tom. How are you? Good morning, 6 00:00:23,335 --> 00:00:25,675 Russ. I'm doing well. How about yourself? Fine. 7 00:00:25,734 --> 00:00:27,089 One of two today. 8 00:00:27,969 --> 00:00:29,730 Yes. Yes. It's gonna be a good day. 9 00:00:29,730 --> 00:00:31,730 Yeah. Once we finish this podcast, we'll be 10 00:00:31,730 --> 00:00:33,329 halfway done with all the work you have 11 00:00:33,329 --> 00:00:34,310 to get done today. 12 00:00:36,289 --> 00:00:38,929 I love Friday. Other other than turning the 13 00:00:38,929 --> 00:00:39,429 plant 14 00:00:40,815 --> 00:00:42,895 to make sure it doesn't grow crazy in 15 00:00:42,895 --> 00:00:44,335 one direction. I don't know what to do 16 00:00:44,335 --> 00:00:44,995 about that. 17 00:00:45,375 --> 00:00:48,094 So today, we are joined by Austin Fedor 18 00:00:48,094 --> 00:00:48,594 Fedorra? 19 00:00:49,054 --> 00:00:50,895 Fedorra? How do you say it? Fedorra? I'm 20 00:00:50,895 --> 00:00:53,635 not sure. Fedorra. Okay. Awesome. Great. 21 00:00:54,229 --> 00:00:56,630 And where are you, Austin? Other than sitting 22 00:00:56,630 --> 00:00:58,170 in front of the empty cat bed. 23 00:00:58,549 --> 00:01:00,409 Yes. I am based in New York City. 24 00:01:00,710 --> 00:01:03,049 Okay. Awesome. What part? Just curious. 25 00:01:03,509 --> 00:01:04,569 I'm over in Brooklyn. 26 00:01:04,870 --> 00:01:07,349 Okay. I used to, I used to work 27 00:01:07,349 --> 00:01:08,650 at a small reseller, 28 00:01:09,435 --> 00:01:11,114 EMC squared, and we did a lot of 29 00:01:11,114 --> 00:01:11,614 work 30 00:01:12,234 --> 00:01:14,474 in that area. And we actually did work 31 00:01:14,474 --> 00:01:16,314 at the hot one of our clients was 32 00:01:16,314 --> 00:01:17,694 the Haagen Dazs factory, 33 00:01:18,314 --> 00:01:21,275 and I believe it's in Brooklyn. And there's 34 00:01:21,275 --> 00:01:24,340 nothing like an ice cream bar directly off 35 00:01:24,340 --> 00:01:26,680 the line before it's been blast freeze. 36 00:01:27,060 --> 00:01:29,299 That's like, that's like crazy good ice cream 37 00:01:29,299 --> 00:01:30,040 right there. 38 00:01:30,340 --> 00:01:30,840 So, 39 00:01:32,500 --> 00:01:34,819 yeah, it was really good. And then I 40 00:01:34,819 --> 00:01:35,799 worked at the 41 00:01:36,254 --> 00:01:38,334 United Engineering Center for a couple of years, 42 00:01:38,334 --> 00:01:40,174 which is gone now, so whatever. It's called 43 00:01:40,575 --> 00:01:41,875 across the street from the UN. 44 00:01:42,254 --> 00:01:44,494 So that building is now gone. But at 45 00:01:44,494 --> 00:01:47,055 any rate. So New York City, Brooklyn. Yep. 46 00:01:47,055 --> 00:01:48,515 Very familiar with the area. 47 00:01:48,840 --> 00:01:51,479 Alright. Awesome. So today, we are talking about 48 00:01:51,479 --> 00:01:51,979 blockchain, 49 00:01:52,599 --> 00:01:55,319 and Austin's gonna explain everything there is no. 50 00:01:55,319 --> 00:01:56,140 Forget that. 51 00:01:57,799 --> 00:01:59,000 Austin. Let's see what we can do. We 52 00:01:59,000 --> 00:02:01,099 only have an hour. Yeah. So, 53 00:02:02,064 --> 00:02:05,344 Austin's gonna talk about scaling. So I know 54 00:02:05,344 --> 00:02:08,385 that, like, there's been stories about water and 55 00:02:08,385 --> 00:02:10,865 power and all sorts of issues around data 56 00:02:10,865 --> 00:02:13,284 center space. And, of course, we know that 57 00:02:14,080 --> 00:02:16,580 blockchain relies on proof of work, 58 00:02:17,199 --> 00:02:18,960 which others think is hilarious. I see these 59 00:02:18,960 --> 00:02:21,120 news stories. We're gonna reduce the amount of 60 00:02:21,120 --> 00:02:22,979 proof of work in blockchain. I'm like, 61 00:02:23,599 --> 00:02:25,219 don't you, like, 62 00:02:27,995 --> 00:02:29,935 do you not understand blockchain? 63 00:02:32,314 --> 00:02:33,375 Yeah. You really, 64 00:02:34,155 --> 00:02:36,075 it's kinda one of these funny things where 65 00:02:36,314 --> 00:02:39,355 so back when blockchain originally started, right, if 66 00:02:39,355 --> 00:02:40,539 we go back to sort of, like, the 67 00:02:40,539 --> 00:02:43,120 Bitcoin days and the original versions of Ethereum, 68 00:02:44,299 --> 00:02:46,459 proof of work was was the standard. It 69 00:02:46,459 --> 00:02:48,060 was what everything was using. 70 00:02:48,379 --> 00:02:51,180 And it basically runs on its security model. 71 00:02:51,180 --> 00:02:53,180 People think about it as hash power, but 72 00:02:53,180 --> 00:02:54,944 hash power is really just a proxy for 73 00:02:54,944 --> 00:02:55,685 energy usage. 74 00:02:56,064 --> 00:02:58,465 Right? The security of Bitcoin comes from the 75 00:02:58,465 --> 00:02:59,205 fact that 76 00:02:59,504 --> 00:03:02,064 you can't buy enough coal plants to attack 77 00:03:02,064 --> 00:03:04,465 the network. The actual power demand that you 78 00:03:04,465 --> 00:03:06,485 need to assemble to run the miners 79 00:03:06,784 --> 00:03:08,669 to try and attack the network 80 00:03:09,050 --> 00:03:12,010 is, like, you know, the entire production of 81 00:03:12,010 --> 00:03:13,310 a mid sized 82 00:03:13,689 --> 00:03:16,409 European country. It's just not Yeah. Viable for 83 00:03:16,409 --> 00:03:18,510 someone to even obtain that much electricity. 84 00:03:19,610 --> 00:03:21,550 The good news, though, is that, you know, 85 00:03:22,544 --> 00:03:25,984 starting in sort of roughly 2019 and really 86 00:03:25,984 --> 00:03:28,784 the transition, you know, continued into 2022 and 87 00:03:28,784 --> 00:03:29,764 2023. 88 00:03:29,905 --> 00:03:32,064 But at this point, Bitcoin and Doge are 89 00:03:32,064 --> 00:03:35,125 really the only two prominent protocols in crypto 90 00:03:35,264 --> 00:03:37,349 that are still on proof of work. Everything 91 00:03:37,349 --> 00:03:39,289 else has switched over to proof of stake, 92 00:03:39,509 --> 00:03:41,669 which basically is the same energy profile as 93 00:03:41,669 --> 00:03:43,849 any typical web two application would. 94 00:03:44,229 --> 00:03:45,909 Yeah. It's interesting because when I talked to 95 00:03:45,909 --> 00:03:47,849 Radia Perlman many years ago 96 00:03:48,229 --> 00:03:51,025 about Bitcoin and proof of work, she was 97 00:03:51,025 --> 00:03:53,344 very insistent that she would never invest in 98 00:03:53,344 --> 00:03:54,724 a proof of work system 99 00:03:55,824 --> 00:03:58,724 because, because of what you just said, like 100 00:03:58,784 --> 00:04:00,645 some country could dedicate 101 00:04:01,664 --> 00:04:03,264 some large country. It wouldn't be a mid 102 00:04:03,264 --> 00:04:05,284 sized country. That'd be a large company, China, 103 00:04:05,344 --> 00:04:08,520 India, Russia, picked something, you know, something like 104 00:04:08,520 --> 00:04:11,080 that, or some first world like US or 105 00:04:11,080 --> 00:04:12,219 Australia or something 106 00:04:12,760 --> 00:04:15,340 could actually dedicate enough energy resources 107 00:04:16,360 --> 00:04:18,459 to control more than 50% 108 00:04:18,600 --> 00:04:21,180 of the processing power thrown at Bitcoin. 109 00:04:21,595 --> 00:04:23,915 And then they could just basically take the 110 00:04:23,915 --> 00:04:26,574 chain over and there would be no undoing. 111 00:04:26,714 --> 00:04:28,495 It would be just a done thing 112 00:04:28,875 --> 00:04:30,254 because it's it's unidirectional. 113 00:04:30,954 --> 00:04:32,875 So that's that's interesting that you talk about 114 00:04:32,875 --> 00:04:35,060 proof of stake. So can you explain the 115 00:04:35,060 --> 00:04:37,300 difference between proof of stake and proof of 116 00:04:37,300 --> 00:04:38,279 proof of work? 117 00:04:38,819 --> 00:04:42,019 Yeah. So proof of stake uses basically the 118 00:04:42,019 --> 00:04:44,600 scarcity of an asset class as a security 119 00:04:44,660 --> 00:04:47,480 mechanism. And so with Bitcoin, right, Bitcoin is 120 00:04:47,620 --> 00:04:50,225 valuable for you have a thousand theses for 121 00:04:50,225 --> 00:04:52,064 why it's valuable. But the reason it's secure 122 00:04:52,064 --> 00:04:53,824 is because as we talked about, the energy 123 00:04:53,824 --> 00:04:56,165 budget is so high to attack. Yeah. 124 00:04:57,024 --> 00:04:58,545 With proof of stake, it's a little bit 125 00:04:58,545 --> 00:05:00,625 different. So if you look at a network 126 00:05:00,625 --> 00:05:03,879 like Solana, there's approximately, you know, 500,000,000 127 00:05:03,879 --> 00:05:05,800 tokens out there. And there are some that 128 00:05:05,800 --> 00:05:08,680 are programmatically created due to, like, inflation on 129 00:05:08,680 --> 00:05:09,339 the protocol 130 00:05:11,000 --> 00:05:13,500 to pull the same type of attack. Right? 131 00:05:13,639 --> 00:05:15,319 The attack vector for a proof of work 132 00:05:15,319 --> 00:05:15,819 protocol 133 00:05:16,435 --> 00:05:19,014 is you physically build enough miners, 134 00:05:19,314 --> 00:05:22,274 basically, you know, compute units to attack the 135 00:05:22,274 --> 00:05:24,514 network. With proof of stake, you have to 136 00:05:24,514 --> 00:05:26,675 actually buy them and acquire them. So you 137 00:05:26,675 --> 00:05:27,975 can do that through hacking, 138 00:05:28,595 --> 00:05:30,194 as, you know, North Korea is prone to 139 00:05:30,194 --> 00:05:32,750 do, or you can, you know, do that 140 00:05:32,750 --> 00:05:35,629 through purchasing them. But, you know, because there's 141 00:05:35,629 --> 00:05:38,110 not actually, like, enough supply on the market 142 00:05:38,110 --> 00:05:40,589 at any given time to attack them, it's 143 00:05:40,589 --> 00:05:42,269 not like someone could just go on and 144 00:05:42,269 --> 00:05:43,835 buy 30% 145 00:05:43,835 --> 00:05:45,834 of the supply of Solana. The market's not 146 00:05:45,834 --> 00:05:47,995 that deep. And even if it was, you'd 147 00:05:47,995 --> 00:05:50,235 send the price, you know, up to the 148 00:05:50,235 --> 00:05:52,475 moon just from trying to acquire all of 149 00:05:52,475 --> 00:05:54,955 those all of those shares. It's very similar 150 00:05:54,955 --> 00:05:57,115 to what happens in a traditional company if 151 00:05:57,115 --> 00:05:59,629 there's, like, a a takeover that someone's trying 152 00:05:59,629 --> 00:06:00,209 to pull. 153 00:06:00,669 --> 00:06:03,550 It it takes years usually to acquire enough 154 00:06:03,550 --> 00:06:05,709 shares and that, you know, to actually be 155 00:06:05,709 --> 00:06:07,069 able to get a board seat on a 156 00:06:07,069 --> 00:06:09,469 company. The the even the activist hedge funds 157 00:06:09,469 --> 00:06:11,329 have to be very, you know, 158 00:06:11,789 --> 00:06:12,289 surreptitious 159 00:06:12,589 --> 00:06:14,634 about how they're purchasing shares and those sorts 160 00:06:14,634 --> 00:06:15,215 of things. 161 00:06:15,835 --> 00:06:17,455 So it's just a different security, 162 00:06:18,154 --> 00:06:19,835 model. And so because of that, they have 163 00:06:19,835 --> 00:06:21,615 different energy profiles because 164 00:06:22,314 --> 00:06:24,795 the energy isn't securing proof of stake in 165 00:06:24,795 --> 00:06:26,399 the same way it is proof of work. 166 00:06:26,560 --> 00:06:28,740 Right. So so we could say that Bitcoin 167 00:06:28,879 --> 00:06:29,620 is basically 168 00:06:30,079 --> 00:06:32,339 a proof of stake, but it's in power, 169 00:06:33,279 --> 00:06:33,779 essentially, 170 00:06:34,240 --> 00:06:36,639 is what it is. Right? How much power 171 00:06:36,639 --> 00:06:38,180 can you bring to the game? 172 00:06:39,285 --> 00:06:41,925 How much Although the difference is, and this 173 00:06:41,925 --> 00:06:43,785 is kind of really important from, 174 00:06:44,485 --> 00:06:45,144 like, a 175 00:06:45,764 --> 00:06:47,625 a economic and dynamic perspective. 176 00:06:48,164 --> 00:06:50,104 In Bitcoin, your rewards are 177 00:06:51,209 --> 00:06:52,829 their contribution to the network. 178 00:06:53,129 --> 00:06:55,230 So if I contribute 10% 179 00:06:55,370 --> 00:06:57,769 of the of the hash power, which, you 180 00:06:57,769 --> 00:06:59,689 know, will be 10% of the power budget 181 00:06:59,689 --> 00:07:02,089 to the network, I will get 10% of 182 00:07:02,089 --> 00:07:04,490 the rewards for a given period of time. 183 00:07:04,490 --> 00:07:05,805 Now, obviously, because Bitcoin, 184 00:07:06,425 --> 00:07:07,865 it's you have to average it out over 185 00:07:07,865 --> 00:07:09,144 a period of time. But, you know, over 186 00:07:09,144 --> 00:07:11,324 enough time, I will average out to 10%. 187 00:07:11,704 --> 00:07:13,384 That's not the case in a proof of 188 00:07:13,384 --> 00:07:13,884 stake 189 00:07:14,185 --> 00:07:16,264 system. It's based on how many tokens you 190 00:07:16,264 --> 00:07:18,764 have attached to to your node, 191 00:07:19,144 --> 00:07:20,899 your earn rate. And so it's a it's 192 00:07:20,899 --> 00:07:22,899 a pretty interesting dynamic where one of the 193 00:07:22,899 --> 00:07:25,620 advantages of Bitcoin is it has an attribute 194 00:07:25,620 --> 00:07:28,039 of fairness to it, where if you're 195 00:07:28,339 --> 00:07:30,740 bringing x percent of the security budget, you 196 00:07:30,740 --> 00:07:33,000 receive exactly x percent of the reward. 197 00:07:33,394 --> 00:07:35,475 Proof of stake networks don't have that same 198 00:07:35,475 --> 00:07:35,975 dynamic. 199 00:07:36,675 --> 00:07:39,714 So if your validator is 10 times more 200 00:07:39,714 --> 00:07:40,935 performant than mine, 201 00:07:41,314 --> 00:07:43,154 but I have more tokens than you, I 202 00:07:43,154 --> 00:07:45,154 will earn more rewards than you even if 203 00:07:45,154 --> 00:07:47,735 you are technically helping the network more. 204 00:07:48,355 --> 00:07:50,699 Yeah. So so this is closer to a 205 00:07:50,699 --> 00:07:53,279 physical asset then. It's closer to 206 00:07:53,899 --> 00:07:56,459 real estate, for instance, where you happen to 207 00:07:56,459 --> 00:07:58,300 buy the piece of property that the highway 208 00:07:58,300 --> 00:07:59,279 is going through, 209 00:07:59,659 --> 00:08:01,500 and therefore, you get to sell to a 210 00:08:01,500 --> 00:08:03,360 gas station instead of a farm, 211 00:08:03,685 --> 00:08:05,144 and therefore you suddenly 212 00:08:05,524 --> 00:08:07,925 are much wealthier. Or even in the gold 213 00:08:07,925 --> 00:08:10,824 market where, you know, if somebody tries to 214 00:08:10,884 --> 00:08:13,285 to corner the gold market, well, guess what? 215 00:08:13,285 --> 00:08:15,044 They're helping you just as much as they're 216 00:08:15,044 --> 00:08:17,764 helping themselves in the long run. Or gold 217 00:08:17,764 --> 00:08:20,039 miners even. Right? People who go out and 218 00:08:20,039 --> 00:08:21,259 physically mine gold. 219 00:08:22,279 --> 00:08:24,919 The the most productive mine doesn't always win 220 00:08:24,919 --> 00:08:26,219 the financial game. 221 00:08:26,680 --> 00:08:29,800 It depends on marketing and the ability to 222 00:08:29,800 --> 00:08:31,800 get the gold to market and stuff like 223 00:08:31,800 --> 00:08:32,174 that. 224 00:08:32,894 --> 00:08:34,014 So it's interesting that 225 00:08:34,495 --> 00:08:37,134 so this is closer to me just thinking 226 00:08:37,134 --> 00:08:38,195 about it economically 227 00:08:38,654 --> 00:08:41,134 to something like that land, gold, something that's 228 00:08:41,134 --> 00:08:41,634 physical. 229 00:08:42,495 --> 00:08:42,995 So 230 00:08:43,695 --> 00:08:45,294 the result of that seems to be that 231 00:08:45,294 --> 00:08:46,470 it will be closer to 232 00:08:47,350 --> 00:08:48,409 physical financial 233 00:08:49,509 --> 00:08:50,490 results, like 234 00:08:51,189 --> 00:08:51,689 more 235 00:08:52,629 --> 00:08:53,129 random 236 00:08:53,750 --> 00:08:55,049 in the real world, 237 00:08:56,230 --> 00:08:57,209 but yet 238 00:08:58,309 --> 00:08:58,809 less 239 00:08:59,355 --> 00:09:01,914 consumption of resources to get there. Is that 240 00:09:01,914 --> 00:09:02,815 kind of the point? 241 00:09:03,514 --> 00:09:05,514 Like, I mean, random by I mean, random, 242 00:09:05,514 --> 00:09:07,034 like, you bought the land that was next 243 00:09:07,034 --> 00:09:08,815 to the right next to the 244 00:09:09,434 --> 00:09:11,195 highway, and you didn't realize you were doing 245 00:09:11,195 --> 00:09:12,970 that, you know, a hundred years ago or 246 00:09:12,970 --> 00:09:15,289 whatever it is. The the the thing that 247 00:09:15,289 --> 00:09:17,549 makes proof of stake network secure 248 00:09:17,929 --> 00:09:20,350 is the fact that they are valuable. 249 00:09:20,889 --> 00:09:22,970 Right? Which is sort of a tautology if 250 00:09:22,970 --> 00:09:25,095 you think about it. But, like, the the 251 00:09:25,095 --> 00:09:26,315 more valuable, 252 00:09:27,095 --> 00:09:29,575 you know, Apple gets, the harder it is 253 00:09:29,575 --> 00:09:30,774 to get a board to get them to 254 00:09:30,774 --> 00:09:32,535 buy enough shares to get a board seat 255 00:09:32,535 --> 00:09:34,695 on Apple. Right? It it it's that same 256 00:09:34,695 --> 00:09:37,029 exact type of structure. Structure. The place that 257 00:09:37,029 --> 00:09:38,309 proof of work and so in both of 258 00:09:38,309 --> 00:09:39,129 them, obviously, 259 00:09:39,509 --> 00:09:40,169 they break 260 00:09:42,070 --> 00:09:44,389 up caring enough to run the infrastructure. Right? 261 00:09:44,389 --> 00:09:45,990 And so if the price drops to a 262 00:09:45,990 --> 00:09:47,750 low enough point that someone can just buy 263 00:09:47,750 --> 00:09:50,070 it, you know, at all, well, there's an 264 00:09:50,070 --> 00:09:52,565 attack factor there with Bitcoin. If people stop 265 00:09:52,565 --> 00:09:55,524 caring enough about Bitcoin to secure it, it 266 00:09:55,524 --> 00:09:57,225 gets easier to attack it. 267 00:09:57,845 --> 00:09:59,365 So they both sort of have the same 268 00:09:59,365 --> 00:10:01,764 failure case. But in success case, they have 269 00:10:01,764 --> 00:10:03,065 very different dynamics, 270 00:10:03,970 --> 00:10:05,889 that are are you know, one of them, 271 00:10:05,889 --> 00:10:07,590 I would say, more closely mimics, 272 00:10:08,049 --> 00:10:11,090 probably, like, existing financial products and instruments that 273 00:10:11,090 --> 00:10:12,789 would be a proof of stake network. 274 00:10:13,330 --> 00:10:15,750 And the ones that more closely mirror commodities 275 00:10:16,289 --> 00:10:18,345 in terms of, like, the properties of gold 276 00:10:18,345 --> 00:10:19,945 or real estate or something along those lines, 277 00:10:19,945 --> 00:10:21,304 even though real estate is not exactly a 278 00:10:21,304 --> 00:10:21,804 commodity. 279 00:10:22,264 --> 00:10:23,945 That that's kind of the proof of work 280 00:10:23,945 --> 00:10:24,445 system. 281 00:10:25,625 --> 00:10:27,304 So then so then what are the what 282 00:10:27,304 --> 00:10:29,325 are the scaling challenges that you see? 283 00:10:29,785 --> 00:10:31,804 Yeah. So, you know, it depends kind of 284 00:10:32,029 --> 00:10:33,870 what you're looking at. So there's really only 285 00:10:34,350 --> 00:10:36,350 Bitcoin is really the only protocol that has 286 00:10:36,350 --> 00:10:37,090 no aspirations 287 00:10:37,470 --> 00:10:40,190 to get faster. Right? One of the values 288 00:10:40,190 --> 00:10:42,350 of Bitcoin is that it doesn't change. Right? 289 00:10:42,350 --> 00:10:44,529 It doesn't need to get faster because 290 00:10:45,070 --> 00:10:46,910 it as much as people talk about it 291 00:10:46,910 --> 00:10:48,985 as a currency, you know, it really is 292 00:10:48,985 --> 00:10:51,785 not no one's buying small things in Bitcoin 293 00:10:51,785 --> 00:10:53,485 for the most part. Or if they do, 294 00:10:53,625 --> 00:10:55,545 they're using it on top of some other 295 00:10:55,545 --> 00:10:58,504 network like Lightning or Wrapped Bitcoin or something 296 00:10:58,504 --> 00:11:00,684 along those lines. They're effectively using a derivative 297 00:11:00,745 --> 00:11:03,039 of Bitcoin. They're not using Bitcoin. Yeah. For 298 00:11:03,039 --> 00:11:04,879 those who don't know what lightning is, lightning 299 00:11:04,879 --> 00:11:07,440 bunches up. It's like it's almost like a 300 00:11:07,440 --> 00:11:10,720 bank, an intermediary bank, where they bunch up 301 00:11:10,720 --> 00:11:12,720 15 or 20 or a 100 or a 302 00:11:12,720 --> 00:11:13,779 thousand transactions 303 00:11:14,634 --> 00:11:16,975 into a single transaction. So you have, like, 304 00:11:17,194 --> 00:11:17,694 subtrees, 305 00:11:18,154 --> 00:11:19,834 and you have, like, okay, I'm gonna take 306 00:11:19,834 --> 00:11:20,875 this $10,000 307 00:11:20,875 --> 00:11:23,194 in Bitcoin, and I'm gonna make that a 308 00:11:23,194 --> 00:11:25,274 subtree, and people can do one and two 309 00:11:25,274 --> 00:11:27,834 dollar transactions within that $10,000 310 00:11:27,834 --> 00:11:28,334 scope. 311 00:11:28,690 --> 00:11:29,990 And then the whole 312 00:11:30,529 --> 00:11:31,029 transaction 313 00:11:31,889 --> 00:11:32,710 as a bundle 314 00:11:33,090 --> 00:11:34,870 gets pushed back to the blockchain. 315 00:11:35,970 --> 00:11:37,110 So it's so it's, 316 00:11:37,649 --> 00:11:39,410 so it allows you to do because a 317 00:11:39,410 --> 00:11:41,090 lot of the problem with with Bitcoin in 318 00:11:41,090 --> 00:11:42,769 this room is, like you said, it's it's 319 00:11:42,769 --> 00:11:44,070 power based. Right? 320 00:11:44,454 --> 00:11:46,134 Like, you don't wanna spend the power on 321 00:11:46,134 --> 00:11:48,634 a 50ยข transaction. Like, that's crazy. 322 00:11:49,014 --> 00:11:51,274 You're never gonna Yeah. There's there's not much 323 00:11:51,334 --> 00:11:53,095 there's not much what's called sort of block 324 00:11:53,095 --> 00:11:54,855 space, which is sort of how much data 325 00:11:54,855 --> 00:11:56,774 you can fit in a block. There's not 326 00:11:56,774 --> 00:11:58,319 much you can fit in a Bitcoin block, 327 00:11:58,319 --> 00:12:00,480 plus the finality on them is very long. 328 00:12:00,480 --> 00:12:03,139 You may depending on how much security confidence 329 00:12:03,199 --> 00:12:04,980 you want, you may be waiting an hour 330 00:12:05,360 --> 00:12:08,079 for your transaction to really settle, which, you 331 00:12:08,079 --> 00:12:10,000 know, maybe for buying a house, that's fine. 332 00:12:10,000 --> 00:12:11,860 But, like, it's not a day to day 333 00:12:12,164 --> 00:12:14,644 payment. So Right. When we look out for 334 00:12:14,644 --> 00:12:17,365 other types of networks, though, we we start 335 00:12:17,365 --> 00:12:20,245 to see, like, smart contract networks, which was 336 00:12:20,245 --> 00:12:21,464 sort of the thing Ethereum 337 00:12:21,845 --> 00:12:24,404 pioneered back in the early days. And smart 338 00:12:24,404 --> 00:12:26,164 contracts are really a fancy way of saying 339 00:12:26,164 --> 00:12:28,190 there's a there's a runtime. Right? And the 340 00:12:28,190 --> 00:12:30,509 minute there's a runtime, you can run programs 341 00:12:30,509 --> 00:12:31,649 on it. And so, 342 00:12:32,350 --> 00:12:33,970 the whole point of blockchains, 343 00:12:34,670 --> 00:12:36,509 you know, have evolved from being this sort 344 00:12:36,509 --> 00:12:37,330 of, like, cryptographically 345 00:12:37,710 --> 00:12:39,330 secure asset type 346 00:12:39,674 --> 00:12:41,914 to effectively what it is is it's it's 347 00:12:41,914 --> 00:12:45,054 globally verifiable compute that has no 348 00:12:45,434 --> 00:12:45,934 dependencies 349 00:12:46,394 --> 00:12:48,815 that rely on trust in the process. 350 00:12:49,355 --> 00:12:51,115 And that really is, like, what the whole 351 00:12:51,115 --> 00:12:53,115 thing is is built for. So, for example, 352 00:12:53,115 --> 00:12:54,720 when when I'm running code on a server 353 00:12:54,720 --> 00:12:55,460 on AWS, 354 00:12:56,080 --> 00:12:58,160 there's a lot of trust involved in that 355 00:12:58,160 --> 00:13:01,220 assumption. I'm trusting that AWS is not inspecting 356 00:13:01,360 --> 00:13:04,399 and monitoring the virtualized instance. I don't actually 357 00:13:04,399 --> 00:13:06,254 know where that server is, you know, all 358 00:13:06,254 --> 00:13:07,934 these sorts of things. And the way that 359 00:13:07,934 --> 00:13:09,235 the program is constructed 360 00:13:09,855 --> 00:13:10,355 is 361 00:13:11,615 --> 00:13:13,615 minded because I assume I'm running in a 362 00:13:13,615 --> 00:13:15,875 in a secure environment for the most part. 363 00:13:16,095 --> 00:13:19,634 When a program runs on, blockchain, what's effectively 364 00:13:19,774 --> 00:13:22,009 running it is is, you know, the 365 00:13:22,470 --> 00:13:25,610 world's slowest virtualized instance of a computer. 366 00:13:26,309 --> 00:13:28,889 And it's running, you know, like, most blockchains 367 00:13:28,950 --> 00:13:30,389 do one thing at a time or they 368 00:13:30,389 --> 00:13:32,629 have some amounts of parallelization, but not a 369 00:13:32,629 --> 00:13:33,129 ton. 370 00:13:33,764 --> 00:13:35,684 And that really is like a giant in 371 00:13:35,684 --> 00:13:37,845 order operation computer, and that was the first 372 00:13:37,845 --> 00:13:40,004 version. This was the Ethereum virtual machine. It 373 00:13:40,004 --> 00:13:42,085 was kind of the first network you could 374 00:13:42,085 --> 00:13:45,384 deploy programs onto. These programs were very compute 375 00:13:45,445 --> 00:13:49,039 starved. They actually more closely remembered, resembled computer 376 00:13:49,039 --> 00:13:51,440 programs from, like, the seventies and eighties where, 377 00:13:51,440 --> 00:13:53,519 like, you can't really do like, it took 378 00:13:53,519 --> 00:13:54,960 people a while to be able to do 379 00:13:54,960 --> 00:13:57,279 floating point math on blockchains. Right? There just 380 00:13:57,279 --> 00:14:00,000 wasn't a lot of capacity available for these 381 00:14:00,000 --> 00:14:00,644 types of things. 382 00:14:01,205 --> 00:14:02,884 And so that was kind of one of 383 00:14:02,884 --> 00:14:05,684 the examples that that, you know, we look 384 00:14:05,684 --> 00:14:08,165 at as, like, a scaling problem is how 385 00:14:08,165 --> 00:14:10,725 do you get more compute available? How do 386 00:14:10,725 --> 00:14:13,125 you get more block space available? There were 387 00:14:13,125 --> 00:14:14,884 a bunch of those types of questions that 388 00:14:14,884 --> 00:14:16,879 were sort of the original ideas of of 389 00:14:16,879 --> 00:14:18,259 how do you scale a blockchain. 390 00:14:19,039 --> 00:14:20,480 So why do you want more compute? Direction 391 00:14:20,480 --> 00:14:21,840 we can go from there. Yeah. Why why 392 00:14:21,840 --> 00:14:23,679 do you want more compute? I guess, is 393 00:14:23,679 --> 00:14:25,840 the first question. It feels like it should 394 00:14:25,840 --> 00:14:26,580 just be 395 00:14:27,039 --> 00:14:30,080 a storage of past transactions. So, like, what 396 00:14:30,080 --> 00:14:31,860 what does the compute get me? 397 00:14:33,315 --> 00:14:35,154 So there's a whole bunch of, like if 398 00:14:35,154 --> 00:14:36,595 you sort of think about it from the 399 00:14:36,595 --> 00:14:39,254 perspective of, like, a database. 400 00:14:39,714 --> 00:14:40,214 Right? 401 00:14:40,834 --> 00:14:42,834 The most basic thing of a database, its 402 00:14:42,834 --> 00:14:45,090 function is to write data into and store 403 00:14:45,250 --> 00:14:46,769 Well, you also need to do things like 404 00:14:46,769 --> 00:14:48,290 build an index. And what if you want 405 00:14:48,290 --> 00:14:50,290 to run analysis on the data that's within 406 00:14:50,290 --> 00:14:53,809 it? So the first version of blockchain was 407 00:14:53,809 --> 00:14:55,889 exactly that. It was Bitcoin. Right? And all 408 00:14:55,889 --> 00:14:57,105 Bitcoin cares about 409 00:14:57,584 --> 00:14:59,745 is recording entries in a ledger and keeping 410 00:14:59,745 --> 00:15:00,804 very good accounting. 411 00:15:01,184 --> 00:15:02,304 And the whole point of it is to 412 00:15:02,304 --> 00:15:04,384 make sure you can't do something called double 413 00:15:04,384 --> 00:15:04,884 spend. 414 00:15:05,985 --> 00:15:08,065 And double spend is exactly what it sounds 415 00:15:08,065 --> 00:15:09,605 like. It's I do a transaction, 416 00:15:10,544 --> 00:15:12,884 with each of you at the same time. 417 00:15:13,639 --> 00:15:15,720 Only one of those transaction can go through 418 00:15:15,720 --> 00:15:17,720 because I only have a certain amount of 419 00:15:17,720 --> 00:15:19,799 of tokens. I can't spend it in two 420 00:15:19,799 --> 00:15:21,179 places at the same time. 421 00:15:21,639 --> 00:15:24,679 And, you know, the double spend problem in 422 00:15:24,679 --> 00:15:26,679 the traditional world is solved by one or 423 00:15:26,679 --> 00:15:29,294 two ways. Either we have cash. Right? Right, 424 00:15:29,294 --> 00:15:30,894 and there's only x amount of bills. And 425 00:15:30,894 --> 00:15:32,894 so if I have a $10 bill, I 426 00:15:32,894 --> 00:15:34,115 can only spend it once. 427 00:15:35,054 --> 00:15:37,455 Or we have a bunch of systems of 428 00:15:37,455 --> 00:15:37,955 chargebacks 429 00:15:38,335 --> 00:15:39,075 and authorizations 430 00:15:39,455 --> 00:15:40,195 and preauthorizations 431 00:15:41,455 --> 00:15:44,870 that don't actually prevent double spend. They just, 432 00:15:46,529 --> 00:15:48,850 add some checks and balances in the process, 433 00:15:48,850 --> 00:15:50,490 and then we sort of solve it through 434 00:15:50,490 --> 00:15:52,370 a bunch of, like, credit card fees and 435 00:15:52,370 --> 00:15:54,929 other types of things. So, like, I can 436 00:15:54,929 --> 00:15:56,884 theoretically get a credit card that allows me 437 00:15:56,884 --> 00:15:58,725 to spend more money than I have, and 438 00:15:58,725 --> 00:16:00,565 that is a double spend problem in a 439 00:16:00,565 --> 00:16:02,804 sort. We've just got a complicated court system 440 00:16:02,804 --> 00:16:04,485 behind it. So to get to the question 441 00:16:04,485 --> 00:16:06,565 of, like, why do we want compute on 442 00:16:06,565 --> 00:16:07,065 chain, 443 00:16:07,605 --> 00:16:09,524 we basically won't compute on chain for the 444 00:16:09,524 --> 00:16:09,924 same reason 445 00:16:10,980 --> 00:16:13,699 we everything else. The minute you can add 446 00:16:13,699 --> 00:16:15,720 compute to a system, you add programmability, 447 00:16:16,100 --> 00:16:16,839 and programmability 448 00:16:17,220 --> 00:16:18,039 adds flexibility. 449 00:16:18,899 --> 00:16:21,220 So it's kind of the same argument of, 450 00:16:21,220 --> 00:16:23,480 like, oh, like, why do we need computers 451 00:16:23,539 --> 00:16:25,539 inside of, like, most of our electrical grid 452 00:16:25,539 --> 00:16:27,634 infrastructure? It's like, well, for management. 453 00:16:28,735 --> 00:16:30,815 And on blockchain management field can be a 454 00:16:30,815 --> 00:16:32,575 very different type of thing. So there's this 455 00:16:32,575 --> 00:16:33,795 concept of, like, programmable 456 00:16:34,335 --> 00:16:37,134 money, and it's both a technical concept and 457 00:16:37,134 --> 00:16:39,695 a real concept. But the minute you have 458 00:16:39,695 --> 00:16:41,795 something that is an asset where 459 00:16:43,000 --> 00:16:45,740 you can program qualities into it, 460 00:16:46,360 --> 00:16:48,199 the entire dynamic of how we relate to 461 00:16:48,199 --> 00:16:50,600 an asset changes. So I can give you 462 00:16:50,600 --> 00:16:52,279 a bunch of examples, but, like, one of 463 00:16:52,279 --> 00:16:55,000 the simplest ones was actually Andrew Yang had 464 00:16:55,000 --> 00:16:56,414 this whole idea for 465 00:16:56,975 --> 00:16:57,475 rebuilding 466 00:16:58,095 --> 00:17:00,274 food assistance in New York City 467 00:17:00,654 --> 00:17:02,975 onto a crypto based network that would be 468 00:17:02,975 --> 00:17:05,054 run by the city. And the reason for 469 00:17:05,054 --> 00:17:06,654 this is you can do a bunch of 470 00:17:06,654 --> 00:17:09,375 things with compute that you can't do in 471 00:17:09,375 --> 00:17:11,075 the traditional system. So 472 00:17:12,070 --> 00:17:13,849 a great example of this is 473 00:17:14,150 --> 00:17:16,809 they could give you money that, for example, 474 00:17:16,950 --> 00:17:19,609 increases or decreases based on a time interval. 475 00:17:20,070 --> 00:17:22,230 Right? So you could have a system where, 476 00:17:22,230 --> 00:17:22,970 let's say, 477 00:17:23,430 --> 00:17:25,509 you know, we go into COVID two. There's 478 00:17:25,509 --> 00:17:27,430 another stimulus check that goes out. What's the 479 00:17:27,430 --> 00:17:28,845 point of a stimulus check? The point of 480 00:17:28,845 --> 00:17:30,445 a stimulus check is for people to spend 481 00:17:30,445 --> 00:17:32,365 it to keep the economy going. If you 482 00:17:32,365 --> 00:17:33,964 take the stimulus check and you put it 483 00:17:33,964 --> 00:17:35,964 in your bank account, that actually doesn't do 484 00:17:35,964 --> 00:17:38,384 any good. That's not like what the people 485 00:17:38,525 --> 00:17:40,125 at the Fed want. They want you to 486 00:17:40,125 --> 00:17:40,625 go 487 00:17:41,309 --> 00:17:44,029 to stay open. So you could write, for 488 00:17:44,029 --> 00:17:46,289 example, a programmable money contract 489 00:17:46,829 --> 00:17:47,490 that says 490 00:17:48,190 --> 00:17:48,690 every 491 00:17:49,230 --> 00:17:52,349 week, we decree we we basically delete 5% 492 00:17:52,349 --> 00:17:53,549 of the money we sent you if you 493 00:17:53,549 --> 00:17:54,690 haven't spent it. 494 00:17:55,174 --> 00:17:57,095 And so suddenly, you have an incentive to 495 00:17:57,095 --> 00:17:59,015 spend that money as opposed to save that 496 00:17:59,015 --> 00:18:02,075 money. And that's something that, like, in theory, 497 00:18:02,375 --> 00:18:04,875 you could, I guess, do in a traditional 498 00:18:05,015 --> 00:18:07,654 stimulus check. But it's something that, like, the 499 00:18:07,654 --> 00:18:10,200 entire infrastructure of how we manage this stuff 500 00:18:10,200 --> 00:18:11,880 is not set up at all to allow 501 00:18:11,880 --> 00:18:13,019 that type of computation. 502 00:18:13,559 --> 00:18:15,579 With blockchain, like, rebalancing 503 00:18:16,039 --> 00:18:18,539 and auto compounding and these sorts of things, 504 00:18:18,599 --> 00:18:20,599 they're just baked in because the systems are 505 00:18:20,599 --> 00:18:23,305 built from the ground up, like, to be 506 00:18:23,305 --> 00:18:25,785 accounting and tracking ledgers. And that means making 507 00:18:25,785 --> 00:18:28,605 programmatic modifications to them is like quite trivial. 508 00:18:29,545 --> 00:18:31,005 So is it sounds 509 00:18:31,625 --> 00:18:34,025 it isn't there though some security in the 510 00:18:34,025 --> 00:18:36,809 fact that the money supply like, if if 511 00:18:36,809 --> 00:18:38,910 we just look at traditional economy, not Bitcoin, 512 00:18:38,970 --> 00:18:40,890 isn't there some security in the fact that 513 00:18:40,890 --> 00:18:42,990 the money supply itself can't do anything? 514 00:18:43,369 --> 00:18:45,210 That it has that someone else has to 515 00:18:45,210 --> 00:18:47,130 use it for like, an intelligent agent has 516 00:18:47,130 --> 00:18:49,464 to use it for a purpose? I mean, 517 00:18:49,464 --> 00:18:52,184 it doesn't it represent a a security risk? 518 00:18:52,184 --> 00:18:54,284 I would say, maybe it's just a philosophical 519 00:18:54,424 --> 00:18:55,404 thing, but, like, 520 00:18:55,704 --> 00:18:57,784 that that the money itself can't do anything. 521 00:18:57,784 --> 00:18:59,625 In this in this sense, it seems that 522 00:18:59,625 --> 00:19:01,625 the the, you know, the the money is 523 00:19:01,625 --> 00:19:04,460 is is intelligent. Doesn't that represent a risk 524 00:19:04,460 --> 00:19:05,279 to the system? 525 00:19:06,220 --> 00:19:08,299 It sort of represents a risk the same 526 00:19:08,299 --> 00:19:11,019 way that, like, smart traffic lights represent a 527 00:19:11,019 --> 00:19:15,179 risk. Right? And, like, gas pipelines represent a 528 00:19:15,179 --> 00:19:16,240 risk. Right? 529 00:19:16,605 --> 00:19:18,204 All of these types of things have some 530 00:19:18,365 --> 00:19:20,525 anytime you add a computer onto something, like, 531 00:19:20,525 --> 00:19:23,005 you have created in some abstract sense a 532 00:19:23,005 --> 00:19:25,025 security vulnerability into that system. 533 00:19:25,404 --> 00:19:27,244 Right. So it's really just a question of, 534 00:19:27,244 --> 00:19:29,404 like, are the trade offs that you get 535 00:19:29,404 --> 00:19:30,065 from this 536 00:19:30,640 --> 00:19:33,440 worth it? Right? The the safest, most efficient 537 00:19:33,440 --> 00:19:35,039 system in the world is one where we 538 00:19:35,039 --> 00:19:38,240 all pay in a heavy, hard to move 539 00:19:38,240 --> 00:19:39,380 physical asset. 540 00:19:39,920 --> 00:19:40,819 Right? Right. 541 00:19:42,160 --> 00:19:43,759 And there's a lot of downsides to that. 542 00:19:43,759 --> 00:19:45,220 Like, you're not gonna, like, 543 00:19:45,955 --> 00:19:47,875 have a bunch of gold physically in your 544 00:19:47,875 --> 00:19:50,275 house. Like, that's that's a security risk in 545 00:19:50,275 --> 00:19:51,654 its own, right, for different reasons. 546 00:19:52,275 --> 00:19:53,875 But we also do kind of have these 547 00:19:53,875 --> 00:19:54,375 programmability 548 00:19:54,755 --> 00:19:56,615 wrappers for US dollars. 549 00:19:57,634 --> 00:19:59,589 It's the Federal Reserve Bank. Right? It's the 550 00:19:59,589 --> 00:20:01,990 ACH system, the automatic clearing house that's run 551 00:20:01,990 --> 00:20:03,829 by the Fed. It's all of these other 552 00:20:03,829 --> 00:20:05,669 types of systems we've built on top of 553 00:20:05,669 --> 00:20:08,069 it that are are just as critical for 554 00:20:08,069 --> 00:20:09,450 money to flow and operate. 555 00:20:10,549 --> 00:20:12,630 But the difference is is that they're separate 556 00:20:12,630 --> 00:20:14,835 onto a different layer. And so one of 557 00:20:14,835 --> 00:20:17,554 the key points of blockchain is it's actually 558 00:20:17,554 --> 00:20:19,795 the most perfect system of accounting humans have 559 00:20:19,795 --> 00:20:20,295 ever 560 00:20:21,954 --> 00:20:22,694 can verify 561 00:20:23,394 --> 00:20:25,554 that the ledger is indeed correct, and they 562 00:20:25,554 --> 00:20:28,115 can see in real time if anyone has 563 00:20:28,115 --> 00:20:29,339 tried to modify it. 564 00:20:29,819 --> 00:20:32,220 Right? So, like, if Bank of America goes 565 00:20:32,220 --> 00:20:33,980 down you guys I'm I'm guessing maybe you've 566 00:20:33,980 --> 00:20:37,019 seen the show mister robot? Yeah. Mhmm. Yeah. 567 00:20:37,019 --> 00:20:39,019 So that's, like, a perfect example. Right? They 568 00:20:39,019 --> 00:20:40,940 they delete a bunch of bank system records, 569 00:20:40,940 --> 00:20:41,679 and suddenly, 570 00:20:41,980 --> 00:20:43,359 all the records of everyone's 571 00:20:43,664 --> 00:20:44,965 both debt and 572 00:20:45,505 --> 00:20:47,984 wealth has been deleted because the only place 573 00:20:47,984 --> 00:20:50,484 they existed was on the bank's ledger. 574 00:20:50,865 --> 00:20:52,244 The thing about Bitcoin 575 00:20:52,625 --> 00:20:53,285 and crypto 576 00:20:53,825 --> 00:20:57,025 is I can like, you can independently verify 577 00:20:57,025 --> 00:20:58,710 how much money I have in my account, 578 00:20:58,710 --> 00:21:00,309 and I can verify how much money I 579 00:21:00,309 --> 00:21:02,150 have in my account. And so because of 580 00:21:02,150 --> 00:21:03,829 that, we're not none of us is ever 581 00:21:03,829 --> 00:21:06,490 trusting some intermediary like Bank of America 582 00:21:06,869 --> 00:21:08,630 to say, like, here's how much money is 583 00:21:08,630 --> 00:21:11,829 in Austin's account. It's just verifiable by anyone 584 00:21:11,829 --> 00:21:13,554 out there. And there there's downsides of that 585 00:21:13,554 --> 00:21:15,394 too. But one of the main upsides is 586 00:21:15,394 --> 00:21:16,855 it it reduces the 587 00:21:17,474 --> 00:21:17,974 hierarchical 588 00:21:18,355 --> 00:21:20,375 relationship we have to money. 589 00:21:21,554 --> 00:21:23,794 Interesting. So I remember there used to be 590 00:21:23,875 --> 00:21:26,054 there was a Michael Crichton or a 591 00:21:26,470 --> 00:21:28,250 Tom Clancy novel or something 592 00:21:28,549 --> 00:21:31,430 where the entire stock market computer system was 593 00:21:31,430 --> 00:21:32,970 taken over by an attacker, 594 00:21:33,509 --> 00:21:36,470 and they caused the market to crash. And 595 00:21:36,470 --> 00:21:38,309 the solution was just to roll back to 596 00:21:38,309 --> 00:21:39,690 midnight the night before 597 00:21:39,990 --> 00:21:41,769 because it's all computers. 598 00:21:42,224 --> 00:21:44,065 Right? So I can scroll back to any 599 00:21:44,065 --> 00:21:45,825 point I want to in the past. What 600 00:21:45,825 --> 00:21:47,924 was the last fair trade we know about? 601 00:21:47,984 --> 00:21:50,224 Was it midnight? Let's pick something arbitrary. Let's 602 00:21:50,224 --> 00:21:52,384 go do it. So that was that was 603 00:21:52,384 --> 00:21:54,964 the solution they eventually offered to the problem. 604 00:21:55,670 --> 00:21:58,150 Now, when you talk about programmatic money, the 605 00:21:58,150 --> 00:21:59,769 other thing I think about are like, 606 00:22:00,390 --> 00:22:02,089 in the high speed trading market, 607 00:22:02,869 --> 00:22:04,710 one of the ways that people make their 608 00:22:04,710 --> 00:22:07,269 high speed trades faster is they actually program 609 00:22:07,269 --> 00:22:09,589 their trades into the forwarding engine on the 610 00:22:09,589 --> 00:22:10,089 hardware. 611 00:22:11,015 --> 00:22:12,855 Yes. They do. So in this case, you 612 00:22:12,855 --> 00:22:14,714 could program the trade 613 00:22:15,414 --> 00:22:18,234 into the smart money contract, perhaps, 614 00:22:18,694 --> 00:22:19,434 and say, 615 00:22:19,734 --> 00:22:22,315 oh, when the price gets to x, sell. 616 00:22:24,049 --> 00:22:26,630 So this actually already happens on chain. 617 00:22:27,330 --> 00:22:29,170 Not only do you have on chain limit 618 00:22:29,170 --> 00:22:30,690 orders and these sorts of things like you 619 00:22:30,690 --> 00:22:33,009 have in the traditional space, but, you know, 620 00:22:33,009 --> 00:22:35,170 some of the first versions of on chain 621 00:22:35,170 --> 00:22:38,230 trading markets were what's called automatic market maker 622 00:22:38,394 --> 00:22:40,654 systems. And they basically run on, 623 00:22:41,434 --> 00:22:44,714 a mathematical function that tries to balance both 624 00:22:44,714 --> 00:22:46,795 sides of the asset pool pretty much in 625 00:22:46,795 --> 00:22:47,454 real time. 626 00:22:47,835 --> 00:22:50,075 And so what these do is they they 627 00:22:50,075 --> 00:22:51,595 they take much less compute than, like, a 628 00:22:51,595 --> 00:22:53,339 central limit order book. They take a lot 629 00:22:53,339 --> 00:22:55,740 less updating than central limit order book. There's 630 00:22:55,740 --> 00:22:57,839 no cancels. Like, a lot of the complexities 631 00:22:58,059 --> 00:23:00,059 of running a modern order book are don't 632 00:23:00,059 --> 00:23:02,299 exist for an AMM. There's downsides of that 633 00:23:02,299 --> 00:23:04,619 too. And now crypto is largely moving over 634 00:23:04,619 --> 00:23:06,079 to central limit order books. 635 00:23:06,535 --> 00:23:08,855 But it's exactly this. Right? You actually even 636 00:23:08,855 --> 00:23:11,255 on some protocols today, you can have pre 637 00:23:11,255 --> 00:23:13,734 hooks and mid hooks and post hooks. And 638 00:23:13,734 --> 00:23:15,575 so you can sort of program, here's the 639 00:23:15,575 --> 00:23:17,994 trade I wanna do. If the market responds 640 00:23:18,055 --> 00:23:18,954 in this way, 641 00:23:19,279 --> 00:23:21,440 go do this secondary trade coming off of 642 00:23:21,440 --> 00:23:21,940 it. 643 00:23:22,480 --> 00:23:24,579 And so it really is, like, 644 00:23:26,319 --> 00:23:28,880 it's really powerful in terms of what you 645 00:23:28,880 --> 00:23:31,519 can actually program into a system if you 646 00:23:31,519 --> 00:23:32,899 are a professional trader. 647 00:23:34,174 --> 00:23:36,115 And one of the cool things about blockchain 648 00:23:36,174 --> 00:23:37,934 is this is kind of all on a 649 00:23:37,934 --> 00:23:39,775 level playing field. Right? If if I wanna 650 00:23:39,775 --> 00:23:41,294 go trade on the New York Stock Exchange 651 00:23:41,294 --> 00:23:43,154 and I wanna actually be competitive, 652 00:23:43,615 --> 00:23:45,214 it's millions of dollars a year just to 653 00:23:45,214 --> 00:23:46,115 get in the door. 654 00:23:46,519 --> 00:23:48,759 Right? I I need to have a server 655 00:23:48,759 --> 00:23:49,259 collocated 656 00:23:49,799 --> 00:23:52,519 at NYSE in Secaucus with the same length 657 00:23:52,519 --> 00:23:55,000 cables everyone else has and access to the 658 00:23:55,000 --> 00:23:56,759 fast data feeds, which are all done through 659 00:23:56,759 --> 00:23:59,954 multicast. And, you know, there's a whole big 660 00:23:59,954 --> 00:24:02,115 process, and it's expensive for a reason. Right? 661 00:24:02,115 --> 00:24:03,894 All that infrastructure is not free. 662 00:24:04,515 --> 00:24:06,755 But all of that is a system that 663 00:24:06,755 --> 00:24:08,294 creates, you know, sys 664 00:24:09,474 --> 00:24:11,794 in the top tier paying. But for everyone 665 00:24:11,794 --> 00:24:12,534 else downstream, 666 00:24:12,914 --> 00:24:15,180 there's no way, like, you can be competitive. 667 00:24:15,240 --> 00:24:16,460 Whereas on on crypto, 668 00:24:17,000 --> 00:24:19,420 if you just are the world's smartest quant, 669 00:24:19,720 --> 00:24:21,080 you can go on chain and you can 670 00:24:21,080 --> 00:24:22,680 trade. And, you know, you do have to 671 00:24:22,680 --> 00:24:24,759 get, like, a dedicated RPC server and stuff. 672 00:24:24,759 --> 00:24:27,580 But for orders of magnitude lower access, 673 00:24:28,535 --> 00:24:31,255 you actually can start trading and compete with 674 00:24:31,255 --> 00:24:33,335 anyone else trading on chain on a much 675 00:24:33,335 --> 00:24:34,634 more level playing field. 676 00:24:36,214 --> 00:24:38,934 Interesting. Okay. So let's return back to our 677 00:24:38,934 --> 00:24:41,654 original topic because we've gone down a pretty 678 00:24:41,654 --> 00:24:43,049 good rabbit hole here, 679 00:24:44,410 --> 00:24:47,289 which is how does all this relate to 680 00:24:47,289 --> 00:24:49,529 energy consumption? I think we've already kind of 681 00:24:49,529 --> 00:24:50,430 answered this, 682 00:24:50,890 --> 00:24:54,009 but scaling properties. Right? So for my audience, 683 00:24:54,009 --> 00:24:56,410 of course, it's network engineering and, like, what 684 00:24:56,410 --> 00:24:58,170 does this do for network? Seems like to 685 00:24:58,170 --> 00:24:58,670 me, 686 00:24:59,005 --> 00:25:01,164 there would be more traffic on the network 687 00:25:01,164 --> 00:25:02,944 because of this rather than less. 688 00:25:03,724 --> 00:25:05,804 Yeah. This is, you know, it it's funny 689 00:25:05,804 --> 00:25:08,044 because, like, you talk to network engineers and 690 00:25:08,044 --> 00:25:09,964 we actually we've we, you know, double zero 691 00:25:09,964 --> 00:25:12,419 is at its core a OSI layer one, 692 00:25:12,419 --> 00:25:15,619 two, and three networking protocol for distributed systems. 693 00:25:15,619 --> 00:25:17,779 And so we have many network engineers who 694 00:25:17,779 --> 00:25:19,539 work for us who were not in crypto 695 00:25:19,539 --> 00:25:20,899 at all until they they sort of got 696 00:25:20,899 --> 00:25:23,139 into this. And, you know, there's a little 697 00:25:23,139 --> 00:25:24,819 bit of, like, wait. This is seriously how 698 00:25:24,819 --> 00:25:26,664 it works that, like, a lot of network 699 00:25:26,664 --> 00:25:28,205 engineers get because 700 00:25:28,585 --> 00:25:30,744 one of the properties of these systems is 701 00:25:30,744 --> 00:25:31,404 they are 702 00:25:31,865 --> 00:25:32,365 decentralized 703 00:25:32,664 --> 00:25:33,484 and distributed. 704 00:25:33,865 --> 00:25:35,625 And so what does that mean? Let's take 705 00:25:35,625 --> 00:25:37,625 Solana, for example. Solana is one of the 706 00:25:37,625 --> 00:25:38,445 fastest blocks 707 00:25:39,630 --> 00:25:41,390 spent the last four years of my career 708 00:25:41,390 --> 00:25:43,490 before leaving the cofound double zero. 709 00:25:43,789 --> 00:25:45,730 But Solana, the blocks are 710 00:25:46,269 --> 00:25:47,490 about four hundred milliseconds. 711 00:25:47,950 --> 00:25:49,490 And when a block is produced, 712 00:25:50,794 --> 00:25:53,355 it then has to be broadcasted out as 713 00:25:53,355 --> 00:25:54,414 quickly as possible 714 00:25:54,794 --> 00:25:56,494 to the other 2,000 validators 715 00:25:56,875 --> 00:25:57,934 all over the world 716 00:25:58,394 --> 00:26:00,954 to be voted on. Because just because I 717 00:26:00,954 --> 00:26:02,634 built a block doesn't mean that all the 718 00:26:02,634 --> 00:26:05,119 transactions I put in are valid. Everyone else 719 00:26:05,119 --> 00:26:07,119 has to verify that the work I did 720 00:26:07,119 --> 00:26:09,440 is correct, and then they have to send 721 00:26:09,440 --> 00:26:11,679 back basically a vote that says, I sign 722 00:26:11,679 --> 00:26:12,980 it with my private key. 723 00:26:13,759 --> 00:26:15,759 I agree that this is a valid block. 724 00:26:15,759 --> 00:26:17,940 And that's kind of how block production happens, 725 00:26:19,744 --> 00:26:21,505 where we have to get you know, it's 726 00:26:21,505 --> 00:26:22,944 not a ton of data, but we have 727 00:26:22,944 --> 00:26:25,744 to get it around to 2,000 nodes in 728 00:26:25,744 --> 00:26:28,144 the world as quickly as possible so they 729 00:26:28,144 --> 00:26:29,585 can vote on it. And one of the 730 00:26:29,585 --> 00:26:31,744 interesting properties of Solana is it has no 731 00:26:31,744 --> 00:26:33,539 pauses between blocks. 732 00:26:33,920 --> 00:26:36,079 So it's not like four hundred milliseconds. We 733 00:26:36,079 --> 00:26:37,539 wait for everyone to agree. 734 00:26:37,920 --> 00:26:41,200 Another four hundred millisecond block happens. It's producing 735 00:26:41,200 --> 00:26:44,320 blocks basically in a stream. And that leader 736 00:26:44,720 --> 00:26:47,299 the the the person producing the block 737 00:26:47,714 --> 00:26:49,974 rotates every one point two seconds. 738 00:26:50,434 --> 00:26:51,714 So it might be being built in New 739 00:26:51,714 --> 00:26:53,315 York, and then it goes to, you know, 740 00:26:53,315 --> 00:26:55,154 Chicago, and then it's over in Singapore, and 741 00:26:55,154 --> 00:26:56,355 then it's back in New York, and then 742 00:26:56,355 --> 00:26:57,335 it's over in Frankfurt. 743 00:26:57,794 --> 00:26:59,714 And so you kind of have this crazy 744 00:26:59,714 --> 00:27:02,470 problem where there's a ton of data streaming 745 00:27:02,470 --> 00:27:03,990 out in real time. It has to get 746 00:27:03,990 --> 00:27:06,230 to everywhere else, basically, in the world where 747 00:27:06,230 --> 00:27:07,529 people are running these nodes. 748 00:27:08,069 --> 00:27:08,809 And then 749 00:27:09,190 --> 00:27:11,269 the leader's moving. So you then have to 750 00:27:11,269 --> 00:27:13,664 get data blasted out from there to everywhere 751 00:27:13,664 --> 00:27:15,424 in the world, and then the leader moves 752 00:27:15,424 --> 00:27:17,345 again. And so you have this, like, crazy 753 00:27:17,345 --> 00:27:17,845 system. 754 00:27:18,305 --> 00:27:19,904 At the same time, people are trying to 755 00:27:19,904 --> 00:27:20,725 send transactions. 756 00:27:21,265 --> 00:27:22,705 And the people who are trying to send 757 00:27:22,705 --> 00:27:23,205 transactions 758 00:27:23,825 --> 00:27:26,225 are trying to rush their data to that 759 00:27:26,225 --> 00:27:26,725 leader 760 00:27:27,369 --> 00:27:29,849 for possible inclusion in the block, but the 761 00:27:29,849 --> 00:27:32,490 leader moves around. Right? So sometimes it's like, 762 00:27:32,490 --> 00:27:35,149 oh, I'm just gonna send my transaction now, 763 00:27:35,690 --> 00:27:38,009 but I know the leader is gonna rotate 764 00:27:38,009 --> 00:27:41,130 in three hundred milliseconds, and it's gonna probably 765 00:27:41,130 --> 00:27:43,054 take me two hundred and fifty milliseconds to 766 00:27:43,054 --> 00:27:45,134 get to the current leader anyway. So I'm 767 00:27:45,134 --> 00:27:46,494 just gonna wait and send it to the 768 00:27:46,494 --> 00:27:47,234 next leader. 769 00:27:48,255 --> 00:27:49,934 And so it's a kind of a crazy 770 00:27:49,934 --> 00:27:51,294 world of and this is all done on 771 00:27:51,294 --> 00:27:54,015 the public Internet. Right? So we're doing these 772 00:27:54,015 --> 00:27:56,414 things, like, call basically building giant Merkle trees 773 00:27:56,414 --> 00:27:58,734 for state distribution. This is called Turbine. That's 774 00:27:58,734 --> 00:28:01,369 the the software package that does this on 775 00:28:01,369 --> 00:28:01,869 Solana. 776 00:28:02,570 --> 00:28:04,250 And, you know, it's sort of crazy if 777 00:28:04,250 --> 00:28:05,769 you if you think about it because any 778 00:28:05,769 --> 00:28:08,049 network engineer would be like, this sounds like 779 00:28:08,049 --> 00:28:09,630 a great application for multicast. 780 00:28:10,170 --> 00:28:12,250 That's what I was just thinking. A 100%. 781 00:28:12,250 --> 00:28:12,910 But you 782 00:28:13,534 --> 00:28:15,774 do it because you can't run multicast across 783 00:28:15,774 --> 00:28:16,674 the public Internet. 784 00:28:17,214 --> 00:28:17,714 Yeah. 785 00:28:18,494 --> 00:28:20,734 But you could run any cast. And I 786 00:28:20,734 --> 00:28:22,894 would think this is also that any cast 787 00:28:22,894 --> 00:28:25,134 is another possible solution. I always send my 788 00:28:25,134 --> 00:28:26,595 block to this IP address. 789 00:28:27,460 --> 00:28:29,380 Although, you know and then that IP address 790 00:28:29,380 --> 00:28:31,619 can change, or there are DNS solutions for 791 00:28:31,619 --> 00:28:34,100 this that are the location of the of 792 00:28:34,100 --> 00:28:35,400 that IP address can change. 793 00:28:35,700 --> 00:28:37,460 Of course, that only works as a suite 794 00:28:37,460 --> 00:28:38,039 of routing. 795 00:28:38,580 --> 00:28:40,660 Yes. But the problem is all those solutions 796 00:28:40,660 --> 00:28:43,894 require trusting an external third party dependency. Right? 797 00:28:44,835 --> 00:28:46,455 So DNS is not, like, 798 00:28:48,595 --> 00:28:52,115 strictly speaking, secure or decentralized. Right? It it 799 00:28:52,115 --> 00:28:54,615 is it is it is still a centralized 800 00:28:54,835 --> 00:28:57,119 operation. And there there have been many cases 801 00:28:57,119 --> 00:28:59,680 before where a government tells a DNS provider, 802 00:28:59,680 --> 00:29:01,700 hey. Block these records. And they do. 803 00:29:02,080 --> 00:29:04,559 Right? And so with crypto, because it kind 804 00:29:04,559 --> 00:29:06,960 of has these libertarian origins and these sort 805 00:29:06,960 --> 00:29:09,865 of, like, verify everything, trust nothing kind of 806 00:29:09,865 --> 00:29:10,924 ethos to it, 807 00:29:11,704 --> 00:29:13,164 those aren't really acceptable 808 00:29:13,625 --> 00:29:14,765 solutions for it. 809 00:29:15,224 --> 00:29:17,065 So the so on Solana and most other 810 00:29:17,065 --> 00:29:19,625 networks, they they basically have a protocol that 811 00:29:19,625 --> 00:29:20,765 has some sort of gossip, 812 00:29:21,079 --> 00:29:23,559 which is where every every validator is basically 813 00:29:23,559 --> 00:29:25,960 posting updates to say, like, hey. I'm I'm 814 00:29:25,960 --> 00:29:29,099 moving IP addresses or, you know, still online. 815 00:29:29,160 --> 00:29:31,259 You know, those sorts of things. And so, 816 00:29:31,880 --> 00:29:34,440 it's much more similar from a networking standpoint 817 00:29:34,440 --> 00:29:35,654 to something like BitTorrent, 818 00:29:36,914 --> 00:29:38,914 except we're trying to run BitTorrent at four 819 00:29:38,914 --> 00:29:42,375 hundred millisecond blocks at 10,000 transactions per second. 820 00:29:43,234 --> 00:29:44,775 Very interesting. So 821 00:29:45,634 --> 00:29:47,974 it sounds to me like this would be 822 00:29:49,330 --> 00:29:51,350 better able to run over 823 00:29:52,049 --> 00:29:52,549 a 824 00:29:54,049 --> 00:29:56,690 stick over a network. That's not stochastic. That 825 00:29:56,690 --> 00:29:57,190 is 826 00:29:57,490 --> 00:30:00,150 more time division multiplexing, like RDMA 827 00:30:00,850 --> 00:30:02,984 or something like that. But you're running it 828 00:30:02,984 --> 00:30:04,585 over the global Internet, which, of course, is 829 00:30:04,585 --> 00:30:07,144 stochastic. Like, you can't you you have no 830 00:30:07,144 --> 00:30:09,005 guarantees in the global Internet. 831 00:30:09,625 --> 00:30:11,705 Yeah. It's actually it's it's very funny because 832 00:30:11,705 --> 00:30:13,865 Anatoly was one of the cofounders of of 833 00:30:13,865 --> 00:30:16,025 Solana. He worked at Qualcomm for something like 834 00:30:16,025 --> 00:30:17,164 a decade before 835 00:30:17,569 --> 00:30:20,210 founding Solana. And his original vision for the 836 00:30:20,210 --> 00:30:22,390 thing was, like, how do we bring CDMA, 837 00:30:23,970 --> 00:30:25,970 basically to blockchain? And this is this is 838 00:30:25,970 --> 00:30:27,730 sort of the idea of proof of history, 839 00:30:27,730 --> 00:30:28,869 which is like a 840 00:30:29,170 --> 00:30:30,470 a a system that 841 00:30:30,994 --> 00:30:33,555 has, like, time stamping on data as it 842 00:30:33,555 --> 00:30:34,615 moves through the system. 843 00:30:34,994 --> 00:30:36,994 And it's kind of it's this very interesting 844 00:30:36,994 --> 00:30:38,914 idea where there's a lot of stuff that, 845 00:30:38,914 --> 00:30:41,315 like, from, like, the early Internet and early 846 00:30:41,315 --> 00:30:44,355 compute days that actually is really applicable to 847 00:30:44,355 --> 00:30:44,855 blockchain. 848 00:30:46,279 --> 00:30:47,799 And, you know, it's it's funny because you 849 00:30:47,799 --> 00:30:49,319 mentioned, like, oh, wouldn't this be faster if 850 00:30:49,319 --> 00:30:50,839 this would run over some other type of 851 00:30:50,839 --> 00:30:51,659 system? Well, 852 00:30:53,079 --> 00:30:54,679 this is sort of what we've been working 853 00:30:54,679 --> 00:30:56,519 on at double zero for the last year 854 00:30:56,519 --> 00:30:58,759 or so. Right? And it's the idea of, 855 00:30:58,759 --> 00:31:00,059 like, could we build 856 00:31:00,519 --> 00:31:02,139 a high performance private 857 00:31:02,735 --> 00:31:04,914 WAN that is run by multiple parties 858 00:31:05,215 --> 00:31:06,355 and has verified 859 00:31:06,975 --> 00:31:09,455 characteristics of networking built into it? Could we 860 00:31:09,455 --> 00:31:10,434 basically build 861 00:31:10,894 --> 00:31:13,955 a high frequency trading firm grade fiber network, 862 00:31:14,220 --> 00:31:16,299 but have it be run by 10 or 863 00:31:16,299 --> 00:31:18,319 15 or 20 independent parties 864 00:31:20,059 --> 00:31:22,559 resistance, and it has these type of guarantees 865 00:31:23,019 --> 00:31:24,880 that you'd expect from the public Internet. 866 00:31:27,055 --> 00:31:29,615 Yeah. Interesting. Which which the answer to that, 867 00:31:29,615 --> 00:31:30,974 by the way, would be yes. But it's 868 00:31:30,974 --> 00:31:32,494 gonna you have to be really clever to 869 00:31:32,494 --> 00:31:34,494 do it. There's a there's there are ways 870 00:31:34,494 --> 00:31:34,575 of 871 00:31:35,295 --> 00:31:37,855 now another thing that I'm always curious about. 872 00:31:37,855 --> 00:31:39,455 So when we go back to resourcing and 873 00:31:39,455 --> 00:31:41,159 we think about what how we're trying to 874 00:31:41,159 --> 00:31:43,240 solve some of the resourcing problems, so the 875 00:31:43,240 --> 00:31:45,000 first thing is we're doing proof of stake 876 00:31:45,000 --> 00:31:46,700 rather than proof of of work. 877 00:31:47,720 --> 00:31:49,880 Now this increases the amount of network traffic, 878 00:31:49,880 --> 00:31:50,779 but it decreases 879 00:31:51,480 --> 00:31:53,960 the water and power and all those physical 880 00:31:53,960 --> 00:31:55,500 resources, server resources. 881 00:31:55,855 --> 00:31:57,954 So that's basically the target 882 00:31:58,654 --> 00:32:00,815 in some way is to reduce those things. 883 00:32:00,815 --> 00:32:03,294 Like, there's this story about some company going 884 00:32:03,294 --> 00:32:04,974 to a small town and building a data 885 00:32:04,974 --> 00:32:08,174 center. And now all their bathtubs look horrible 886 00:32:08,174 --> 00:32:10,470 because all the water is now hard and, 887 00:32:10,549 --> 00:32:12,710 like, they can't get good water anymore because 888 00:32:12,710 --> 00:32:14,250 the data center just consumes 889 00:32:14,950 --> 00:32:17,190 so much water that they're always at the 890 00:32:17,190 --> 00:32:19,990 dregs of all their tanks. Like, it's all 891 00:32:19,990 --> 00:32:22,470 the rusted water that's been sitting out there 892 00:32:22,470 --> 00:32:24,785 for years and years and years that they're 893 00:32:25,244 --> 00:32:26,865 using. So that's that's an interesting, 894 00:32:27,404 --> 00:32:30,625 perspective of this, but, again, with higher networking 895 00:32:30,765 --> 00:32:32,525 requirements. Right? Is that kind of the trade 896 00:32:32,525 --> 00:32:33,025 off 897 00:32:33,484 --> 00:32:34,865 that we're getting into? 898 00:32:35,884 --> 00:32:37,085 Yeah. You know, the 899 00:32:37,789 --> 00:32:39,390 for for a while, there was a very 900 00:32:39,390 --> 00:32:40,529 legitimate critique 901 00:32:40,909 --> 00:32:42,750 of crypto that it was using a ton 902 00:32:42,750 --> 00:32:45,009 of power to, you know, 903 00:32:45,470 --> 00:32:48,269 just generate financial value. It wasn't doing anything 904 00:32:48,269 --> 00:32:48,769 else. 905 00:32:49,309 --> 00:32:51,224 And at this point, you know, the the 906 00:32:51,224 --> 00:32:54,184 carbon footprint and the energy load of modern 907 00:32:54,184 --> 00:32:55,244 blockchain systems 908 00:32:55,705 --> 00:32:57,945 is at par or less than your typical 909 00:32:57,945 --> 00:33:00,424 web two system out there, which is not 910 00:33:00,424 --> 00:33:02,424 to say it's nothing. Right? Obviously, we we 911 00:33:02,664 --> 00:33:04,125 you know, Spotify and 912 00:33:04,424 --> 00:33:05,964 Zoom and stuff still 913 00:33:06,529 --> 00:33:08,929 require a ton of electricity and a ton 914 00:33:08,929 --> 00:33:11,190 of water to use just simply because, 915 00:33:11,650 --> 00:33:12,630 no, you know, electricity 916 00:33:13,009 --> 00:33:13,509 is 917 00:33:13,970 --> 00:33:16,309 largely still not nuclear today or renewables. 918 00:33:17,169 --> 00:33:18,929 But the one of the nice things about 919 00:33:18,929 --> 00:33:22,445 this is it brings, like, the crypto energy 920 00:33:22,445 --> 00:33:24,845 problem into the same bucket of problems as 921 00:33:24,845 --> 00:33:26,924 any other computer system. It's it's not sort 922 00:33:26,924 --> 00:33:29,005 of out there on its own anymore except 923 00:33:29,005 --> 00:33:30,144 in the case of Bitcoin. 924 00:33:31,085 --> 00:33:32,605 And so the nice thing is is that 925 00:33:32,605 --> 00:33:33,825 you see the same characteristics 926 00:33:34,205 --> 00:33:36,144 that you see everywhere else, which is that 927 00:33:36,259 --> 00:33:38,259 people like their money, and they wanna send 928 00:33:38,259 --> 00:33:40,920 less of it on bills, like server bills. 929 00:33:41,140 --> 00:33:42,759 And so you see people voluntarily 930 00:33:43,059 --> 00:33:45,779 moving to areas where data centers have lower 931 00:33:45,779 --> 00:33:46,680 cost of power 932 00:33:46,980 --> 00:33:48,500 or, you know, for any other number of 933 00:33:48,500 --> 00:33:50,734 reasons, the the cost to actually run the 934 00:33:50,734 --> 00:33:53,315 data center is less. And so, you know, 935 00:33:53,615 --> 00:33:55,694 at some point, this is a problem that 936 00:33:55,694 --> 00:33:58,015 sort of capitalism does a pretty okay job 937 00:33:58,015 --> 00:33:58,674 at solving 938 00:34:00,015 --> 00:34:01,474 marginal cost. And so, 939 00:34:01,775 --> 00:34:04,355 you know, the most expensive single quantity 940 00:34:05,690 --> 00:34:08,329 of cost in a data center is the 941 00:34:08,329 --> 00:34:09,630 power and cooling bill. 942 00:34:10,170 --> 00:34:11,849 Right? It's it's the water bill and it's 943 00:34:11,849 --> 00:34:13,550 the electric bill. And so 944 00:34:13,849 --> 00:34:16,030 everyone is working to reduce those things, 945 00:34:16,570 --> 00:34:18,650 not just because they care about the planet, 946 00:34:18,650 --> 00:34:20,590 but because they care about their bottom line. 947 00:34:22,135 --> 00:34:23,835 Interesting. Yeah. Yeah. 948 00:34:24,775 --> 00:34:27,255 Okay. So I guess another question couple of 949 00:34:27,255 --> 00:34:28,775 courses I have are more on the crypto 950 00:34:28,775 --> 00:34:30,714 side, perhaps, which are things like Sure. 951 00:34:31,335 --> 00:34:33,594 How do you prevent Byzantine attacks 952 00:34:34,054 --> 00:34:36,219 against this kind of a system? Because that 953 00:34:36,219 --> 00:34:37,739 would seem to me to be a major 954 00:34:37,739 --> 00:34:38,239 problem, 955 00:34:39,500 --> 00:34:41,599 unless it's just that it's a shared pool. 956 00:34:41,820 --> 00:34:44,160 It's like No. No. Exactly. So so 957 00:34:44,539 --> 00:34:46,699 Byzantine fault tolerance is one of, like, the 958 00:34:46,699 --> 00:34:47,599 core characteristics 959 00:34:48,380 --> 00:34:50,414 of of block And so there's a ton 960 00:34:50,414 --> 00:34:53,235 of work on what's basically called consensus mechanism 961 00:34:53,295 --> 00:34:53,795 design. 962 00:34:54,175 --> 00:34:55,315 How does the network 963 00:34:56,815 --> 00:35:00,015 establish enough confidence to agree that it's it's 964 00:35:00,015 --> 00:35:01,295 gotten to the point where it can move 965 00:35:01,295 --> 00:35:03,179 on to doing the next thing? And so 966 00:35:03,179 --> 00:35:06,059 there's many different algorithms, and there's many different 967 00:35:06,059 --> 00:35:09,599 protocols out there for how consensus actually works. 968 00:35:10,219 --> 00:35:11,199 And they're optimized 969 00:35:11,579 --> 00:35:13,500 for different types of things. So some of 970 00:35:13,500 --> 00:35:15,244 them, they're optimized for assuming, 971 00:35:15,625 --> 00:35:18,265 like, if most people are good actors, how 972 00:35:18,265 --> 00:35:19,985 fast can we run the system? Right? That 973 00:35:20,105 --> 00:35:22,605 that's sort of an optimistic system that says, 974 00:35:22,664 --> 00:35:24,505 we still do all these checks and balances 975 00:35:24,505 --> 00:35:25,945 and these sorts of things. But if we 976 00:35:25,945 --> 00:35:28,409 assume that at any given point, very few 977 00:35:28,409 --> 00:35:30,730 people are intentionally misbehaving, we can just run 978 00:35:30,730 --> 00:35:33,210 the system a lot quicker. There's also systems 979 00:35:33,210 --> 00:35:35,210 that say, you know, above all else, we 980 00:35:35,210 --> 00:35:37,369 wanna make sure is that we have a 981 00:35:37,369 --> 00:35:38,510 100% certainty 982 00:35:38,889 --> 00:35:40,650 before we move on to anything else. And 983 00:35:40,650 --> 00:35:43,065 those systems usually run more slowly. 984 00:35:43,784 --> 00:35:45,704 So Solana, for example, is a good is 985 00:35:45,704 --> 00:35:47,545 a good example of this. Right now, there's 986 00:35:47,545 --> 00:35:50,045 something called confirmation versus finality. 987 00:35:50,505 --> 00:35:53,885 And finality is, like, full Byzantine fault tolerance. 988 00:35:54,025 --> 00:35:56,539 This will never like, there is a 0% 989 00:35:56,539 --> 00:35:58,559 chance this thing can ever get rewound 990 00:35:59,099 --> 00:36:01,420 unless, of course, everyone agrees to undo it, 991 00:36:01,420 --> 00:36:02,880 which is, like, you know, 992 00:36:03,739 --> 00:36:05,519 a problem you can't solve with code. 993 00:36:06,619 --> 00:36:08,699 But the confirmation times are, you know, four 994 00:36:08,699 --> 00:36:10,320 hundred milliseconds or less. 995 00:36:10,775 --> 00:36:13,195 But it takes a while to get basically 996 00:36:13,335 --> 00:36:15,574 all enough votes in that the thing is 997 00:36:15,574 --> 00:36:18,055 considered fully settled. So Solana right now is 998 00:36:18,055 --> 00:36:18,535 going through an 999 00:36:19,974 --> 00:36:20,474 consensus 1000 00:36:20,775 --> 00:36:21,275 mall, 1001 00:36:21,815 --> 00:36:23,195 on something called AlpenGlow. 1002 00:36:23,960 --> 00:36:26,839 And AlpinGlow is gonna bring that time to 1003 00:36:26,839 --> 00:36:29,579 finality down from twelve seconds to anywhere between 1004 00:36:29,880 --> 00:36:31,400 one and a half seconds and a hundred 1005 00:36:31,400 --> 00:36:32,380 and fifty milliseconds. 1006 00:36:33,799 --> 00:36:35,159 So this is kind of like as we 1007 00:36:35,159 --> 00:36:37,579 look at the the problems of scaling blockchain 1008 00:36:38,445 --> 00:36:38,945 for 1009 00:36:39,244 --> 00:36:42,144 pretty much the first ten years of blockchain, 1010 00:36:42,285 --> 00:36:44,285 it was all software problems. It was that 1011 00:36:44,285 --> 00:36:46,445 the software code wasn't good enough. People hadn't 1012 00:36:46,445 --> 00:36:48,545 come up with smart enough consensus algorithms. 1013 00:36:49,005 --> 00:36:50,684 It wasn't that it was limited by the 1014 00:36:50,684 --> 00:36:52,204 power of the hardware because there's a lot 1015 00:36:52,204 --> 00:36:54,269 of money crypto now. Right? You you can 1016 00:36:54,269 --> 00:36:56,430 buy just as powerful servers to run stuff 1017 00:36:56,430 --> 00:36:57,890 as you can in any industry. 1018 00:36:58,989 --> 00:37:01,010 It was the software was not, like, 1019 00:37:01,390 --> 00:37:02,989 fast enough and built in a way that 1020 00:37:02,989 --> 00:37:05,150 it could be performance enough. And now that 1021 00:37:05,150 --> 00:37:07,454 it's no longer a software problem, like, that's 1022 00:37:07,454 --> 00:37:09,474 that's where we come in at double zero. 1023 00:37:09,614 --> 00:37:10,594 Awesome. Cool. 1024 00:37:10,974 --> 00:37:13,215 Alright. Well, I can't think of much else 1025 00:37:13,215 --> 00:37:15,315 to cover on this topic, so I'll 1026 00:37:15,695 --> 00:37:17,454 let you is there anything in Austin that 1027 00:37:17,454 --> 00:37:19,295 you'd like to talk about before we wrap 1028 00:37:19,295 --> 00:37:19,795 up? 1029 00:37:21,420 --> 00:37:23,420 I would say that there's, like for anyone 1030 00:37:23,420 --> 00:37:25,280 who's interested in this space, 1031 00:37:25,820 --> 00:37:28,400 there's kind of greenfield opportunity here. 1032 00:37:28,940 --> 00:37:31,579 You know, networking has not historically been a 1033 00:37:31,579 --> 00:37:34,059 particularly big part of blockchain. It has sort 1034 00:37:34,059 --> 00:37:35,659 of always operated in this kind of, like, 1035 00:37:35,659 --> 00:37:37,025 magical magical thinking land 1036 00:37:37,405 --> 00:37:39,804 of, like, oh, there's infinite bandwidth. Latency doesn't 1037 00:37:39,804 --> 00:37:42,525 matter. Like, there's infinite compute because it's a 1038 00:37:42,525 --> 00:37:45,324 bunch of software people building stuff. It's not 1039 00:37:45,324 --> 00:37:47,664 hardware people or networking people building stuff. 1040 00:37:48,364 --> 00:37:48,864 But 1041 00:37:49,219 --> 00:37:51,219 things are really, like, starting to change, I 1042 00:37:51,219 --> 00:37:53,059 would say. Like, you know, what I would 1043 00:37:53,139 --> 00:37:55,380 you know, of our engineering team, 90% of 1044 00:37:55,380 --> 00:37:57,079 them are network engineers. 1045 00:37:57,619 --> 00:37:59,380 Like, only, you know, a small percentage of 1046 00:37:59,380 --> 00:38:01,799 them actually, like, come from crypto originally. 1047 00:38:02,420 --> 00:38:04,295 And we're seeing this across more and more 1048 00:38:04,295 --> 00:38:04,795 protocols, 1049 00:38:05,174 --> 00:38:07,034 is that the the folks who are 1050 00:38:07,335 --> 00:38:09,335 able to make some of the most impacts 1051 00:38:09,335 --> 00:38:11,574 now, they're not cryptographers. They're not people who 1052 00:38:11,574 --> 00:38:14,154 are writing smart contracts or or programming languages. 1053 00:38:14,454 --> 00:38:16,329 They're actually on the networking stack And people 1054 00:38:16,329 --> 00:38:18,650 who can think smartly about data flow and 1055 00:38:18,650 --> 00:38:20,489 how it moves through both the hardware and 1056 00:38:20,489 --> 00:38:21,389 software system, 1057 00:38:22,089 --> 00:38:23,769 those are the folks that nowadays are making 1058 00:38:23,769 --> 00:38:24,989 some of the largest contributions 1059 00:38:25,289 --> 00:38:26,269 to to crypto. 1060 00:38:26,650 --> 00:38:29,394 Cool. So, Tom, any other questions? Any other 1061 00:38:29,394 --> 00:38:31,174 thoughts before we wrap up? 1062 00:38:32,034 --> 00:38:33,414 No. No. It's been great. 1063 00:38:33,795 --> 00:38:35,315 So, Tom, where can people get in touch 1064 00:38:35,315 --> 00:38:36,694 with you if they want to? 1065 00:38:37,315 --> 00:38:39,315 They can find me on LinkedIn. He actually 1066 00:38:39,315 --> 00:38:40,994 said a whole sentence and not just a 1067 00:38:40,994 --> 00:38:43,090 single word. Yeah. I'm I'm amazed. 1068 00:38:44,269 --> 00:38:46,349 So, Austin, I assume you blog or you 1069 00:38:46,349 --> 00:38:48,769 have regular stuff that you do? 1070 00:38:49,230 --> 00:38:50,690 Working people We do. Yeah. 1071 00:38:51,309 --> 00:38:53,329 Double 0 dot xyz. 1072 00:38:53,630 --> 00:38:55,484 You can see all of our our our 1073 00:38:55,484 --> 00:38:57,005 blogs and write ups and kind of read 1074 00:38:57,005 --> 00:38:59,025 our our white paper there and follow along. 1075 00:38:59,244 --> 00:39:01,184 There's a link to our GitHub as well, 1076 00:39:01,565 --> 00:39:03,565 and we're pretty active on Twitter as well. 1077 00:39:03,565 --> 00:39:05,484 Cool. Alright. I'm Russ White. You can always 1078 00:39:05,484 --> 00:39:06,764 find me here at the hedge at rule 1079 00:39:06,764 --> 00:39:07,824 eleven dot tech. 1080 00:39:09,070 --> 00:39:11,469 X slash Twitter, whatever. I'm really easy to 1081 00:39:11,469 --> 00:39:13,469 find. Just look for me. I'm around some 1082 00:39:13,469 --> 00:39:14,449 place or another. 1083 00:39:15,469 --> 00:39:15,969 So 1084 00:39:16,510 --> 00:39:18,109 we know that we live in an attention 1085 00:39:18,109 --> 00:39:19,010 driven economy 1086 00:39:19,309 --> 00:39:21,869 and that your time and your attention are 1087 00:39:21,869 --> 00:39:24,964 valuable resources, and we're happy you spent this 1088 00:39:25,105 --> 00:39:27,844 time with us talking about blockchain and blockchain 1089 00:39:28,704 --> 00:39:31,184 scaling and the new ways that we're looking 1090 00:39:31,184 --> 00:39:31,924 at blockchain. 1091 00:39:32,864 --> 00:39:34,384 Thank you for listening all the way to 1092 00:39:34,384 --> 00:39:35,519 the bitter end, and we will 1093 00:39:42,960 --> 00:39:43,460 catch 1094 00:39:47,360 --> 00:39:48,110 you next