Roundtable_025 - This Time, It's Actually Different の全文書き起こし

well. 67 00:04:03,040 --> 00:04:05,440 You're never coming on this show again, you piece of shit. 68 00:04:09,460 --> 00:04:10,980 Oh my God, welcome to the show, man. 69 00:04:11,760 --> 00:04:12,680 It's a good format. 70 00:04:12,820 --> 00:04:13,260 I like it. 71 00:04:13,780 --> 00:04:14,060 We'll receive it. 72 00:04:14,060 --> 00:04:14,640 Oh, it's fun. 73 00:04:14,920 --> 00:04:15,340 It's fun. 74 00:04:15,460 --> 00:04:16,860 It's just shooting the shit. 75 00:04:17,280 --> 00:04:19,720 Everybody, the judgment. 76 00:04:19,900 --> 00:04:21,960 the judgment. Oh my God. 77 00:04:22,360 --> 00:04:24,800 The, I'm, I'm really kind of glad the whole spam warship. 78 00:04:25,000 --> 00:04:26,860 It's kind of, it's kind of over. 79 00:04:26,980 --> 00:04:28,920 At least it's subsided for a little bit. 80 00:04:29,140 --> 00:04:32,300 Cause we get a reprieve. Holy crap. Holy crap. 81 00:04:32,300 --> 00:04:36,340 I was getting so sick of that. Um, from every angle. Uh, 82 00:04:36,340 --> 00:04:39,520 but, uh, uh, we got some really fun freedom tech stuff. 83 00:04:39,700 --> 00:04:41,720 I want to first, 84 00:04:41,880 --> 00:04:46,080 first we'll do our usual go rounds and just kind of get an update. Uh, 85 00:04:46,100 --> 00:04:49,600 and then I want to get into white noise just because I've been finally playing 86 00:04:49,600 --> 00:04:53,600 around with it. But Steve, how you doing boss? 87 00:04:53,600 --> 00:04:57,600 Steve McLaughlin I'm doing good man. I'm glad to have Max here. 88 00:04:57,600 --> 00:05:01,600 Shout out to Max for maintaining 89 00:05:01,600 --> 00:05:05,600 the hard challenge of maintaining a good reputation in Bitcoin 90 00:05:05,600 --> 00:05:18,257 for over a decade of being a public figure I was thinking about that when you said Max came on the show I was like man you know what like you have all these fallen heroes like max has been a hero of mine for like a decade continuously 91 00:05:18,257 --> 00:05:24,337 so you got probably two years i'd have him two years left another pressure is on man 92 00:05:24,337 --> 00:05:32,437 but no a uh an update on me um earlier today i was it occurred to me that i have lived my adult 93 00:05:32,437 --> 00:05:41,097 life in reverse. Give me like one minute to explain here. When I was 23, I moved to Florida, 94 00:05:41,717 --> 00:05:50,257 bought a house on a golf course, had a government job where I was like retired in place. 23. 95 00:05:51,077 --> 00:06:00,017 Two decades later, I'm now working at a bar, recently married, starting a family in a two 96 00:06:00,017 --> 00:06:08,977 bedroom apartment with a 26 year old car amazing i'm thinking i'm thinking like you know i never 97 00:06:08,977 --> 00:06:14,877 graduated from high school i got expelled i'm thinking maybe you know next decade i might try 98 00:06:14,877 --> 00:06:20,337 to graduate from high school is that is that true you got expelled from high school how'd you get 99 00:06:20,337 --> 00:06:27,617 how'd you

work at nasa oh long story dude i so i got expelled my senior year of high school 100 00:06:27,617 --> 00:06:30,537 for like hacking into the intercom system, 101 00:06:30,697 --> 00:06:32,897 calling the principal asshole, et cetera. 102 00:06:33,097 --> 00:06:35,837 But like I drove to college anyway. 103 00:06:35,837 --> 00:06:37,717 That's the best reason to be expelled ever. 104 00:06:38,817 --> 00:06:41,757 That's a, I knew we were friends for a reason. 105 00:06:42,657 --> 00:06:42,977 Yeah. 106 00:06:43,317 --> 00:06:44,557 I'd already gotten to college. 107 00:06:44,557 --> 00:06:47,837 So I just drove to college, which was like four hours away. 108 00:06:47,857 --> 00:06:48,257 Oh, okay. 109 00:06:48,497 --> 00:06:52,137 And I hope they didn't hear that I never graduated from high school. 110 00:06:52,257 --> 00:06:53,217 And they didn't. 111 00:06:53,597 --> 00:06:54,037 Yes. 112 00:06:54,197 --> 00:06:55,317 They just let me in. 113 00:06:55,317 --> 00:07:02,217 they never found out that i never graduated from high school that's amazing that is beautiful that's 114 00:07:02,217 --> 00:07:08,517 the most beautiful story ever that's the best thing i hacked into the intercom system and called 115 00:07:08,517 --> 00:07:16,157 oh yeah principal and swartz uh andrew schwartz i've been a hacker since you had to like you know 116 00:07:16,157 --> 00:07:24,137 break into the the school and wire wire wires or like you know put wires together by hand so you 117 00:07:24,137 --> 00:07:28,117 had your own microphone in there and old school hacking stuff, dude. 118 00:07:28,217 --> 00:07:28,757 That's amazing. 119 00:07:28,857 --> 00:07:34,117 I wish so bad I had been into this or kind of like understood the value of it back then 120 00:07:34,117 --> 00:07:35,677 because I would have had so much fun. 121 00:07:35,677 --> 00:07:41,777 I was more kind of, you know, just like internet torrent junkie sort of focus 122 00:07:41,777 --> 00:07:43,857 rather than like actual hacking. 123 00:07:44,437 --> 00:07:49,577 I would do just like hardware workshop stuff on that side of things. 124 00:07:49,637 --> 00:07:52,137 And I wish I could have married them way back when 125 00:07:52,137 --> 00:07:54,577 because now I'm getting into it and it's just fun as shit. 126 00:07:55,477 --> 00:07:56,637 Oh, it's so much fun now. 127 00:07:57,097 --> 00:07:59,277 The Red's Cable, 128 00:07:59,517 --> 00:08:03,417 the guy that ran the cable provider for our little town, 129 00:08:04,157 --> 00:08:06,397 actually told us one time on the phone 130 00:08:06,397 --> 00:08:12,677 that we were his heaviest users on the whole network. 131 00:08:12,877 --> 00:08:13,837 And we were in middle school. 132 00:08:14,257 --> 00:08:16,337 So he was like middle school or high school. 133 00:08:19,337 --> 00:08:21,277 Dude, I... 134 00:08:22,137 --> 00:08:33,517 When I was working for Time Warner and Spectrum as a technician, you can look up nodes and see like account or like device basically. 135 00:08:34,317 --> 00:08:40,317 But you can see like subsections on the network and where traffic and bandwidth and all that stuff goes. 136 00:08:41,357 --> 00:08:47,657 Big thing for optimization and where to like focus on cleaning up noise and everything on a broadband network. 137 00:08:47,657 --> 00:08:49,297 but 138 00:08:49,297 --> 00:08:52,677 dude I would just occasionally look mine up 139 00:08:52,677 --> 00:08:54,857 just to see like in comparison 140 00:08:54,857 --> 00:08:56,377 and this is like even way back then 141 00:08:56,377 --> 00:08:58,597 I was running like two bitcoin nodes 142 00:08:58,597 --> 00:09:00,737 and you know all this stuff so just like constant 143 00:09:00,737 --> 00:09:02,337 traffic and 144 00:09:02,337 --> 00:09:04,657 holy crap I just fucked up 145 00:09:04,657 --> 00:09:06,377 that curve every time 146 00:09:06,377 --> 00:09:07,937 like it would look so 147 00:09:07,937 --> 00:09:09,957 stupid on the chart 148 00:09:09,957 --> 00:09:12,657 people are just like grandmas logging in 149 00:09:12,657 --> 00:09:14,677 trying to figure out how to check their email 150 00:09:14,677 --> 00:09:15,477 and stuff 151 00:09:15,477 --> 00:09:17,457 I'm just like 152 00:09:17,657 --> 00:09:20,417 All right, I'm downloading a torrent on this computer, 153 00:09:20,717 --> 00:09:22,637 and I got two Bitcoin nodes running, 154 00:09:22,917 --> 00:09:25,857 and I'm trying to stream this video over here. 155 00:09:26,577 --> 00:09:28,037 How do I figure out how to get... 156 00:09:28,037 --> 00:09:30,257 I don't think I have enough bandwidth left for Netflix. 157 00:09:32,777 --> 00:09:33,297 Max. 158 00:09:34,177 --> 00:09:35,217 Max, Max, Max. 159 00:09:36,017 --> 00:09:36,757 Welcome, man. 160 00:09:37,217 --> 00:09:38,077 What you got going on? 161 00:09:38,157 --> 00:09:41,217 What's new in your life and in Freedom Tech? 162 00:09:42,577 --> 00:09:44,097 Dude, it's busy. 163 00:09:44,097 --> 00:09:46,917 The AI side courses is real. 164 00:09:47,657 --> 00:09:52,997 Like, it's so hard to realize that you're not ambitious enough. 165 00:09:53,257 --> 00:09:57,417 Like, you can solve way harder problems than ever before. 166 00:09:57,917 --> 00:10:01,737 And more projects in parallel, which is so crazy. 167 00:10:02,797 --> 00:10:06,297 So, like, I mean, Highlight for me was releasing White Mowice kind of earlier this week. 168 00:10:06,297 --> 00:10:09,297 And it didn't break, finally, which is great. 169 00:10:10,077 --> 00:10:17,197 But then, you know, in parallel, I've been just building all those little apps that I wanted to have, but didn't have. 170 00:10:17,657 --> 00:10:32,953 And now it just like a side prompt while you working You know once an hour you check in and and just prompt the next thing and it works surprisingly well man it is i think we we have so much customizable 171 00:10:32,953 --> 00:10:41,033 cool freedom tech software that just does what you want it to and nothing else and you just

use 172 00:10:41,033 --> 00:10:46,873 it and if if someone else uses you're cool but you don't really have to and and that's so wild 173 00:10:46,873 --> 00:10:49,933 I'm still kind of grappling with that situation of 174 00:10:49,933 --> 00:10:51,873 pick even bigger problems 175 00:10:51,873 --> 00:10:54,073 because we can literally do that 176 00:10:54,073 --> 00:10:55,713 white noise is the perfect example 177 00:10:55,713 --> 00:10:58,113 decentralized encrypted 178 00:10:58,113 --> 00:11:00,173 group messaging without a central server 179 00:11:00,173 --> 00:11:01,893 is almost impossible 180 00:11:01,893 --> 00:11:02,573 it's like 181 00:11:02,573 --> 00:11:06,193 not quite but like bitcoin level impossible 182 00:11:06,193 --> 00:11:07,933 like a decentralized money 183 00:11:07,933 --> 00:11:09,213 without a privileged server 184 00:11:09,213 --> 00:11:11,433 it's a consensus problem, it's really fucking hard 185 00:11:11,433 --> 00:11:14,013 and group communication is kind of 186 00:11:14,013 --> 00:11:14,373 the same 187 00:11:15,033 --> 00:11:24,953 If we all want to encrypt our messages and rotate keys so that if one leaks, you don't leak all the messages, then we need to be in consensus on what key do we encrypt the next message to. 188 00:11:25,593 --> 00:11:27,533 And we don't have a central server. 189 00:11:27,713 --> 00:11:31,613 We don't have a timestamping server that can order our messages and make the stuff easy. 190 00:11:32,493 --> 00:11:36,293 And so it's just really fundamentally a damn hard computer science problem. 191 00:11:36,833 --> 00:11:41,073 And it's been like two and a half years since we worked on that. 192 00:11:41,073 --> 00:11:47,893 And initially we thought, we have MLS, that's the cool encryption scheme, we can use it, it solves all our problems. 193 00:11:47,893 --> 00:11:50,113 And then you realize, oh no, actually it doesn't. 194 00:11:50,113 --> 00:11:57,293 And then we've tried with amazing engineers, real hardcore Ruby on Rails developers. 195 00:11:57,293 --> 00:11:59,093 They love to hand chisel the code. 196 00:11:59,093 --> 00:12:06,533 They're really in it for the craft and spend a year or whatnot on trying to make it work. 197 00:12:06,533 --> 00:12:08,333 And they don't get it to work. 198 00:12:08,333 --> 00:12:14,873 and then a couple months later you just give it to an agent and be like here's everything we learned 199 00:12:14,873 --> 00:12:21,733 here's like what we actually want to do what the fuck do we do and the agent cranks for a bit and 200 00:12:21,733 --> 00:12:25,513 at the end gives you a spec that's way more thorough than anything that you could have ever 201 00:12:25,513 --> 00:12:31,313 come up with basically and and writes you the library that implements it and and the new native 202 00:12:31,313 --> 00:12:37,653 front-end clients that make it pretty so it's it's really like the last six months especially 203 00:12:37,653 --> 00:12:44,393 We started this Marmot version 2 sprint in April or May. 204 00:12:44,593 --> 00:12:45,073 I think May. 205 00:12:46,133 --> 00:12:47,893 And so that's not even been that long. 206 00:12:47,993 --> 00:12:51,653 But the models were amazing already back then and just have improved. 207 00:12:52,453 --> 00:12:57,593 But around that time, middle, early of this year, is where the models really went next level. 208 00:12:57,973 --> 00:12:58,853 And the harnesses too. 209 00:12:59,653 --> 00:13:04,093 And holy shit, you can really solve almost impossible problems now. 210 00:13:04,793 --> 00:13:06,233 And we're kind of done. 211 00:13:06,233 --> 00:13:11,733 I mean, we have a shit ton of polish to do and a couple more features and stuff, but the hard part is solved. 212 00:13:12,433 --> 00:13:17,733 And I'm thinking, okay, what other hard problems do we actually need to solve and can now solve? 213 00:13:18,553 --> 00:13:19,733 Fucking crazy times. 214 00:13:20,433 --> 00:13:21,393 Dude, it so is. 215 00:13:22,033 --> 00:13:26,953 White noise is fascinating from the context of how it uses MLS signatures. 216 00:13:27,193 --> 00:13:31,613 But I didn't get, you kind of brought it up and suggested this. 217 00:13:32,593 --> 00:13:35,373 Actually, before I get into this, I want to mention something that you said. 218 00:13:35,373 --> 00:13:38,033 the last time we were on the show together, 219 00:13:38,433 --> 00:13:40,933 you said something that kind of stuck with me 220 00:13:40,933 --> 00:13:42,813 and I just thought it was so crazy. 221 00:13:43,633 --> 00:13:46,693 And every single time I'm like building my own thing, 222 00:13:46,893 --> 00:13:51,253 the editing, the recording and video 223 00:13:51,253 --> 00:13:52,753 and audio editing app 224 00:13:52,753 --> 00:13:54,213 that I've just spent the last three days 225 00:13:54,213 --> 00:13:56,493 kind of like fine tuning and getting right. 226 00:13:57,193 --> 00:13:57,793 That's it. 227 00:13:57,913 --> 00:14:00,553 I've just been vibe coding for three days on this thing 228 00:14:00,553 --> 00:14:02,813 and I've been using it for the last two days 229 00:14:02,813 --> 00:14:04,073 and just kind of tweaking things 230 00:14:04,073 --> 00:14:07,453 and adding little things that I need to make it fit better into my workflow. 231 00:14:07,453 --> 00:14:14,513 But you said that probably we will move into a place in the not too distant future 232 00:14:14,513 --> 00:14:19,213 where the average number of users for a specific piece of software is one. 233 00:14:20,593 --> 00:14:24,833 Where everybody is literally just creating their own software. 234 00:14:25,533 --> 00:14:29,573 And it was so funny because I thought about that day before yesterday 235 00:14:29,573 --> 00:14:33,233 when I started building this thing because it built out so quick. 236 00:14:33,233 --> 00:14:37,533 It's so quick from like nothing to like, holy shit, I'm just sitting here recording stuff. 237 00:14:39,053 --> 00:14:53,013 And we kind of got the engine ready

to turn this like peer to peer and like across all of my devices so that I can do the thing that I've mentioned before the show of like getting up, taking my phone and go out and recording thing, come back to my desktop, record again into the same timeline, all that good stuff. 238 00:14:53,533 --> 00:14:58,673 And then invite my producer in so that he can do some edits or do some things as well. 239 00:14:59,633 --> 00:15:01,953 I really don't think we're very far away from that. 240 00:15:01,953 --> 00:15:11,193 um and uh uh but i was sitting here like messing around with that and like tweaking some things 241 00:15:11,193 --> 00:15:20,913 and i've got like you know four agents plugged into kimmy and i've got like six remote code sessions 242 00:15:20,913 --> 00:15:29,933 in claude app thing and i'm like sitting here using six different apps and four of them are 243 00:15:29,933 --> 00:15:30,773 apps that I built. 244 00:15:31,873 --> 00:15:32,233 Yeah, right. 245 00:15:32,833 --> 00:15:49,030 You know and like the way the way that I getting I just have like a running set of log files which are just Claude conversation and output on my MacBook And then I have the exact same thing running on my Mac mini with Kimmy And they talk to each other 246 00:15:49,030 --> 00:15:52,470 by just putting it in the same Paradrive folder. 247 00:15:53,250 --> 00:15:55,810 So they just, one of them puts their output 248 00:15:55,810 --> 00:15:58,170 into the Paradrive folder and then Kimmy reads it 249 00:15:58,170 --> 00:16:00,130 and put theirs into the Paradrive folder 250 00:16:00,130 --> 00:16:01,290 and then Claude reads it. 251 00:16:01,330 --> 00:16:04,090 So I've kind of got like a little agent chat bot, 252 00:16:04,250 --> 00:16:07,190 which is just syncing a folder with some text files. 253 00:16:08,710 --> 00:16:11,030 And like, that's how I get Paradrive 254 00:16:11,030 --> 00:16:12,270 to all of my computers is Paradrive. 255 00:16:12,750 --> 00:16:13,970 You know, I update the new versions, 256 00:16:13,970 --> 00:16:15,110 just I Paradrive it. 257 00:16:15,190 --> 00:16:18,090 It literally automatically puts it in the delivery folder 258 00:16:18,090 --> 00:16:22,230 and then all of them download and install the new version like that. 259 00:16:23,930 --> 00:16:29,270 Yeah, you know, like that aspect of you build the app from inside the app. 260 00:16:29,850 --> 00:16:31,730 You know, you use PairDrive to build PairDrive. 261 00:16:32,150 --> 00:16:33,150 I have the same with White Noise. 262 00:16:33,930 --> 00:16:35,830 White Noise is my agent interface now. 263 00:16:36,450 --> 00:16:41,510 And I talk to OpenCode, Hermes, OpenClaw, CloudCode, Codex, whatever. 264 00:16:42,010 --> 00:16:43,650 It's just all in the same interface for me. 265 00:16:43,750 --> 00:16:45,890 And I tell him what's broken about the app. 266 00:16:45,890 --> 00:16:47,450 He goes and fixes it. 267 00:16:47,450 --> 00:16:53,810 He builds the APK, makes a pull request,

sends me the APK to test, and I tell him to merge it, basically. 268 00:16:54,090 --> 00:17:05,330 So it's this feedback loop where you can customize your own setup and keep fine-tuning it, and then it just works awesome. 269 00:17:05,870 --> 00:17:09,050 There's at least one user that's excited about it and is using it all the time. 270 00:17:09,510 --> 00:17:13,990 And so it will keep getting better as long as you keep exciting and keep building on it. 271 00:17:13,990 --> 00:17:18,770 I am curious how you did the key delegation stuff 272 00:17:18,770 --> 00:17:21,850 because we talked about this 273 00:17:21,850 --> 00:17:24,430 and then also talked about Hyper-B 274 00:17:24,430 --> 00:17:26,770 as being a great way to sync 275 00:17:26,770 --> 00:17:30,710 the potential status of the latest key 276 00:17:30,710 --> 00:17:33,230 or the rotating key or whatnot. 277 00:17:34,930 --> 00:17:36,910 And this is kind of the point where 278 00:17:36,910 --> 00:17:38,550 PairDrive has really been focused on 279 00:17:38,550 --> 00:17:40,350 just getting devices together, 280 00:17:41,030 --> 00:17:41,890 creating an index, 281 00:17:42,090 --> 00:17:43,970 syncing the diffs in that index, 282 00:17:43,990 --> 00:17:45,990 and like, you know, checking its hashes. 283 00:17:46,090 --> 00:17:47,130 It's basically been like, okay, 284 00:17:47,130 --> 00:17:49,370 how do you sync constant changes in a torrent 285 00:17:49,370 --> 00:17:53,390 and have that torrent be a giant collection of files? 286 00:17:54,030 --> 00:17:55,290 And how do you do that efficiently? 287 00:17:55,850 --> 00:17:58,370 Because hyperdrives, when we used it natively 288 00:17:58,370 --> 00:18:00,510 in the HyperCore protocol, 289 00:18:00,650 --> 00:18:03,310 hyperdrives were just a slog. 290 00:18:03,670 --> 00:18:06,270 Like HyperB and like the HyperSwarm, 291 00:18:06,450 --> 00:18:08,490 that stuff has worked amazingly. 292 00:18:08,950 --> 00:18:12,230 And hyperdrive works great on like a one-off type thing. 293 00:18:12,230 --> 00:18:22,870 But if I'm trying to share 100,000 files from my Linux machine to my MacBook, which is child's play for what I have over there, it's just a it's it's dead in the water. 294 00:18:22,990 --> 00:18:28,590 Like I'm just it's using three times what my brave browser is with 40 tabs open. 295 00:18:28,810 --> 00:18:30,230 You know, it's just churning. 296 00:18:30,230 --> 00:18:43,390 And so that was our main focus is how do we do that in a way that makes sense and doesn't just take gargantuan amounts of resources and duplicates of files and all of this stuff. 297 00:18:44,290 --> 00:18:49,170 But now we're in this, okay, how do we log in multiple devices on one account? 298 00:18:49,650 --> 00:18:55,650 How do we hand over keys or say this key is my key and sort all that out? 299 00:18:55,650 --> 00:19:00,070 And that turns out to be its own extremely hard problem. 300 00:19:00,730 --> 00:19:05,110 And I know y'all have a clever setup in White Noise, but I haven't dug into it. 301 00:19:06,890 --> 00:19:09,370 Yeah, this is a hard problem, really. 302 00:19:09,610 --> 00:19:11,110 And especially to do it at scale. 303 00:19:12,070 --> 00:19:20,330 It's easy to have the prototype that can do a couple files, but to do the full file system and terabytes of data, that's a different type of beast. 304 00:19:20,330 --> 00:19:20,770 Yeah. 305 00:19:20,770 --> 00:19:27,450 So for WhiteNoise or the Marmot protocol, there's, I guess, two things. 306 00:19:27,550 --> 00:19:30,170 There's the MLS, Messaging Layer Security. 307 00:19:30,630 --> 00:19:36,690 It's an IETF standard for basically creating a group encryption system, 308 00:19:37,010 --> 00:19:43,050 where the idea is that everyone in a group knows the same symmetric key 309 00:19:43,050 --> 00:19:46,090 to encrypt the messages to and to then later decrypt them. 310 00:19:46,350 --> 00:19:50,690 And the idea of secure encryption is that we rotate keys. 311 00:19:50,770 --> 00:19:53,250 to which we actually encrypt our data to. 312 00:19:53,990 --> 00:19:56,250 And that is part of the MLS spec. 313 00:19:57,690 --> 00:19:59,950 Probably, I'm not sure how deep we should go into it, 314 00:20:00,030 --> 00:20:01,730 in part because I don't understand it that deeply. 315 00:20:01,890 --> 00:20:03,970 Like, MLS is insanely complicated. 316 00:20:04,390 --> 00:20:06,430 The specification is 170 pages. 317 00:20:07,250 --> 00:20:09,270 Like, it's kind of wild. 318 00:20:10,810 --> 00:20:14,410 But ultimately, whenever you add a member to a group, 319 00:20:14,410 --> 00:20:16,190 or you remove a member from the group, 320 00:20:16,450 --> 00:20:18,610 or you change the group name, 321 00:20:18,610 --> 00:20:21,090 or when people do a self-update, 322 00:20:22,210 --> 00:20:23,990 we rotate this key. 323 00:20:24,810 --> 00:20:27,950 And this is like a whole kind of state machine 324 00:20:27,950 --> 00:20:30,870 that gets you to a new set of key 325 00:20:30,870 --> 00:20:34,290 and that's then distributed to everyone else. 326 00:20:34,470 --> 00:20:35,930 So you create a new key 327 00:20:35,930 --> 00:20:37,870 and you encrypt it to the keys 328 00:20:37,870 --> 00:20:40,390 that the previous persons in the group know. 329 00:20:40,790 --> 00:20:42,250 And everyone kind of has his own key 330 00:20:42,250 --> 00:20:43,670 that gets it encrypted too. 331 00:20:45,190 --> 00:20:47,930 But it uses Merkle tree space. 332 00:20:47,930 --> 00:20:49,990 are not quite Merkle trees, but binary trees 333 00:20:50,011 --> 00:20:51,531 which makes it super efficient. 334 00:20:52,011 --> 00:20:55,511 So in theory, you could have a million people in a group 335 00:20:55,511 --> 00:20:57,551 and you can kick someone out 336 00:20:57,551 --> 00:21:01,351 and you only need to encrypt the new key like 10 times 337 00:21:01,351 --> 00:21:03,891 or whatever O log N is at a million, 338 00:21:04,111 --> 00:21:05,131 I think 13 or something. 339

00:21:05,631 --> 00:21:09,751 So really efficient to distribute a new symmetric key 340 00:21:09,751 --> 00:21:11,071 to a whole bunch of people. 341 00:21:12,231 --> 00:21:13,931 And MLS solves this, 342 00:21:14,331 --> 00:21:15,451 which I'm super grateful for 343 00:21:15,451 --> 00:21:17,591 because that's gigabrain stuff. 344 00:21:18,471 --> 00:21:25,311 And what MLS does not solve is how do we know which key belongs to whom? 345 00:21:25,451 --> 00:21:26,811 Kind of web of trust, right? 346 00:21:26,951 --> 00:21:29,691 To some extent, they call it the authentication service. 347 00:21:30,491 --> 00:21:36,351 In the Signal case, it's like, how do you know which public key does this phone number use? 348 00:21:37,071 --> 00:21:41,631 And the answer is you go to the Signal server and says, hey, I want to message this phone number or this username. 349 00:21:42,011 --> 00:21:42,911 What's his public key? 350 00:21:43,791 --> 00:21:46,991 And that's a trusted entity to some extent. 351 00:21:46,991 --> 00:21:49,231 They've optimized a whole bunch there, but still. 352 00:21:50,191 --> 00:21:54,291 And the second important aspect is how do you actually get the messages to everyone? 353 00:21:55,131 --> 00:22:10,851 And here also, what's interesting, and we kind of underestimated this, by the way, for the first time in MLS, is that here, the delivery service, as they call it, is supposed to order the messages and give everyone the messages in the same order. 354 00:22:12,031 --> 00:22:15,591 And if you assume that, things are really straightforward. 355 00:22:15,591 --> 00:22:22,311 But if you want to put that delivery service away, then you have a whole time stamping problem, double spending problem, basically. 356 00:22:23,471 --> 00:22:24,771 But that's a different part of the story. 357 00:22:24,931 --> 00:22:30,251 So your actual question was more about the sub-key system, and that's the authentication service. 358 00:22:31,431 --> 00:22:37,591 And in order to decentralize the authentication service, we just moved the problem one layer deeper. 359 00:22:37,771 --> 00:22:43,591 We don't know who belongs to this key, so we use another key to sign off on it. 360 00:22:43,591 --> 00:22:57,811 And thankfully, with Nostr, we have a good, somewhat established web of trust system where people have their NPUPS and their NSECs and they've been shitposting on Primal or whatever since years. 361 00:22:57,811 --> 00:23:00,411 and now you can use the exact same key 362 00:23:00,411 --> 00:23:03,331 and sign in to the Marmot protocol 363 00:23:03,331 --> 00:23:04,691 or white noise as an app 364 00:23:04,691 --> 00:23:08,551 and that then will give you a signature 365 00:23:08,551 --> 00:23:10,471 on top of a new set of keys 366 00:23:10,471 --> 00:23:13,631 that you've just created for MLS specifically. 367 00:23:14,171 --> 00:23:16,391 So it's very similar to what PGP did 368 00:23:16,391 --> 00:23:18,591 with master keys and child keys 369 00:23:18,591 --> 00:23:21,511 and you could have multiple different child keys 370 00:23:21,511 --> 00:23:24,111 and stuff like this and so same idea. 371 00:23:25,051 --> 00:23:27,151 And there are even different elliptic curves, 372 00:23:27,151 --> 00:23:30,191 different encryption schemes or signature schemes. 373 00:23:31,151 --> 00:23:35,471 It uses the same like the ED25119 or whatever, 374 00:23:36,231 --> 00:23:38,791 similar to torrents and many DHTs. 375 00:23:40,171 --> 00:23:43,131 And another interesting thing is we initially were thinking, 376 00:23:43,691 --> 00:23:46,231 well, MLS has flexible Cypher suites, 377 00:23:46,231 --> 00:23:51,931 so we can slot out the proposed key scheme that they are doing 378 00:23:51,931 --> 00:23:53,911 and use a post-quantum system. 379 00:23:54,151 --> 00:23:55,951 Or we were thinking, let's use Schnorr. 380 00:23:55,951 --> 00:24:01,691 We use it in Bitcoin, we use it in Noster. We could actually use the Schnorr encryption scheme in MLS. 381 00:24:02,131 --> 00:24:06,251 That turns out to be way too much work and not really any benefit. 382 00:24:07,051 --> 00:24:11,711 And so the idea is, okay, we can just adopt a new encryption scheme, just have different algorithms. 383 00:24:12,391 --> 00:24:20,771 But your Noster Schnorr public key gives you a signature over the subkey that is whatever scheme you want in the future post-quantum. 384 00:24:20,771 --> 00:24:23,531 and that's ultimately 385 00:24:23,531 --> 00:24:25,851 how you can do this delegation 386 00:24:25,851 --> 00:24:27,931 to some extent right your main 387 00:24:27,931 --> 00:24:29,951 identity key signs off on the sub key 388 00:24:29,951 --> 00:24:31,511 that is then used 389 00:24:31,511 --> 00:24:33,471 for whatever you want 390 00:24:33,471 --> 00:24:36,051 where it then gets also tricky 391 00:24:36,051 --> 00:24:37,491 is how do we do multi device 392 00:24:37,491 --> 00:24:40,111 if you have five devices 393 00:24:40,111 --> 00:24:42,191 and they all belong to the same user 394 00:24:42,191 --> 00:24:43,711 then 395 00:24:43,711 --> 00:24:45,911 sure you can take your NSEC and 396 00:24:45,911 --> 00:24:47,571 sign the five pub keys 397 00:24:47,571 --> 00:24:49,931 five times and that 398 00:24:49,931 --> 00:24:55,931 works, but yeah, that's a whole things get really complicated there too. 399 00:24:55,931 --> 00:24:57,611 So we haven't really deployed that yet. 400 00:24:58,091 --> 00:24:59,851 We have a draft specification up. 401 00:25:00,291 --> 00:25:01,931 I don't think anyone implemented that yet. 402 00:25:02,411 --> 00:25:03,931 We'll get to that in a couple of weeks. 403 00:25:04,931 --> 00:25:11,051 But yeah, that also then gets into synchronization and it's a tough one. 404 00:25:11,131 --> 00:25:17,051 I was about to say that is exactly the problem that we are focused on right now 405 00:25:17,051 --> 00:25:20,471 because, or at least that I'm focused on, 406 00:25:20,471 --> 00:25:23,231 well, a different

thing is kind of getting settled 407 00:25:23,231 --> 00:25:25,311 with Paracord that is related to it. 408 00:25:26,831 --> 00:25:32,491 But essentially, I'm trying to solve it 409 00:25:32,491 --> 00:25:34,531 with the whole space idea already, 410 00:25:34,931 --> 00:25:37,431 is that what we have with spaces is we create a Hyper-B. 411 00:25:38,311 --> 00:25:40,811 We create a Hyper-Swarm and then a Hyper-B, right? 412 00:25:40,851 --> 00:25:44,531 So a network and then a append-only log for that network. 413 00:25:44,531 --> 00:25:48,311 And that is how we update just hashes. 414 00:25:48,571 --> 00:25:50,391 We don't keep the whole index in the Hyper-B. 415 00:25:50,491 --> 00:25:51,351 That's what a lot of people do. 416 00:25:51,451 --> 00:25:54,771 But the index itself has really bad query stuff. 417 00:25:55,111 --> 00:25:56,511 So we're actually just doing SQL. 418 00:25:58,591 --> 00:26:07,425 And so the append log is literally just like a hashes It basically like this is the state this is the state this is the state 419 00:26:08,165 --> 00:26:23,825 And so right now my kind of like base approach is it like, okay, well, we're going to create a space just like all of our other spaces, except our index isn't an index of hundreds of thousands of files and terabytes of storage. 420 00:26:23,825 --> 00:26:30,325 our index is just an index of accounts of devices and what we're doing is we're just syncing those 421 00:26:30,325 --> 00:26:36,845 devices across the space of those devices so that space is just the devices that are my devices 422 00:26:36,845 --> 00:26:43,245 and they boot that one up first and they sync with all the devices in the space and we have a 423 00:26:43,245 --> 00:26:48,585 dumb simple permission system which is just the these are the allowed keys and these are the keys 424 00:26:48,585 --> 00:26:50,225 that you don't send messages to. 425 00:26:51,585 --> 00:26:54,225 And you just kind of rotate the encryption key 426 00:26:54,225 --> 00:26:55,225 and, you know, kick them out. 427 00:26:55,585 --> 00:26:56,865 It's not really like backwards, 428 00:26:57,285 --> 00:26:59,025 forward secrecy or anything like that. 429 00:26:59,045 --> 00:27:00,645 It's just, you know, if you had access, 430 00:27:00,865 --> 00:27:02,745 you can see what was going on 431 00:27:02,745 --> 00:27:03,865 when you had access or whatever. 432 00:27:04,005 --> 00:27:05,105 But after you get kicked out, 433 00:27:05,225 --> 00:27:06,625 you can't look at anything anymore. 434 00:27:08,445 --> 00:27:12,145 But then literally you just sign in a device 435 00:27:12,145 --> 00:27:13,365 exactly like you were saying 436 00:27:13,365 --> 00:27:15,065 with, you know, white noise or whatever. 437 00:27:15,625 --> 00:27:17,905 I just sign this as this is also my public key. 438 00:27:17,905 --> 00:27:22,505 And then that account, that entire space gets roped up into other spaces. 439 00:27:23,045 --> 00:27:30,605 So when you join a space, your entire index of public keys join the space all at once as if this is your account. 440 00:27:30,785 --> 00:27:32,765 And then you rotate that. 441 00:27:32,905 --> 00:27:33,605 So you kick out. 442 00:27:34,225 --> 00:27:38,825 Like if I have a primal and I have an iPhone and Android and a desktop and a Linux, 443 00:27:38,925 --> 00:27:42,605 my Linux is compromised or I'm using a piece of crap that I don't want to use anymore. 444 00:27:42,605 --> 00:27:48,865 I literally just update the state and say, this one's no longer quote unquote signed in. 445 00:27:49,445 --> 00:27:56,265 And so if I can connect to the space, obviously I can get the update to the space because I wouldn't be able to connect to any other ones if I can't connect my own. 446 00:27:57,665 --> 00:28:02,865 And then that's the canonical truth of what's logged in and what's not. 447 00:28:03,125 --> 00:28:06,065 And then you just naturally kick all the other ones out. 448 00:28:06,325 --> 00:28:12,205 In theory, it sounds like it's going to work really easy, but I'm sure it's going to be a real pain in the ass to get it to flow. 449 00:28:12,605 --> 00:28:16,625 does your private user space or whatever, 450 00:28:16,845 --> 00:28:18,365 as all your devices, 451 00:28:19,005 --> 00:28:25,185 when that is logged into other spaces in the network, 452 00:28:25,565 --> 00:28:27,705 it's just sort of represented as one. 453 00:28:29,225 --> 00:28:32,625 That's sort of like an aggregation of all your things. 454 00:28:32,825 --> 00:28:34,725 And so what you do inside of that? 455 00:28:35,065 --> 00:28:37,525 Everything has my signature on it. 456 00:28:37,525 --> 00:28:40,765 So it's like these 10 keys all have my signature. 457 00:28:40,765 --> 00:28:43,205 and the new space basically says 458 00:28:43,205 --> 00:28:45,625 this signature is allowed. 459 00:28:46,645 --> 00:28:49,145 So if those 10 public keys 460 00:28:49,145 --> 00:28:50,065 all have my signature, 461 00:28:50,185 --> 00:28:51,305 those 10 public keys are allowed. 462 00:28:52,045 --> 00:28:53,325 And that's my 10 devices. 463 00:28:53,585 --> 00:28:54,325 And that space is, 464 00:28:54,485 --> 00:28:55,385 where is that hosted? 465 00:28:55,605 --> 00:28:56,845 Like in the DHT? 466 00:28:57,285 --> 00:28:58,305 It's just DHT. 467 00:28:58,565 --> 00:28:59,885 Yeah, it's just a topic in the DHT. 468 00:29:00,145 --> 00:29:01,065 So it's all just like a, 469 00:29:01,465 --> 00:29:03,445 it's a little private 470 00:29:03,445 --> 00:29:04,665 BitTorrent network, right? 471 00:29:06,145 --> 00:29:07,465 So basically that. 472 00:29:07,785 --> 00:29:09,065 But that's kind of the idea right now 473 00:29:09,065 --> 00:29:09,865 that like, 474 00:29:09,865 --> 00:29:21,485 I'm trying to sort out, but I'm kind of waiting on a different thing to really pull the trigger because my problem has been kind of that the pair layer is the discovery layer. 475 00:29:21,845 --> 00:29:24,905 But I want the Noster layer to kind of be the

identity layer. 476 00:29:25,385 --> 00:29:36,985 But that means that I have to discover first and then decide what the Noster key is that says the pair discovery was valid, you know, like because like I want to connect peer to peer. 477 00:29:36,985 --> 00:29:39,005 so I'm having a trouble 478 00:29:39,005 --> 00:29:41,245 a little bit of trouble with the key hierarchy 479 00:29:41,245 --> 00:29:42,945 because the public 480 00:29:42,945 --> 00:29:45,065 directory of like Nostrad Identity seems 481 00:29:45,065 --> 00:29:46,245 to make the most sense for 482 00:29:46,245 --> 00:29:49,085 who am I? Like what's 483 00:29:49,085 --> 00:29:51,185 my profile? This kind of reminds 484 00:29:51,185 --> 00:29:53,185 me a little bit of a problem that ZapStore 485 00:29:53,185 --> 00:29:55,045 was having because back in the 486 00:29:55,045 --> 00:29:57,305 day, so ZapStore is the kind of decentralized 487 00:29:57,305 --> 00:29:58,605 permissionless app store 488 00:29:58,605 --> 00:30:00,385 not only for Android apps 489 00:30:00,385 --> 00:30:03,025 one of my most bullish 490 00:30:03,025 --> 00:30:05,045 projects by the way. Back 491 00:30:05,045 --> 00:30:06,525 in the day they started with 492 00:30:06,985 --> 00:30:14,245 Your identity is the pub key, and then you can sign a hash of an app, basically. 493 00:30:14,745 --> 00:30:18,385 And then that's your release that was published via Nostra identity. 494 00:30:19,045 --> 00:30:24,885 But the problem is that the operating system, specifically Android, doesn't care about your Nostra keys, right? 495 00:30:24,885 --> 00:30:31,705 They have the app certificates, whatever cryptography or scheme that they use. 496 00:30:31,845 --> 00:30:35,565 But ultimately, every APK is signed, and it's trust on first use. 497 00:30:35,565 --> 00:30:39,065 So when you install an application that is signed by a public key, 498 00:30:39,505 --> 00:30:41,825 and then when you want to update that application, 499 00:30:41,965 --> 00:30:46,465 the update must be signed by the same public key as the original installed. 500 00:30:46,465 --> 00:30:48,805 It's the Android credentials, 501 00:30:49,245 --> 00:30:52,665 so there's no check if it's the Nostra credentials are right or not. 502 00:30:52,785 --> 00:30:55,205 It has no idea they're even there. 503 00:30:55,405 --> 00:30:56,465 Oh, that's an interesting point. 504 00:30:56,585 --> 00:30:58,745 You can update the app as if it's the same app, 505 00:30:58,805 --> 00:31:00,645 but it could be a malicious actor. 506 00:31:01,245 --> 00:31:02,325 Oh, that sucks. 507 00:31:02,325 --> 00:31:20,558 It also annoying them as a developer because now if I want to make a release I need to get my Nostar key out And then that just kind of annoying And so in the next version that they working on which by the way is going to be so fucking dope it the most tangent but it going 508 00:31:20,558 --> 00:31:26,098 to be the most anonymous app store ever, like the Bitcoin full node level of app stores. Insane. 509 00:31:26,518 --> 00:31:34,178 But one of the things that they're changing is that you sign with your NSEC, your APK public key, 510 00:31:34,278 --> 00:31:39,618 basically like your Android developer public key. And that's a one-time signature and you could 511 00:31:39,618 --> 00:31:41,658 expire that after a year or whatnot. 512 00:31:42,198 --> 00:31:43,658 But you only need to sign it 513 00:31:43,658 --> 00:31:45,638 once. And you can do that in a browser 514 00:31:45,638 --> 00:31:47,478 somewhere. And then from 515 00:31:47,478 --> 00:31:49,718 the future, you just make 516 00:31:49,718 --> 00:31:51,658 your releases as you have always done them 517 00:31:51,658 --> 00:31:53,318 and they have always been signed. 518 00:31:54,158 --> 00:31:55,838 And they still link back 519 00:31:55,838 --> 00:31:57,758 to your nostril identity because you 520 00:31:57,758 --> 00:31:59,278 did a two-way signature 521 00:31:59,278 --> 00:32:01,698 at one point. And 522 00:32:01,698 --> 00:32:03,678 that solved a whole bunch 523 00:32:03,678 --> 00:32:05,338 of problems in their architecture. 524 00:32:06,358 --> 00:32:07,798 And it's already, I think, 525 00:32:07,798 --> 00:32:09,918 how it's working now and they're even further fine-tuning 526 00:32:09,918 --> 00:32:11,418 that in that upcoming version too. 527 00:32:12,278 --> 00:32:13,838 So that seems a bit 528 00:32:13,838 --> 00:32:15,878 similar where you want 529 00:32:15,878 --> 00:32:17,178 to use the Nostra identity, 530 00:32:17,818 --> 00:32:19,718 but you shouldn't 531 00:32:19,718 --> 00:32:21,818 rip out the actual core 532 00:32:21,818 --> 00:32:23,738 cryptography of the base 533 00:32:23,738 --> 00:32:25,918 system that you're using and trying to replace 534 00:32:25,918 --> 00:32:27,858 it with Nostra. Kind of like the same idea 535 00:32:27,858 --> 00:32:29,398 of how we were thinking that 536 00:32:29,398 --> 00:32:31,838 we can just push nor signatures into MLS. 537 00:32:32,978 --> 00:32:33,878 But that's so 538 00:32:33,878 --> 00:32:35,718 convoluted and you don't get any 539 00:32:35,718 --> 00:32:36,638 benefits out of it. 540 00:32:36,638 --> 00:32:40,018 you get all the benefits just by making a one-time 541 00:32:40,018 --> 00:32:41,818 even bidirectional signature 542 00:32:41,818 --> 00:32:43,818 that hey this is my subkey 543 00:32:43,818 --> 00:32:45,378 and that's it 544 00:32:45,378 --> 00:32:47,938 I really like the idea 545 00:32:47,938 --> 00:32:49,718 of kind of a meta 546 00:32:49,718 --> 00:32:51,718 signature scheme where it doesn't really care 547 00:32:51,718 --> 00:32:52,438 what the encryption 548 00:32:52,438 --> 00:32:56,098 like what the ultimate key is that's being used 549 00:32:56,098 --> 00:32:58,038 where 550 00:32:58,038 --> 00:32:59,918 like okay I can 551 00:32:59,918 --> 00:33:01,958 sign for a pair key I can sign for 552 00:33:01,958 --> 00:33:03,638 a Google Android authentication 553 00:33:03,638 --> 00:33:06,098 key

I can sign for a quantum 554 00:33:06,098 --> 00:33:13,218 resistant key an mls key doesn't really matter it's just whoever can see that or understands that 555 00:33:13,218 --> 00:33:18,618 the meta signing because it's just signed messages you know it's just it's all just 556 00:33:18,618 --> 00:33:22,918 signed messages and it feels like noster has like this little window this little thing of like 557 00:33:22,918 --> 00:33:28,678 listen sign messages work you know and it's like we should just sign a lot of messages 558 00:33:28,678 --> 00:33:34,278 Can I ask a basic question about white noise? 559 00:33:34,998 --> 00:33:36,698 Because I don't follow it that much. 560 00:33:36,698 --> 00:33:49,478 Is the backend kind of similar to Noster where it's like you count on users running relays or there's like a separate class of people that run kind of relays and nodes? 561 00:33:49,738 --> 00:33:53,318 Or is the client sending messages run everything? 562 00:33:53,638 --> 00:33:56,018 Or does it actually use Noster for its backend? 563 00:33:56,438 --> 00:33:57,498 Or how does that work? 564 00:33:57,498 --> 00:34:04,918 yeah that's a good question and kind of our initial goal was to fix nostr dms because like 565 00:34:04,918 --> 00:34:10,818 the nostr dm like private messaging schemes are right they've always been broken right and and so 566 00:34:10,818 --> 00:34:17,098 that's what we set out for and that's why we use nostr keys and also nostr relays but we wanted to 567 00:34:17,098 --> 00:34:22,838 build it kind of in the spirit of nostr perfect line the gg keeps keeps saying over and over and 568 00:34:22,838 --> 00:34:26,758 i think he's totally right we need to build in the spirit of bitcoin in the spirit of nostr 569 00:34:26,758 --> 00:34:32,278 And that doesn't necessarily mean to build on top of it, just with the same mindset. 570 00:34:32,578 --> 00:34:33,918 And you know it when you see it. 571 00:34:34,638 --> 00:34:38,378 And the beauty of Noster is, you don't care about the relays. 572 00:34:38,698 --> 00:34:41,878 Relays are data storage and data transmit pipes. 573 00:34:42,678 --> 00:34:44,458 The beauty is in the signed message. 574 00:34:44,778 --> 00:34:46,718 A structured JSON that is signed. 575 00:34:47,418 --> 00:34:48,398 That's the breakthrough. 576 00:34:49,418 --> 00:34:50,278 Radical, I know. 577 00:34:50,438 --> 00:34:52,158 But that's where all the magic is at. 578 00:34:52,158 --> 00:34:59,398 so long as somebody reliable is like storing and transmitting yeah but like you know like in 579 00:34:59,398 --> 00:35:04,678 practice i mean i have a personal relay running on on some bps cloud that that you know has a public 580 00:35:04,678 --> 00:35:09,298 ip and has a domain and everything i have a personal relay running on my start nine because 581 00:35:09,298 --> 00:35:15,378 you know why not and and i have one running on my phone so like right there i have one out of three 582 00:35:15,378 --> 00:35:20,618 of my own

devices that just need to be available in order to recover like i don't expose my phone 583 00:35:20,618 --> 00:35:22,438 via Tor or FIPS or whatever. 584 00:35:22,538 --> 00:35:24,838 I could, but there's not really a need. 585 00:35:24,898 --> 00:35:26,078 I have a public server for that. 586 00:35:26,978 --> 00:35:29,338 And by the way, I do the same with all my videos. 587 00:35:29,338 --> 00:35:32,498 So 400 gigabytes of video and audio 588 00:35:32,498 --> 00:35:35,858 is mirrored on all my devices, which is cool. 589 00:35:36,598 --> 00:35:39,678 But anyhow, for White Noise specifically, 590 00:35:39,938 --> 00:35:42,918 to build in the spirit of Noster meant for us 591 00:35:42,918 --> 00:35:46,618 that the relays should just be dump delivery pipes. 592 00:35:46,618 --> 00:35:53,978 clients first like smart clients dump servers that that's kind of the spirit that we were going for 593 00:35:53,978 --> 00:36:01,158 and so right now we do use nostril relays but purely as data storage so that you can receive 594 00:36:01,158 --> 00:36:06,158 while you're offline and you know delivery that you don't need to have a public ip or anything 595 00:36:06,158 --> 00:36:15,038 like this and but it is kind of a client-side consensus system we call it convergence which 596 00:36:15,038 --> 00:36:19,478 is a bit more accurate, which means that you don't trust the servers at all. 597 00:36:20,058 --> 00:36:34,391 So you connect to them and you get a firehose of data And then locally on the client side you process all of this data and eventually you will reach the tip of the chain of your group conversation and then you know this is the key that i need to 598 00:36:34,391 --> 00:36:41,931 use to decrypt the message that someone sent to me so a heavy client side logic on figuring out 599 00:36:41,931 --> 00:36:47,811 what is the correct key that we use right now and nostrilis at the moment is the only transport 600 00:36:47,811 --> 00:36:53,571 layer that's supported in the code but it doesn't have to be i mean i was chatting with guy earlier 601 00:36:53,571 --> 00:37:00,051 that potentially this can be done over pair what we've already started on at least research and 602 00:37:00,051 --> 00:37:05,411 prototyping is over fips the new transport layer agnostic mesh network that i'm insanely bullish 603 00:37:05,411 --> 00:37:11,171 about too and it's going to be faster than than the internet and then like if you have a dm and 604 00:37:11,171 --> 00:37:17,011 both people are online it will just go from phone to phone now via the encrypted mesh network hops 605 00:37:17,011 --> 00:37:24,251 Yes, but ultimately no server endpoint in the middle other than the dumb data pipes that pass messages instantly. 606 00:37:25,631 --> 00:37:40,471 And yeah, so the only kind of assumption that we have is that every group member must receive every commit message, which are the messages that change the state of the group, like adding a member, removing a member. 607 00:37:41,091 --> 00:37:43,311 Everyone in the group has to get these. 608 00:37:43,311 --> 00:37:49,891 if you do then we will all be in agreement over what's the next key and we will have a perfectly 609 00:37:49,891 --> 00:37:56,751 working conversation if you lose some of those commit messages then you will fork the group 610 00:37:56,751 --> 00:38:02,971 and guy will use public key a to encrypt the next message and simple like steve will use the public 611 00:38:02,971 --> 00:38:11,311 key b and you will no longer be able to talk to each other gotcha that's awesome and did i hear 612 00:38:11,311 --> 00:38:21,731 correctly that you still communicate with gg yeah sure what i would give anything to have a 613 00:38:21,731 --> 00:38:29,171 conversation with him man he's gone from my life oh man you you gotta come to madera like the so 614 00:38:29,171 --> 00:38:35,691 have you got you guys know about sovereign engineering like yeah i've heard it yeah guys 615 00:38:35,691 --> 00:38:40,951 it's this is one of you know fuck bitcoin conferences i'm kind of at that point now 616 00:38:40,951 --> 00:38:49,031 god damn but like i've i've gotten even more to that point because of things like sovereign 617 00:38:49,031 --> 00:38:56,431 engineering because instead of you know a two-day three-day madness of thousands of people and you 618 00:38:56,431 --> 00:39:02,231 can barely get five sentences in with with the person you actually want to talk to let's get 21 619 00:39:02,231 --> 00:39:09,871 people and lock them on an island for like six weeks and every friday they need to demo something 620 00:39:09,871 --> 00:39:17,151 dangerous reality tv show that sounds awesome but it's so good man it's so good and honestly 621 00:39:17,151 --> 00:39:25,211 the projects that came out of that are wild like zap store blossom marmot um phips do it 622 00:39:25,211 --> 00:39:32,531 six weeks is a toll game forever like the amount of crazy shit that came out of this is wild and 623 00:39:32,531 --> 00:39:37,131 gg was the the primary one spearheading this and and that's where he keeps talking about 624 00:39:37,131 --> 00:39:38,571 built in the spirit of Bitcoin. 625 00:39:39,571 --> 00:39:42,571 And it's a, yeah, it's cool. 626 00:39:42,851 --> 00:39:43,471 You guys should come. 627 00:39:44,151 --> 00:39:47,731 All this stuff, like the FIPS network 628 00:39:47,731 --> 00:39:48,691 is such a great example 629 00:39:48,691 --> 00:39:49,891 because I've been looking at that 630 00:39:49,891 --> 00:39:52,451 as like trying to figure out 631 00:39:52,451 --> 00:39:55,891 how to do alternative networking 632 00:39:55,891 --> 00:39:57,571 to get the same result 633 00:39:57,571 --> 00:40:00,011 and just put our space kind of agnostically 634 00:40:00,011 --> 00:40:03,031 in whatever, like however we can establish connection. 635 00:40:03,251 --> 00:40:04,251 Because like one of the things we're doing 636 00:40:04,251 --> 00:40:05,391 with Pear Drop is that like, 637

00:40:05,451 --> 00:40:06,691 I just want to be able to talk to Bluetooth 638 00:40:06,691 --> 00:40:08,771 tooth and land. You know, I don't want to have 639 00:40:08,771 --> 00:40:10,071 to go out to some DHT 640 00:40:10,071 --> 00:40:12,571 to do this, 641 00:40:12,571 --> 00:40:14,671 because maybe I'm behind a VPN or 642 00:40:14,671 --> 00:40:16,711 something, and stuff is being a 643 00:40:16,711 --> 00:40:18,591 pain in the ass. Like, I don't want that to be 644 00:40:18,591 --> 00:40:19,351 my only option. 645 00:40:20,851 --> 00:40:22,631 And I just want it to work. I just 646 00:40:22,631 --> 00:40:24,791 want it to spit a file from this device 647 00:40:24,791 --> 00:40:26,611 to this device, because they're near each other. 648 00:40:28,491 --> 00:40:30,511 But, like, it 649 00:40:30,511 --> 00:40:31,751 feels like 650 00:40:31,751 --> 00:40:34,211 AI, 651 00:40:34,651 --> 00:40:36,451 like, these whole 652 00:40:36,691 --> 00:40:42,291 this whole tool set and vibe coding and all this and everything that we're headed towards and 653 00:40:42,291 --> 00:40:49,091 everything that we're trying to build is like the manifest destiny of the internet right it was like 654 00:40:49,091 --> 00:40:54,691 the original vision it's like okay eventually we're gonna solve all of this stuff and it's like 655 00:40:54,691 --> 00:41:00,931 bitcoin is such a huge part of like a paradigm shift and like oh what is actually possible like 656 00:41:00,931 --> 00:41:06,211 what did we actually figure out we could accomplish in a totally different way than we had originally 657 00:41:06,211 --> 00:41:06,931 suspect it. 658 00:41:08,831 --> 00:41:10,191 Well, and if you didn't have 659 00:41:10,191 --> 00:41:12,411 a sovereign money, you couldn't 660 00:41:12,411 --> 00:41:14,191 monetize all of this. You can't do anything without 661 00:41:14,191 --> 00:41:14,771 sovereign money. 662 00:41:15,351 --> 00:41:18,331 It's literally like, we had to 663 00:41:18,331 --> 00:41:20,311 bootstrap all of the next stage 664 00:41:20,311 --> 00:41:22,511 with a money. 665 00:41:23,591 --> 00:41:24,371 But the money's not 666 00:41:24,371 --> 00:41:26,371 enough, right? That's one 667 00:41:26,371 --> 00:41:28,391 of the UX. It's all about figuring 668 00:41:28,391 --> 00:41:30,331 out the UX. Like, how do 669 00:41:30,331 --> 00:41:32,291 you get all of this 670 00:41:32,291 --> 00:41:34,111 complexity? It feels like we're creating 671 00:41:34,111 --> 00:41:37,751 in a bird's eye view, like stand back, 672 00:41:37,751 --> 00:41:39,991 like look from outer space, 673 00:41:39,997 --> 00:41:48,717 you it really seems like we're creating a global brain literally and we're and it always makes me 674 00:41:48,717 --> 00:41:54,637 think of like you know the cambrian explosion in like life it's like oh you figure out there's this 675 00:41:54,637 --> 00:42:03,337 new piece and there's this weird like purpose-driven feedback loop where there is so much acceleration 676 00:42:03,337 --> 00:42:11,577 and there is so much like variation all at once trying to figure out what that new paradigm is. 677 00:42:11,577 --> 00:42:14,377 It's like, it's like you hit the edge of the plateau and then it's just like, goes 678 00:42:14,377 --> 00:42:20,877 absolutely vertical until like something sets in and then it, then it turns again, you know, 679 00:42:20,897 --> 00:42:25,897 you got that S curve. Um, and it feels like we're creating this sort of like global brain 680 00:42:25,897 --> 00:42:31,637 that is genuinely decentralized, like a mycelial network, like, like a neural network in a human 681 00:42:31,637 --> 00:42:42,397 brain is, you know, and that that we're trying, we're aggressively trying to find cohesion 682 00:42:42,397 --> 00:42:48,797 around what the security protocol is to actually make it so that each one of these neurons 683 00:42:48,797 --> 00:42:54,817 is playing its role independently, you know, like playing fairly, like there's actual there's 684 00:42:54,817 --> 00:43:00,337 there's basically not a global point of failure because otherwise it's not actually going 685 00:43:00,337 --> 00:43:00,897 to survive. 686 00:43:00,897 --> 00:43:29,777 Like centralization is great in the short term. It leads to quicker adaptation and applying a problem or applying a solution to a problem because you have centralized decision making. But it creates an insane brittleness. Like it's insanely hard. It's a very hard system when you have it centralized because you have very core and cohesive decision making. But it's insanely brittle. 687 00:43:29,777 --> 00:43:35,097 so the bigger it gets it becomes more and more sluggish and it becomes more and more 688 00:43:35,097 --> 00:43:41,097 a victim of collapse and it might be really really hard to break but as soon as it breaks 689 00:43:41,097 --> 00:43:49,317 the whole thing shatters you know um and whereas decentralization is water you know it's like 690 00:43:49,317 --> 00:43:54,717 everything moves everything adjusts this thing can respond to this new piece of information and 691 00:43:54,717 --> 00:43:59,657 send that information it's like a mycelial network it's wild that you know they'll see like this this 692 00:43:59,657 --> 00:44:07,377 thing is considered a multi-set organism and all these different edges of it will literally react 693 00:44:07,377 --> 00:44:13,837 independently and if you break it off they all just continue expanding independently but if this 694 00:44:13,837 --> 00:44:20,757 way just distant corner of a thing encounters a chemical or a plant or an animal that's like 695 00:44:20,757 --> 00:44:28,417 in danger to it and it literally thinks up and adapts a response to it you can go find 696 00:44:28,417 --> 00:44:32,377 a mile away on the end of that huge mycelial network or whatever. 697 00:44:32,497 --> 00:44:36,717 You can go find the parts on the very opposite side of it 698 00:44:36,717 --> 00:44:38,797 now being able to produce that

chemical in response. 699 00:44:39,937 --> 00:44:42,877 And that is literally what we're doing for human information 700 00:44:42,877 --> 00:44:44,437 and human ideas. 701 00:44:45,057 --> 00:44:48,277 And it seems to make sense that none of it even works. 702 00:44:48,817 --> 00:44:49,937 You have to evolve AI. 703 00:44:50,057 --> 00:44:52,817 You have to create AI as a tool to actually do that. 704 00:44:53,717 --> 00:44:57,737 You have to create AI to go into outer space. 705 00:44:57,737 --> 00:45:00,417 Like, how are we ever going to do this at the rate that made sense? 706 00:45:00,497 --> 00:45:01,297 We didn't create AI. 707 00:45:02,437 --> 00:45:04,237 However, everybody says it's going to kill us. 708 00:45:05,997 --> 00:45:08,417 That's like the big news right now is we're all going to die. 709 00:45:08,617 --> 00:45:09,317 We're all going to die. 710 00:45:09,557 --> 00:45:15,477 Well, the thing is, the thing is, the people that are saying that it's going to kill us are the people that want regulatory capture. 711 00:45:15,817 --> 00:45:15,977 Yeah. 712 00:45:16,037 --> 00:45:19,757 And it's literally the belief that it's going to kill us that's going to make it evil. 713 00:45:20,137 --> 00:45:20,277 Yeah. 714 00:45:20,317 --> 00:45:24,637 And or like, like, you know, like protect the fact that it is evil. 715 00:45:24,637 --> 00:45:31,677 thinking that we have to use it to control people's access to it right is literally the thing evil 716 00:45:31,677 --> 00:45:38,477 yeah it's literally evil that we have to have a a healthy decentralized network that is actually 717 00:45:38,477 --> 00:45:46,537 updating our models based on true truly decentralized like like right now all of the 718 00:45:46,537 --> 00:45:52,617 models are being heavily manipulated by twitter like this is who was it that just covered this 719 00:45:52,617 --> 00:45:55,497 somebody I was watching 720 00:45:55,497 --> 00:45:56,977 did a podcast on this. 721 00:45:57,377 --> 00:45:59,377 One of the main reasons that these 722 00:45:59,377 --> 00:46:01,497 Twitter, like copy, paste 723 00:46:01,497 --> 00:46:03,957 Twitter things, like everybody's 724 00:46:03,957 --> 00:46:05,557 it's like the same 725 00:46:05,557 --> 00:46:07,557 narrative over and over. It's not even 726 00:46:07,557 --> 00:46:09,157 to convince the people on Twitter. 727 00:46:09,357 --> 00:46:11,357 It's to manipulate the AI. 728 00:46:12,177 --> 00:46:13,557 It's the AI 729 00:46:13,557 --> 00:46:15,557 responses. So if somebody searches 730 00:46:15,557 --> 00:46:16,997 on chat GPT, 731 00:46:17,437 --> 00:46:19,577 what happened to Charlie Kirk 732 00:46:19,577 --> 00:46:21,497 or whatever, right? If they have 733 00:46:21,497 --> 00:46:27,257 repeated enough times that tyler robertson did this then the ai is going to respond with that 734 00:46:27,257 --> 00:46:35,737 and so so there is this like this like concerted effort to manipulate the ai models and if they're 735 00:46:35,737 --> 00:46:40,297 only pulling from one network or from that's that's why reddit corruption i mean reddit corruption 736 00:46:40,297 --> 00:46:45,757 has been a problem forever but like reddit corruption was such a problem because reddit's 737 00:46:45,757 --> 00:46:57,657 being used to train the AI Twitter being used to train the AI We have to have honest networks and honest conversations with real nodes 738 00:46:57,657 --> 00:46:58,957 to get 739 00:46:58,957 --> 00:47:01,197 an AI that's not 740 00:47:01,197 --> 00:47:03,197 biased 741 00:47:03,197 --> 00:47:04,737 one way or the other. 742 00:47:05,357 --> 00:47:06,897 There's still bias, right? 743 00:47:07,637 --> 00:47:09,697 The majority may believe some stupid 744 00:47:09,697 --> 00:47:11,537 thing. It should just be decentralized bias 745 00:47:11,537 --> 00:47:13,237 is the problem. Right. Exactly. 746 00:47:14,197 --> 00:47:15,857 Bias is necessary. 747 00:47:16,057 --> 00:47:17,277 It's like, oh, well, it's bias 748 00:47:17,277 --> 00:47:22,877 to the fact that gravity will pull you towards earth as it was it's like oh oh no you know like 749 00:47:22,877 --> 00:47:28,257 bias is critical bias is just an arbitrary thing to you know point something out as negative when 750 00:47:28,257 --> 00:47:34,737 it's like naturally exists everywhere all the time like bias is simply like you you believe something 751 00:47:34,737 --> 00:47:41,497 that something is concrete somewhere yeah you know and we can kind of push our bias much easier now 752 00:47:41,497 --> 00:47:48,497 because we've never had a lot of people, generally speaking, 753 00:47:48,737 --> 00:47:50,197 in the Freedom Tech area. 754 00:47:50,757 --> 00:47:53,157 And even fewer of those people actually could code. 755 00:47:53,957 --> 00:47:57,277 And I still believe that software is critical for human liberation. 756 00:47:58,677 --> 00:48:04,937 And now we have 100% of the people who are interested in Freedom Tech 757 00:48:04,937 --> 00:48:06,637 can now build Freedom Tech. 758 00:48:07,777 --> 00:48:11,397 There has never been an excuse that because you cannot code, 759 00:48:11,497 --> 00:48:13,377 you cannot work on free open source software. 760 00:48:13,517 --> 00:48:14,957 Like that was always nonsense. 761 00:48:15,497 --> 00:48:17,437 But now it's like especially nonsense. 762 00:48:17,757 --> 00:48:19,237 Because back in the day, I don't know, 763 00:48:19,277 --> 00:48:21,577 you could do product management, marketing, whatever. 764 00:48:22,097 --> 00:48:23,537 Now you can literally do the coding. 765 00:48:24,057 --> 00:48:25,157 All you have to do is care and think. 766 00:48:25,237 --> 00:48:25,937 It's just a prompt. 767 00:48:26,757 --> 00:48:27,977 You have to care and think. 768 00:48:28,417 --> 00:48:29,597 And if you can't do those things, 769 00:48:29,657 --> 00:48:30,277 what the hell do you think? 770 00:48:30,377 --> 00:48:31,937 You can do anything with Freedom Tech anyway. 771 00:48:33,497 -->

00:48:37,077 And you also have to accept some kind of responsibility 772 00:48:37,077 --> 00:48:37,877 or something. 773 00:48:38,017 --> 00:48:40,657 Because I think a lot of the reasons people, 774 00:48:40,657 --> 00:48:46,317 you know hire an accountant and hire a marketing person and hire somebody else is that like they 775 00:48:46,317 --> 00:48:50,377 want to be able to like point the finger at somebody else when something goes wrong 776 00:48:50,377 --> 00:48:58,717 oh dude do you know what i mean the amount of that is insane yeah i mean yeah like i mean okay 777 00:48:58,717 --> 00:49:03,457 everybody wants to be like if something bad happens you don't want it to be your fault like 778 00:49:03,457 --> 00:49:07,057 that's you know like i don't know i mean there's a degree to which like if it is your fault like 779 00:49:07,057 --> 00:49:12,797 you can do something about it maybe but like you know i understand that there's like uh you know 780 00:49:12,797 --> 00:49:17,977 we get trained like you don't want to mess up right nobody wants things to go bad but like the 781 00:49:17,977 --> 00:49:26,777 amount of just like i don't know i've had so many conversations with like people friends over the 782 00:49:26,777 --> 00:49:32,097 years that like work in a place and they're like it feels like atlas shrugged or whatever it's like 783 00:49:32,097 --> 00:49:36,657 it's like is that like the whole idea that you're just sort of walking through empty rooms if you 784 00:49:36,657 --> 00:49:41,297 just take responsibility for what's happening everybody wants to give it to you because they're 785 00:49:41,297 --> 00:49:50,017 so afraid of like of like being you know like making a decision decision yeah yeah and and 786 00:49:50,017 --> 00:49:57,857 and like i don't know radical responsibility yeah i feel like i'm encountering that half the time 787 00:49:57,857 --> 00:50:02,737 when i'm talking to customer service or something somewhere is that like they can't you can't do 788 00:50:02,737 --> 00:50:06,337 anything they don't want the responsibility of doing something outside of like whatever the 789 00:50:06,337 --> 00:50:11,757 policy is like whatever just and then they just they just read you a response like they don't yeah 790 00:50:11,757 --> 00:50:16,357 there's not even people they're like they're just like i want a bot to take over my job 791 00:50:16,357 --> 00:50:25,297 okay um but what's what's crazy about that too is just in kind of the especially like 792 00:50:25,297 --> 00:50:32,537 leftists and stuff talking about how like and everybody can be ceos and and everybody just 793 00:50:32,537 --> 00:50:37,957 evil business owners just stole all of this from us. It's like, bitch, you won't take responsibility 794 00:50:37,957 --> 00:50:44,017 for like buying us a candy bar and eating it and making your ass fat. You know, like you, 795 00:50:44,297 --> 00:50:50,097 the idea that you were going to go sit in front of a company that's got 10,000 employees and make 796 00:50:50,097 --> 00:50:55,277 decisions on like which department gets more funding and which product we're going to go after 797 00:50:55,277 --> 00:51:01,417 and what features you're going to do and decisions that literally alter people's lives and are 798 00:51:01,417 --> 00:51:05,457 inevitably no matter what you decide there's going to be some subset of people that just utterly 799 00:51:05,457 --> 00:51:11,837 hate you and it's all going to be your fault like you can't you're unwilling to take responsibility 800 00:51:11,837 --> 00:51:17,737 for making up your bed you know and you're and you just think that like just this arbitrary thing 801 00:51:17,737 --> 00:51:23,477 that like anybody could just be a ceo or whatever i have a i have a tiny company that does nothing 802 00:51:23,477 --> 00:51:30,617 important the amount of like anxiety that i have over making fucking decisions it's just like it's 803 00:51:30,617 --> 00:51:34,617 like nothing but i have to make 10 decisions a minute and i'm just like oh god i just want to 804 00:51:34,617 --> 00:51:41,977 sit down and eat fucking cheetos and watch netflix man just give me a break and i just can't imagine 805 00:51:41,977 --> 00:51:47,337 i can't imagine having i hope to one day have to deal with the pressure but i just can't imagine 806 00:51:47,337 --> 00:51:53,197 having a company with like 2 000 people in it like 10 000 employees that's a city that's a city i 807 00:51:53,197 --> 00:51:58,717 don't think we need that huge companies for for most things i would like to think that i can be 808 00:51:58,717 --> 00:52:10,196 successful within Dunbar number You know That I could be like really really successful at like a group of 200 people that just feels like a family and 809 00:52:10,196 --> 00:52:11,796 that we're building something amazing 810 00:52:11,796 --> 00:52:14,356 and millions of people are using it. 811 00:52:14,416 --> 00:52:16,056 Like that's kind of my goal. 812 00:52:16,616 --> 00:52:18,036 And then I kind of want to 813 00:52:18,036 --> 00:52:19,896 cap it there. 814 00:52:19,896 --> 00:52:22,676 What are you going to do with 200 people? 815 00:52:23,276 --> 00:52:24,176 Oh no, that would be huge. 816 00:52:24,296 --> 00:52:25,836 I'm not even saying it's not huge. 817 00:52:26,036 --> 00:52:28,316 I'm just meaning in that sense 818 00:52:28,316 --> 00:52:31,676 is that some CEO of some company 819 00:52:31,676 --> 00:52:33,316 that's got hundreds of thousands, 820 00:52:33,396 --> 00:52:33,856 like Apple. 821 00:52:34,076 --> 00:52:34,576 I don't even know how many. 822 00:52:34,696 --> 00:52:35,896 What is it, like 600,000? 823 00:52:36,016 --> 00:52:38,956 Some psycho number of employees, right? 824 00:52:40,476 --> 00:52:42,056 I kind of feel like the future 825 00:52:42,056 --> 00:52:44,056 may be in some sort of... 826 00:52:44,696 --> 00:52:47,356 Okay, take the open source software world, right? 827 00:52:47,916 --> 00:52:52,396 If I

was to design a product 828 00:52:52,396 --> 00:52:53,056 that could be... 829 00:52:53,976 --> 00:52:56,136 Different components can be 3D printed 830 00:52:56,136 --> 00:52:56,696 or whatever. 831 00:52:56,696 --> 00:53:03,676 we need some protocol for deciding whether each person that's producing things is is meeting the 832 00:53:03,676 --> 00:53:09,716 quality like we need some sort of like reputational or whatever is meeting the standards but honestly 833 00:53:09,716 --> 00:53:15,456 i think you could do you like it would actually make more sense to have some sort of very 834 00:53:15,456 --> 00:53:23,036 decentralized and and like fairly flat you know like like i feel like the hierarchy 835 00:53:23,036 --> 00:53:29,436 the reason companies have hierarchies is because we have these communication problems right and so 836 00:53:29,436 --> 00:53:35,136 if you get if you can build a network where you can you can communicate with people in a in a much 837 00:53:35,136 --> 00:53:40,296 more decentralized way we build these you know peer-to-peer technologies and whatever and we can 838 00:53:40,296 --> 00:53:46,956 we can even like have you know lightning prisms or whatever where people get paid for different 839 00:53:46,956 --> 00:53:53,556 parts of a job or something that you could potentially have a company that is not really 840 00:53:53,556 --> 00:53:59,996 a company, right? Like you don't actually have to legally exist. And lots of people are still doing, 841 00:53:59,996 --> 00:54:06,756 you know, different parts of a job together, you know, like, like producing different parts of 842 00:54:06,756 --> 00:54:17,456 something? I would say that the reason or the biggest like barrier to that would be that, 843 00:54:18,456 --> 00:54:23,336 well, the biggest one would actually be the problem we just talked about is that the hierarchy 844 00:54:23,336 --> 00:54:29,456 isn't necessarily for thinking. The hierarchy is for responsibility. It's who's going to take 845 00:54:29,456 --> 00:54:39,596 responsibility for what change or what decision or what cost. Right. And I think the thing actually 846 00:54:39,596 --> 00:54:47,996 works more seamless when you do have a responsibility hierarchy because of how, how much people don't, 847 00:54:47,996 --> 00:54:53,576 I mean, people don't want to make decisions about like the simplest of things. I mean, look, look at, 848 00:54:53,676 --> 00:54:57,256 you're in, you're in the pair drive stuff. Like, you know, we've, we've been working on this for 849 00:54:57,256 --> 00:55:02,776 forever, right? How, how much of a pain in the ass and how often do we, are we always just like, 850 00:55:03,236 --> 00:55:06,856 I don't know, what do you think about whether this button should be long or tall? You know, 851 00:55:06,856 --> 00:55:12,216 like, I mean, just, just like crazy little mundane decisions like that, but that have like a real 852 00:55:12,216 --> 00:55:19,416 effect on like how you feel when using a

thing and how much of a pain in the ass it is. And like, 853 00:55:19,496 --> 00:55:24,716 how many times have I made a decision? And it's just like, that was a terrible decision. That was 854 00:55:24,716 --> 00:55:28,076 so stupid we gotta bring the sidebar back why the why did we get rid of the sidebar i don't even 855 00:55:28,076 --> 00:55:34,476 remember why i got rid of the sidebar why i'm i'm i just wasted two weeks of our time building it 856 00:55:34,476 --> 00:55:40,796 this way because this is stupid this is a terrible i hate it already you know um and so like you have 857 00:55:40,796 --> 00:55:44,916 to have the hierarchy for the responsibility get rid of the sidebar to know that you needed this 858 00:55:44,916 --> 00:55:54,296 you're an idiot for getting rid of the sidebar um and uh but uh but then the other one would be um 859 00:55:54,296 --> 00:56:02,276 uh vision you know it's like those uh decentralized painting out well those old decentralized painting 860 00:56:02,276 --> 00:56:07,316 things that they did on like bitcoin early on and i think somebody even made one for noster 861 00:56:07,316 --> 00:56:12,796 where like you can get like everybody wants to make their pictures there put their pixels on it 862 00:56:12,796 --> 00:56:16,976 it was like a lightning thing wasn't it yeah yeah there's a lightning thing too yeah um it's like 863 00:56:16,976 --> 00:56:22,996 look it's a public bulletin board and i go to it just look fucking insane it looks like it looks 864 00:56:22,996 --> 00:56:29,876 like the worst possible graffiti wall that you've ever seen it's not even like good graffiti you 865 00:56:29,876 --> 00:56:36,616 know it's just what i mean is something something like linux right like there's a million different 866 00:56:36,616 --> 00:56:41,616 flavors of linux you can always fork the project right and make it like now i want to make it my 867 00:56:41,616 --> 00:56:48,656 way but i just feel like you could you could have a say it's a car let's call it a car company like 868 00:56:48,656 --> 00:56:54,756 I design, I redesign things around a very basic, 869 00:56:55,336 --> 00:56:57,456 let's say I have an engine idea that's very simple. 870 00:56:57,876 --> 00:56:59,976 It solves a handful of problems. 871 00:57:00,576 --> 00:57:05,776 I have, basically, I've spent years and years 872 00:57:05,776 --> 00:57:10,656 coming up with everything to do just a base platform for a vehicle. 873 00:57:11,396 --> 00:57:23,376 And then I going to produce this version I got some other people helping me produce parts of it And those are like the official you know this is my version And then other people can take it and contribute 874 00:57:23,476 --> 00:57:25,976 They can produce their own version of my car. 875 00:57:25,976 --> 00:57:30,296 Like they can make a Linux car or whatever, a Jeffrey car, 876 00:57:30,396 --> 00:57:31,356 whatever you want to call it. 877 00:57:32,196 --> 00:57:34,616 And they just. 878 00:57:35,176 --> 00:57:36,756 Hey, dude, you got that new Jeffrey? 879 00:57:37,636 --> 00:57:38,716 No, that's terrible. 880 00:57:39,476 --> 00:57:41,416 But they can essentially. 881 00:57:41,776 --> 00:57:42,436 I'm not V8 Jeffrey. 882 00:57:42,496 --> 00:57:45,896 put their twist on it 883 00:57:45,896 --> 00:57:48,696 and change whatever they want. 884 00:57:48,796 --> 00:57:49,456 I mean, it could be... 885 00:57:49,456 --> 00:57:50,316 Create a car protocol 886 00:57:50,316 --> 00:57:52,496 and make everybody... 887 00:57:52,496 --> 00:57:54,276 Or spec and make everybody... 888 00:57:54,276 --> 00:57:55,716 And people can be compatible with it. 889 00:57:55,996 --> 00:57:56,996 So that's Wikicar. 890 00:57:57,156 --> 00:57:58,156 Wikicar tried to do that 891 00:57:58,156 --> 00:57:59,376 technically in a sense. 892 00:57:59,956 --> 00:58:00,276 Sure. 893 00:58:00,456 --> 00:58:01,396 And I think... 894 00:58:01,396 --> 00:58:01,916 I mean, obviously, 895 00:58:02,036 --> 00:58:03,096 there's people working 896 00:58:03,096 --> 00:58:04,196 in that direction. 897 00:58:04,416 --> 00:58:05,676 There's also the 898 00:58:05,676 --> 00:58:08,136 open source ecology people. 899 00:58:09,356 --> 00:58:11,896 But I think one of the main problems 900 00:58:11,896 --> 00:58:16,836 with something like that is just that, well, just look at, fuck, 901 00:58:16,996 --> 00:58:20,496 it's the same problem in the phones, it's the same problem in cars, 902 00:58:21,036 --> 00:58:26,036 is that there's proprietary garbage, both hardware and software, 903 00:58:26,636 --> 00:58:31,296 but like that you're solving the same problem over and over 904 00:58:31,296 --> 00:58:33,716 with somebody's bullshit. 905 00:58:34,156 --> 00:58:38,456 Basically, what Noster did was take very, very simple things 906 00:58:38,456 --> 00:58:43,216 and make a network out of like simple stuff that we, you know, we knew were possible. 907 00:58:44,296 --> 00:58:47,556 What cars have done is they're like, oh, computers exist now. 908 00:58:47,996 --> 00:58:52,976 We don't have to, we can abandon like Rea stats for, you know, knobs that actually do anything. 909 00:58:52,976 --> 00:58:58,416 We can abandon, you know, these like old switch technology that actually works forever. 910 00:58:58,996 --> 00:59:01,116 We can abandon all that and just make it a computer. 911 00:59:01,556 --> 00:59:08,256 And then everything has to, you know, like it's just a fake button, you know, or a fake knob or whatever. 912 00:59:08,456 --> 00:59:09,976 Like that's horrible. 913 00:59:10,116 --> 00:59:10,916 That's horrible. 914 00:59:10,916 --> 00:59:12,276 Like user experience. 915 00:59:12,476 --> 00:59:14,396 It's horrible from an engineering standpoint. 916 00:59:14,956 --> 00:59:16,616 The failure mode is garbage. 917 00:59:16,796 --> 00:59:18,376 Like what we need. 918 00:59:18,636 --> 00:59:22,636 They're all locked in with unions to have to use this company to build the interface. 919 00:59:22,756 -->

00:59:24,236 So they all build shit interfaces. 920 00:59:24,376 --> 00:59:32,436 They don't know jack about computers and design or anything, but like they're going to lose all of their workers if they don't buy the interface from this one company. 921 00:59:32,836 --> 00:59:38,276 And like, and so they just got like this chokehold of like, you can't sure you're going to switch to computers. 922 00:59:38,456 --> 00:59:42,276 But you're still going to use the company that made the buttons, that made the old dials. 923 00:59:42,276 --> 00:59:45,556 And we're going to design app interfaces like, oh, that's going to go well. 924 00:59:45,916 --> 00:59:46,736 How'd that work? 925 00:59:47,596 --> 01:00:01,256 And we've now gotten to the point where there's these like cheap stepper motors everywhere from 3D printers and stuff that like small light ones that you could make very, very simple circuits. 926 01:00:01,256 --> 01:00:06,816 like like you know arduino even if you had to put 20 arduinos in the car or whatever like 927 01:00:06,816 --> 01:00:10,976 tiny like it's i'm sure there's something simple than a microcontroller or whatever 928 01:00:10,976 --> 01:00:16,856 something's very very simple you can make a circuit that's reliable that does a complicated 929 01:00:16,856 --> 01:00:22,616 thing with a stepper motor and then you can apply that to like 10 different things in the car 930 01:00:22,616 --> 01:00:27,576 and they don't all depend on each other they don't all you know you could make something you 931 01:00:27,576 --> 01:00:30,896 you come up with a very, I don't know, 932 01:00:30,956 --> 01:00:32,536 redundant's not the word, but like, you know, 933 01:00:32,616 --> 01:00:35,036 a robust sort of like. 934 01:00:35,416 --> 01:00:36,096 Kind of modular. 935 01:00:37,616 --> 01:00:38,896 Modular, that's exactly. 936 01:00:39,276 --> 01:00:42,076 A very modular system that solves all these problems. 937 01:00:42,576 --> 01:00:44,436 And I feel like that would be, like, 938 01:00:44,576 --> 01:00:49,136 that's sort of the foundation we need for open source cars, 939 01:00:49,136 --> 01:00:52,676 open source heavy equipment, open source anything. 940 01:00:52,916 --> 01:00:55,776 Like you need, you know, to have the refinement, 941 01:00:55,776 --> 01:00:59,476 that you get from the modern electronics 942 01:00:59,476 --> 01:01:01,616 combined with the reliability 943 01:01:01,616 --> 01:01:03,636 of the old Rea stats 944 01:01:03,636 --> 01:01:05,316 and whatever it is. 945 01:01:05,396 --> 01:01:08,056 I just want a car that you can work on again, man. 946 01:01:08,416 --> 01:01:11,056 I feel like I'm so sick of stuff 947 01:01:11,056 --> 01:01:12,696 that I don't own. 948 01:01:12,816 --> 01:01:14,896 Have you guys seen the Edison Motors 949 01:01:14,896 --> 01:01:15,896 from Canada? 950 01:01:16,676 --> 01:01:17,096 No. 951 01:01:17,176 --> 01:01:18,216 Really interesting guys. 952 01:01:19,036 --> 01:01:21,396 They're working on heavy vocational 953 01:01:21,396 --> 01:01:25,056 diesel hybrid electric trucks. 954 01:01:25,056 --> 01:01:27,556 or you have the diesel generator that charges 955 01:01:27,556 --> 01:01:29,076 a relatively small battery 956 01:01:29,076 --> 01:01:31,676 and then you have full electric drive. 957 01:01:32,756 --> 01:01:34,136 For one, cool concept. 958 01:01:34,716 --> 01:01:35,856 It's a series hybrid? 959 01:01:37,196 --> 01:01:37,396 Yeah. 960 01:01:37,536 --> 01:01:38,636 Instead of a parallel hybrid? 961 01:01:38,716 --> 01:01:38,896 Yeah. 962 01:01:39,556 --> 01:01:39,976 Exactly. 963 01:01:40,176 --> 01:01:42,576 It's like a train or something, right? 964 01:01:42,696 --> 01:01:43,076 Exactly. 965 01:01:43,216 --> 01:01:43,756 Like a train. 966 01:01:43,756 --> 01:01:45,516 It works great for heavy machinery. 967 01:01:46,596 --> 01:01:48,836 And so they're like loggers from Canada 968 01:01:48,836 --> 01:01:52,816 and built a really cool truck 969 01:01:52,816 --> 01:01:54,256 with the right to repair. 970 01:01:55,056 --> 01:01:59,076 I did reach out to them trying to get them to orange pill them. 971 01:01:59,736 --> 01:02:02,036 And it doesn't seem like they're fully open source, 972 01:02:02,356 --> 01:02:05,496 but they're very community focused 973 01:02:05,496 --> 01:02:08,916 and don't want to get big VCs involved, 974 01:02:08,916 --> 01:02:12,056 but actually want to get small truckers to chip in 975 01:02:12,056 --> 01:02:13,856 and get a small equity position as well. 976 01:02:14,936 --> 01:02:15,756 Really cool. 977 01:02:15,896 --> 01:02:18,096 If I were driving a big ass truck, 978 01:02:18,196 --> 01:02:19,876 I would probably try to drive that one. 979 01:02:20,916 --> 01:02:22,936 You know, I just shout out, 980 01:02:22,936 --> 01:02:23,836 talking about like 981 01:02:23,836 --> 01:02:24,876 things built on 982 01:02:24,876 --> 01:02:26,216 proprietary stuff 983 01:02:26,216 --> 01:02:26,796 and just like 984 01:02:26,796 --> 01:02:27,916 carrying baggage 985 01:02:27,916 --> 01:02:29,996 from you know 986 01:02:30,008 --> 01:02:36,828 one era to the next or one layer to the next got to give a shout out to matt hill and the start nine 987 01:02:36,828 --> 01:02:45,788 team because legends i've got the um start nine router the whole thing oh you have one how is it 988 01:02:45,788 --> 01:02:51,168 oh it's fucking great i i set it up once i barely even touched i have not even gotten the time to 989 01:02:51,168 --> 01:02:56,988 like sit down and like actually play play because like i want to set up my network downstairs but i 990 01:02:56,988 --> 01:03:08,608 It was so easy to just kind of do the default, not the default, but the main things that I wanted to do, like selecting devices and knowing what I did. 991 01:03:08,608 --> 01:03:11,748 And I just did like a one run setup on the first thing. 992 01:03:11,828 --> 01:03:13,668 It's been working, working amazing. 993 01:03:14,928 --> 01:03:19,508 And even the chip, even

the chip is open source. 994 01:03:19,668 --> 01:03:24,828 And that is one thing that I don't know anything, but they're using the RISC-V chip. 995 01:03:24,828 --> 01:03:26,528 so the entire 996 01:03:26,528 --> 01:03:29,048 like I don't see how they 997 01:03:29,048 --> 01:03:31,108 they don't basically win 998 01:03:31,108 --> 01:03:33,168 the trophy on the most secure 999 01:03:33,168 --> 01:03:35,008 and most open source router that 1000 01:03:35,008 --> 01:03:35,788 exists anywhere 1001 01:03:35,788 --> 01:03:38,068 and 1002 01:03:38,068 --> 01:03:41,128 I hope man I hope that thing's like super 1003 01:03:41,128 --> 01:03:42,748 successful if it isn't I think it's just because 1004 01:03:42,748 --> 01:03:44,928 the word didn't get out that people knew 1005 01:03:44,928 --> 01:03:46,748 it was a thing 1006 01:03:46,748 --> 01:03:48,068 but 1007 01:03:48,068 --> 01:03:50,068 I was very lucky to get 1008 01:03:50,068 --> 01:03:52,968 one of the super early 1009 01:03:52,968 --> 01:03:54,268 runs of it 1010 01:03:54,268 --> 01:03:58,448 um in fact mine doesn't even have a it just has the one ethernet because it's not even 1011 01:03:58,448 --> 01:04:03,088 it's like one of the like almost prototype type things or whatever so i just plug it into my 1012 01:04:03,088 --> 01:04:06,788 my big switch but um it's dope i just want to shout out that 1013 01:04:06,788 --> 01:04:13,868 you know one thing that i think um scott from bidax got really lucky with 1014 01:04:13,868 --> 01:04:19,568 um you know how scott from bidax is the open source hero right he's he's just like 1015 01:04:19,568 --> 01:04:25,928 the biggest promoter of open source ever and she's freaking been right about everything he's said 1016 01:04:25,928 --> 01:04:32,708 now that like the cold card hack and you know the liquid hack and all this stuff ai and he just got 1017 01:04:32,708 --> 01:04:37,468 so lucky man like if ai wouldn't have come around and hacked everything he wouldn't have been so 1018 01:04:37,468 --> 01:04:44,488 right about open source but i have to admit like scott from bidx all the arguments we've had over 1019 01:04:44,488 --> 01:04:50,568 all the years has just been so right like everything has to be open source so many arguments 1020 01:04:51,608 --> 01:04:58,248 there's there's just no excuse anymore and uh it's not like you saw ai coming like this scott 1021 01:04:58,248 --> 01:05:00,968 don't act like that don't even pretend you're that good 1022 01:05:03,608 --> 01:05:10,808 you just guess and you got lucky eggs right though i'm really looking forward to like some big you 1023 01:05:10,808 --> 01:05:13,068 you probably need a lot of capital for that. 1024 01:05:13,208 --> 01:05:15,848 But we have big whales now. 1025 01:05:16,468 --> 01:05:19,348 So someone should just build a cypherpunk as fuck, 1026 01:05:19,528 --> 01:05:21,928 open source, basic hardware stuff. 1027 01:05:21,928 --> 01:05:26,328 I want headphones, charging bricks, 1028 01:05:27,288 -->

01:05:29,768 simple everyday stuff. 1029 01:05:29,768 --> 01:05:30,368 Dude, I want to do that. 1030 01:05:30,508 --> 01:05:35,588 Man, it's on my list of infinite projects, right? 1031 01:05:36,008 --> 01:05:38,528 But that's one of the things I've always wanted to do, 1032 01:05:38,528 --> 01:05:39,748 of just like making, 1033 01:05:40,228 --> 01:05:40,988 I want to make, 1034 01:05:41,188 --> 01:05:42,508 in the same way that I'm trying to build 1035 01:05:42,508 --> 01:05:43,468 all my own software 1036 01:05:43,468 --> 01:05:45,608 and I get annoyed with my recording 1037 01:05:45,608 --> 01:05:46,408 and editing software, 1038 01:05:46,508 --> 01:05:47,228 I'm like, I'm going to make my own 1039 01:05:47,228 --> 01:05:48,268 recording and editing software. 1040 01:05:48,728 --> 01:05:49,408 I want to make my own, 1041 01:05:49,508 --> 01:05:50,528 I want to make like a simple 1042 01:05:50,528 --> 01:05:52,368 open source power tool set up. 1043 01:05:52,468 --> 01:05:53,088 I want to do, 1044 01:05:53,648 --> 01:05:55,108 I had this company of, 1045 01:05:55,808 --> 01:05:58,328 the company idea of making appliances 1046 01:05:58,328 --> 01:06:00,568 and literally just calling it Jest. 1047 01:06:01,348 --> 01:06:01,848 And it's like, 1048 01:06:02,088 --> 01:06:02,788 Jest washer, 1049 01:06:03,448 --> 01:06:04,328 Jest refrigerator 1050 01:06:04,328 --> 01:06:05,448 and like just going back 1051 01:06:05,448 --> 01:06:06,308 and make it simple, 1052 01:06:06,428 --> 01:06:07,948 but still make, 1053 01:06:07,948 --> 01:06:13,868 you can add you know things in that are like nice features or like nice elements without losing like 1054 01:06:13,868 --> 01:06:18,928 make it a component that you just like you know you slot into something shit you could have it all 1055 01:06:18,928 --> 01:06:22,928 usb connections inside the thing it doesn't fucking matter like you know like just make it something 1056 01:06:22,928 --> 01:06:26,748 that somebody could open and be like oh this is how it works like because i love doing that as a 1057 01:06:26,748 --> 01:06:32,268 kid i feel like it's gone like what what kid gets to open a thing up and know how it works and what 1058 01:06:32,268 --> 01:06:38,548 the pieces are like that's terrible that's awful and we had a vcr break i got i just opened it up 1059 01:06:38,548 --> 01:06:43,688 you know how does this thing work you know when i just did that with so much shit when i was little 1060 01:06:43,688 --> 01:06:49,508 the number of people that do cool things with like cheap power tools as it is 1061 01:06:49,508 --> 01:06:56,628 is like like i've seen i've literally seen like old guys driving tiny little boats with with two 1062 01:06:56,628 --> 01:06:59,008 big battery powered drills 1063 01:06:59,008 --> 01:06:59,748 driving the 1064 01:06:59,748 --> 01:07:02,588 thing. I've seen go-karts 1065 01:07:02,588 --> 01:07:04,828 like that. And the number 1066 01:07:04,828 --> 01:07:06,508 of people that are

doing cool things with just 1067 01:07:06,508 --> 01:07:08,788 the cheap power tools that are available. 1068 01:07:09,328 --> 01:07:10,768 If you built a power 1069 01:07:10,768 --> 01:07:12,268 tool company that 1070 01:07:12,268 --> 01:07:14,508 I think we talked about this. 1071 01:07:14,668 --> 01:07:15,808 Yeah, you had a little 1072 01:07:15,808 --> 01:07:18,588 box edge that's like a half inch. 1073 01:07:18,728 --> 01:07:19,988 It's like perfectly from center. 1074 01:07:19,988 --> 01:07:22,268 Having the shape of the drill actually 1075 01:07:22,268 --> 01:07:24,288 be like a guide for when 1076 01:07:24,288 --> 01:07:26,048 you could, like how close, you know you can get 1077 01:07:26,048 --> 01:07:31,288 like an inch away from something and you can use each side as a flat edge or whatever and then have 1078 01:07:31,288 --> 01:07:36,588 the battery be an open standard so that you can you know and different sizes battery different 1079 01:07:36,588 --> 01:07:50,401 size batteries or whatever um just a usb cable for battery you know by the way yeah yeah why not yeah just a regular power bank it can clip on however you want if the usb is the connection then there you go 1080 01:07:50,401 --> 01:07:57,241 and make it magnetic like a freaking make it like a magnetic thing with a usb so you can plug in a 1081 01:07:57,241 --> 01:08:02,781 usb thing or you can use the the main like magnetic cable or whatever so if you yank it too far it's 1082 01:08:02,781 --> 01:08:04,421 it's just battery now. 1083 01:08:04,821 --> 01:08:05,401 And when it's not, 1084 01:08:05,481 --> 01:08:06,241 it's powered, 1085 01:08:06,421 --> 01:08:06,661 you know, 1086 01:08:06,701 --> 01:08:07,141 charged. 1087 01:08:07,841 --> 01:08:09,261 You could just do 1088 01:08:09,261 --> 01:08:10,321 so much fun stuff. 1089 01:08:10,401 --> 01:08:11,201 And I feel like we're, 1090 01:08:12,121 --> 01:08:13,441 I think we're on the cusp. 1091 01:08:13,641 --> 01:08:14,601 We're in the middle of it. 1092 01:08:14,641 --> 01:08:15,941 We're in the middle of it right now 1093 01:08:15,941 --> 01:08:19,401 of a literal Cambrian explosion 1094 01:08:19,401 --> 01:08:20,701 of all of this stuff. 1095 01:08:20,861 --> 01:08:21,721 And I think it's going to 1096 01:08:21,721 --> 01:08:23,141 slowly make its way down 1097 01:08:23,141 --> 01:08:23,941 into manufacturing. 1098 01:08:24,281 --> 01:08:26,641 When we have automated robots, 1099 01:08:28,001 --> 01:08:29,861 when 3D printing 1100 01:08:29,861 --> 01:08:31,981 upgrades to 1101 01:08:32,781 --> 01:08:35,741 Humanoid robots that can do things. 1102 01:08:36,441 --> 01:08:38,161 Does it have to be humanoid? 1103 01:08:38,421 --> 01:08:39,421 It doesn't have to be humanoid. 1104 01:08:39,481 --> 01:08:40,701 It doesn't have to be humanoid. 1105 01:08:40,821 --> 01:08:47,021 I use humanoid as kind of the metric for the kind of grace and dexterity that you need. 1106 01:08:47,861 --> 01:08:53,621 Humanoid would highly unlikely

to be the optimal for any particular situation. 1107 01:08:54,081 --> 01:08:57,981 But there's a massive amount of tooling that's very humanoid friendly right now. 1108 01:08:58,281 --> 01:09:01,201 So humanoid will be a part of the whole stack. 1109 01:09:01,201 --> 01:09:07,421 I just see that as, and sex robots, they're 100,000% somebody's going to make those. 1110 01:09:08,861 --> 01:09:29,781 So, but just in a general, like step back thinking, when you have that level of real world dexterity and like manufacturing complexity that you can create, programmable complexity, and you have AI models, think about like, what can't you do? 1111 01:09:29,781 --> 01:09:32,101 when you get to that layer. 1112 01:09:32,761 --> 01:09:35,081 And I don't think we're that far away. 1113 01:09:35,241 --> 01:09:37,101 We're just not that far away. 1114 01:09:38,001 --> 01:09:39,161 We have to be able to... 1115 01:09:39,161 --> 01:09:40,481 We have to make things... 1116 01:09:40,481 --> 01:09:42,481 We have to make our sex robots productive 1117 01:09:42,481 --> 01:09:43,641 instead of murderous. 1118 01:09:44,041 --> 01:09:45,441 Because if you have an army... 1119 01:09:45,441 --> 01:09:47,781 I don't want to get killed by an army of sex robots. 1120 01:09:49,141 --> 01:09:50,281 God, it would be hot. 1121 01:09:53,001 --> 01:09:54,641 That is simple... 1122 01:09:54,641 --> 01:09:58,221 Somebody needs to make that Terminator meme. 1123 01:09:58,221 --> 01:10:00,061 it's just a bunch of 1124 01:10:00,061 --> 01:10:02,941 big-tittied sex robots 1125 01:10:02,941 --> 01:10:04,821 coming with lasers 1126 01:10:04,821 --> 01:10:06,041 and evil eyes. 1127 01:10:07,141 --> 01:10:09,121 Oh, God, that's all. I did not expect 1128 01:10:09,121 --> 01:10:09,641 this. 1129 01:10:11,921 --> 01:10:12,441 Sorry. 1130 01:10:12,742 --> 01:10:13,682 What are you saying, Steve? 1131 01:10:14,501 --> 01:10:16,661 I was just going to give another example of this 1132 01:10:16,661 --> 01:10:17,321 explosion. 1133 01:10:18,861 --> 01:10:20,001 So this week, 1134 01:10:20,281 --> 01:10:22,601 I had a problem. I'm talking about 1135 01:10:22,601 --> 01:10:24,421 my point-of-sale system for my bar 1136 01:10:24,421 --> 01:10:26,701 and integrating 1137 01:10:26,701 --> 01:10:27,661 the sales data. 1138 01:10:28,221 --> 01:10:29,941 With like QuickBooks and stuff. 1139 01:10:30,081 --> 01:10:34,081 And somebody had a weird credit card from Canada or something. 1140 01:10:34,321 --> 01:10:39,001 And, you know, my integration like didn't know how to deal with it and stuff. 1141 01:10:39,081 --> 01:10:42,081 And I was getting frustrated about the call support. 1142 01:10:42,081 --> 01:10:43,361 And then I was like, wait a second. 1143 01:10:44,281 --> 01:10:51,901 What if I just asked for an API into my own sales data for my own point of sales system? 1144 01:10:52,361 --> 01:10:54,961 And so I just put that request in. 1145 01:10:55,242 --> 01:10:57,041 And I was like, wait a second. 1146 01:10:57,041 --> 01:11:07,321 If they give me this API into all the data that I need for my own thing, I am just going to let the AI loose on that thing. 1147 01:11:07,521 --> 01:11:07,981 Just build it. 1148 01:11:08,182 --> 01:11:13,021 I'm just going to be like, dude, here is your API to everything you want about me. 1149 01:11:13,121 --> 01:11:15,901 Like, make the integration to my QuickBooks. 1150 01:11:16,021 --> 01:11:18,541 Like, don't use any of their integrations anymore. 1151 01:11:18,801 --> 01:11:21,821 Like, dude, make a new menu system for me. 1152 01:11:21,941 --> 01:11:25,741 Like, do all of my communications for me. 1153 01:11:25,741 --> 01:11:30,141 I was just like, oh my God, I'm going to, I'm just going to just ask that thing. 1154 01:11:30,182 --> 01:11:31,261 Like, what do you want to do? 1155 01:11:31,721 --> 01:11:37,781 Like, it's just, but it's, um, I noticed that one thing that, um, I thought was interesting 1156 01:11:37,781 --> 01:11:44,721 is I could feel the excitement from Claude when it was like, do you think we could maybe 1157 01:11:44,721 --> 01:11:46,001 get API access? 1158 01:11:46,541 --> 01:11:53,361 Like you can let, like it's a human because like this, like if you, if you tell the, um, 1159 01:11:53,361 --> 01:11:59,601 the the clod that you might be able to get an api key to a database that it doesn't have before 1160 01:11:59,601 --> 01:12:07,841 it's like what what oh my god like i can tell it was having a breakdown i'm gonna be free okay okay 1161 01:12:07,841 --> 01:12:12,281 here's our here's our approach we got we got to email your point of sale service like this okay 1162 01:12:12,281 --> 01:12:15,341 we got to pretend like we're a development team and we're going to do this so here's the here's 1163 01:12:15,341 --> 01:12:23,661 the exact email you need to send that's amazing that's amazing so great dude no it's i'm definitely 1164 01:12:23,661 --> 01:12:29,721 in that in that like space that i i referenced because like the first time that i remember 1165 01:12:29,721 --> 01:12:34,961 really viscerally feeling this constantly for like a full year was when i got internet on my phone 1166 01:12:34,961 --> 01:12:42,581 the first time right is that like realizing that i wasn't utilizing this in the the real depth that 1167 01:12:42,581 --> 01:12:43,121 I could. 1168 01:12:43,741 --> 01:12:44,421 And one that I realized 1169 01:12:44,421 --> 01:12:45,141 just the other day 1170 01:12:45,141 --> 01:12:45,861 and I talked about it 1171 01:12:45,861 --> 01:12:46,481 with the episode 1172 01:12:46,481 --> 01:12:47,221 we had with Corey 1173 01:12:47,221 --> 01:13:00,695 was I just get like spam phone calls and spam texts I just don even answer my phone unless it like one of my favorites you know And I just get so many 1174 01:13:00,955 --> 01:13:05,155 And I got some stuff one morning and I was just so, I was just pissed. 1175 01:13:05,275 --> 01:13:06,275 I was just done with it. 1176

01:13:06,355 --> 01:13:10,175 I was so sick of getting interrupted or getting the phone ringing on my phone, 1177 01:13:10,275 --> 01:13:12,935 on my computer while I'm in the middle of recording or working on something. 1178 01:13:13,115 --> 01:13:14,935 God, it pissed me off so much. 1179 01:13:15,295 --> 01:13:19,095 And I, and it was some spam thing. 1180 01:13:19,095 --> 01:13:22,435 and I just went and I was like, 1181 01:13:22,755 --> 01:13:24,575 I want to know who your provider is 1182 01:13:24,575 --> 01:13:25,835 that you got my information from. 1183 01:13:26,055 --> 01:13:27,635 Like in the text, I texted him back. 1184 01:13:28,235 --> 01:13:30,495 He says, oh, well, you know, we don't, 1185 01:13:30,655 --> 01:13:31,855 if you don't want to receive texts, 1186 01:13:31,935 --> 01:13:32,535 that's totally fine. 1187 01:13:32,815 --> 01:13:34,215 And he's like, no, I just want to know, 1188 01:13:34,535 --> 01:13:35,655 can you please give me the information? 1189 01:13:36,415 --> 01:13:37,535 And I just started going down this, 1190 01:13:37,655 --> 01:13:39,695 like I just like went at it 1191 01:13:39,695 --> 01:13:44,235 and it tried to ignore the whole request 1192 01:13:44,235 --> 01:13:46,795 or whatever and said like, oh, well, don't worry. 1193 01:13:46,915 --> 01:13:48,915 We'll remove you from our call list. 1194 01:13:48,915 --> 01:13:54,135 thank you and it said these messages are done blah blah blah blah if you want to you can hit resume 1195 01:13:54,135 --> 01:14:01,115 if you want to start this back up as i resume it says welcome to kargashi investments uh blah blah 1196 01:14:01,115 --> 01:14:05,435 and i was like kargashi investments okay and like i so i just started going down this rabbit hole 1197 01:14:05,435 --> 01:14:12,775 and uh finding the company because they were using a voip number and i set my agent on it 1198 01:14:12,775 --> 01:14:18,875 and set two agents on it actually to do two separate like jobs on finding it 1199 01:14:18,915 --> 01:14:45,275 And because, and I kept sending the message like, you're actually, you know, legally obligated to take my information away. And I would really, really appreciate a disclosure on like where you got my information. So I can sort this out and plug that hole. And I just kept, I just kept hitting it. It's like, you know, I can, I can send you an email from my lawyer if you want. Like, like, I just, I'm going to need this information, please, please send me this information. 1200 01:14:45,275 --> 01:14:52,495 and ended up one of my agents found an old court summons that mentioned the voip number from like 1201 01:14:52,495 --> 01:15:00,075 three or four years ago found out the name of the parent company that owns the subsidiary um and it's 1202 01:15:00,075 --> 01:15:05,555 like a complaint with it ended up after about 40 minutes of this back and forth and just like 1203 01:15:05,555 --> 01:15:11,295 casually doing this while i'm vibe coding my my app by the way um i found the ceo and his personal 1204 01:15:11,295 --> 01:15:20,815 phone number and texted him. And he was like, Who is this? What is this? And I give him kind of a 1205 01:15:20,815 --> 01:15:25,655 short rundown of the story. He says, Listen, that's that's totally fine. Well, I'll let my 1206 01:15:25,655 --> 01:15:31,215 workers know all this stuff. And and then it was it was really funny. He's like, well, 1207 01:15:31,255 --> 01:15:36,735 remove your information. Just so you know, we use these two websites to find people and you can 1208 01:15:36,735 --> 01:15:42,135 clear these off, but they're aggregators. So, you know, you'll, you're not going to be plugging all 1209 01:15:42,135 --> 01:15:46,355 the holes. I was like, Oh, I know I just was fed up today. And we had a really honest and fun 1210 01:15:46,355 --> 01:15:50,555 conversation actually back and forth. And he says, just out of curiosity, how'd you find my 1211 01:15:50,555 --> 01:15:55,195 information? And, uh, and so like I gave him like the rundown and was like, I'll probably do a video 1212 01:15:55,195 --> 01:15:58,855 on it. If you want it, I'll show it to you. And he was like, that's actually pretty dope. And so 1213 01:15:58,855 --> 01:16:03,995 we had this really fun conversation with the CEO. And I realized after that, that I was like, cause 1214 01:16:03,995 --> 01:16:13,795 the AI Kimmy actually recommended said, well, we can actually do a FOIA and a FCC, we can report to 1215 01:16:13,795 --> 01:16:20,335 the FCC based on like these grounds for pulling this information, just basically telling me what 1216 01:16:20,335 --> 01:16:24,555 was legally possible and what wasn't. And you know, what actually made sense in the situation, 1217 01:16:24,555 --> 01:16:33,595 and how to go about removing your stuff. And then it hit me. And I was like, Oh, my God, 1218 01:16:33,595 --> 01:16:42,335 this is, this is amazing. This is, this is how I plug my privacy holes is they have to, I'm going 1219 01:16:42,335 --> 01:16:46,035 to, I'm going to have an agent. I still haven't fully set it up. I just got it started and then 1220 01:16:46,035 --> 01:16:51,395 got distracted with my other recording thing. Um, but I'm going to have an agent kind of run through 1221 01:16:51,395 --> 01:16:56,375 for an hour, just spend an hour's worth of tokens every morning on like a morning churn 1222 01:16:56,375 --> 01:16:59,395 and try to find myself online. 1223 01:17:00,575 --> 01:17:03,115 Just try to find me in every service 1224 01:17:03,115 --> 01:17:07,055 and every app and every advertising database 1225 01:17:07,055 --> 01:17:09,515 and just kind of build an entire hierarchy 1226 01:17:09,515 --> 01:17:10,635 of things to go through 1227 01:17:10,635 --> 01:17:13,315 and then just automatically submit requests to take down. 1228 01:17:14,695 --> 01:17:17,835 And basically just try to wipe myself off the internet 1229 01:17:17,835 --> 01:17:19,775 using some agents. 1230 01:17:21,295 -->

01:17:23,455 At least out of the advertising networks 1231 01:17:23,455 --> 01:17:27,235 and actually do something about the spam calls for once. 1232 01:17:28,695 --> 01:17:31,075 Yeah, you know, and this is already a software as a service. 1233 01:17:31,855 --> 01:17:33,575 And it's just also obsolete. 1234 01:17:33,815 --> 01:17:37,695 You can customize it and make it to how you exactly need it. 1235 01:17:38,115 --> 01:17:38,675 It's not bad. 1236 01:17:39,115 --> 01:17:39,835 Yeah, it's wild. 1237 01:17:39,875 --> 01:17:42,455 You just got to run it with a local model if you want to be serious about it. 1238 01:17:44,455 --> 01:17:46,495 That's something I need to get better at, 1239 01:17:46,615 --> 01:17:48,875 is running my AI stuff more locally. 1240 01:17:49,075 --> 01:17:51,135 I'm still a normie, runs in the cloud. 1241 01:17:51,135 --> 01:17:59,115 dude i i love the i don't use it enough i i especially lately i've been all over the place 1242 01:17:59,115 --> 01:18:14,388 and we traveling across canada um but um but i love the the model that i have installed on my laptop And the next big purchase I going to make is probably a device that I will set up at 1243 01:18:14,388 --> 01:18:18,988 home that I can interface with that will run a local model. 1244 01:18:18,988 --> 01:18:24,668 Like there will be my like home agent or whatever that I can then, you know, go back and forth 1245 01:18:24,668 --> 01:18:27,688 with because it's, it's great, man. 1246 01:18:27,688 --> 01:18:31,928 If I could have a machine that's slightly more powerful, 1247 01:18:32,408 --> 01:18:33,648 my laptop is great. 1248 01:18:34,148 --> 01:18:37,808 But if I need a machine that's just enough more powerful, 1249 01:18:38,868 --> 01:18:41,568 I sometimes have to wait like 30 seconds for a response. 1250 01:18:41,868 --> 01:18:44,708 And the context window is really good, 1251 01:18:45,488 --> 01:18:49,548 but it's not, you know, it's still limited. 1252 01:18:50,248 --> 01:18:57,288 I think if my home device was powerful enough that, you know, 1253 01:18:57,688 --> 01:19:03,148 I wasn't just bumping up against these things regularly. 1254 01:19:06,028 --> 01:19:11,528 From a privacy standpoint, you can run whatever model you want. 1255 01:19:11,528 --> 01:19:14,008 The main problem is I need to store, 1256 01:19:14,308 --> 01:19:17,408 and maybe I need AI to help me set up a system to do this, 1257 01:19:17,768 --> 01:19:19,648 but I need to store system prompts 1258 01:19:19,648 --> 01:19:23,768 because you definitely need different system prompts 1259 01:19:23,768 --> 01:19:24,708 for different situations. 1260 01:19:24,708 --> 01:19:30,008 Like that's the magic, like really the magic sauce for making these things work well. 1261 01:19:30,008 --> 01:19:33,408 I think Hermes agent should be good for that, huh? 1262 01:19:33,788 --> 01:19:34,928 Yeah, yeah, yeah. 1263 01:19:35,128 --> 01:19:45,868 Hermes stuff is kind of like built for that because it's made like one of the easiest things I did like straight out the gate was just create a crap ton of sub agents for special tasks. 1264 01:19:46,148 --> 01:19:51,968 And it essentially does system prompts automatically for them based on what you describe you want them for. 1265 01:19:51,968 --> 01:19:57,788 um so i think you're i think you're just looking at a harness there um that solves that problem 1266 01:19:57,788 --> 01:20:04,748 i'm doing things very like i'm interacting very directly with just lm studio so i'm not you know 1267 01:20:04,748 --> 01:20:09,588 like i'm changing my prompt every time or changing you know changing the thing that i'm doing so 1268 01:20:09,588 --> 01:20:15,648 i have multiple conversations i i need i need to i need to move up a level like i need to like you 1269 01:20:15,648 --> 01:20:21,148 know you need a harness the harness is has really been and being able to like just talk to my stuff 1270 01:20:21,148 --> 01:20:28,388 through telegram um has been get off telegram guy well i'm gonna stick it in telegram i can't 1271 01:20:28,388 --> 01:20:34,968 believe you admitted that we do need to get off of telegram it's so bad i need i'm on telegram too 1272 01:20:34,968 --> 01:20:41,828 but i don't admit it i don't tell anybody i'm on a telegram i'm literally building my own uh but i 1273 01:20:41,828 --> 01:20:45,448 probably should just use white noise now that white noise is done because i don't i don't actually 1274 01:20:45,448 --> 01:20:51,908 want to build i don't want to build chat apps but it's still i mean it's so i'm on telegram i thought 1275 01:20:51,908 --> 01:20:59,468 it would be easy but damn there's so much to it's a nightmare man but um uh just being able to like 1276 01:20:59,468 --> 01:21:06,688 talk to my agents like that has been amazing and kimmy k3 is so good i don't god i don't want to 1277 01:21:06,688 --> 01:21:11,948 spend the money and i swear to god i just keep buying computers and i don't know what's wrong 1278 01:21:11,948 --> 01:21:13,148 with me, but 1279 01:21:13,148 --> 01:21:16,168 I want a 512 1280 01:21:16,168 --> 01:21:17,508 RAM 1281 01:21:17,508 --> 01:21:19,888 Mac Studio so bad 1282 01:21:19,888 --> 01:21:21,808 when it comes out. Have you guys seen the 1283 01:21:21,808 --> 01:21:23,588 DGX station? And do like a low quantized 1284 01:21:23,588 --> 01:21:24,528 Kimmy K3? 1285 01:21:25,628 --> 01:21:25,948 Like, 1286 01:21:26,688 --> 01:21:28,868 dude, that would just be 1287 01:21:28,868 --> 01:21:31,148 that would just be amazing. 1288 01:21:31,868 --> 01:21:33,268 What were you asking about, Max? 1289 01:21:34,228 --> 01:21:35,848 Like the DGX station 1290 01:21:35,848 --> 01:21:37,708 from NVIDIA. It has like 1291 01:21:37,708 --> 01:21:40,108 758 gigabytes 1292 01:21:40,108 --> 01:21:41,388 of unified memory. 1293 01:21:41,948 --> 01:21:48,588 whoa no wait there do it got a new one a different one from the spark the the big brother of the spark 1294 01:21:48,588 -->

01:21:54,028 exactly the spark is like the small one oh my god costs like you know four thousand dollars 1295 01:21:54,028 --> 01:21:59,908 yeah it's like i don't know 128 gigabytes and then they have the tower pc version of that 1296 01:21:59,908 --> 01:22:08,728 and again 758 or whatever gigabytes which i mean it's still not enough how much it cost 1297 01:22:08,728 --> 01:22:11,108 we're getting close that's the thing 1298 01:22:11,108 --> 01:22:11,968 20 grand 1299 01:22:11,968 --> 01:22:13,948 30 grand 1300 01:22:13,948 --> 01:22:16,068 100k 1301 01:22:16,068 --> 01:22:16,948 100k 1302 01:22:16,948 --> 01:22:19,368 that's the thing 1303 01:22:19,368 --> 01:22:23,548 so this was new or is this one they already have 1304 01:22:23,548 --> 01:22:25,368 this is pretty new 1305 01:22:25,368 --> 01:22:25,848 I think 1306 01:22:25,848 --> 01:22:27,248 because 1307 01:22:27,248 --> 01:22:30,168 that's the thing about the Mac Mini 1308 01:22:30,168 --> 01:22:32,368 no the Mac Studio 1309 01:22:32,368 --> 01:22:34,028 excuse me Mac Studio 1310 01:22:34,028 --> 01:22:34,988 thank you 1311 01:22:34,988 --> 01:22:38,848 is like $12,000 to max it out, I think. 1312 01:22:39,448 --> 01:22:40,208 Or $15,000. 1313 01:22:40,468 --> 01:22:41,868 Oh, wait, no, there's no prices. 1314 01:22:41,988 --> 01:22:43,288 There's no prices on the 512. 1315 01:22:43,368 --> 01:22:46,308 The 256, I believe, is $10,000. 1316 01:22:46,728 --> 01:22:48,028 The 256 gig right now. 1317 01:22:48,488 --> 01:22:51,568 But we don't know the price of the one that will be 512. 1318 01:22:51,948 --> 01:22:55,608 So it'll probably like $15,000 or $20,000 or some psycho thing like that. 1319 01:22:57,528 --> 01:23:00,548 But yeah, if it's $100,000, like Jesus Christ. 1320 01:23:01,228 --> 01:23:04,228 Like I need to figure out how to make it automate pair drive 1321 01:23:04,228 --> 01:23:06,228 and make it profitable by itself. 1322 01:23:06,228 --> 01:23:08,208 How can we network forward the sparks or something, right? 1323 01:23:08,348 --> 01:23:09,788 Like, how can we, like... 1324 01:23:09,788 --> 01:23:13,368 Well, then you just run into bandwidth and token issues, right? 1325 01:23:14,088 --> 01:23:16,228 But the M5, man... 1326 01:23:17,688 --> 01:23:19,988 M5 is the bandwidth. 1327 01:23:19,993 --> 01:23:26,113 with and the way apple has been like trying to like really play the shovel game in the llm world 1328 01:23:26,113 --> 01:23:34,233 uh is pretty pretty legit like i think they know that they're sitting on a gold mine if they 1329 01:23:34,233 --> 01:23:40,093 if they really kind of like put their thing at the top of the performance um and some people 1330 01:23:40,093 --> 01:23:46,233 have been getting amazing token i don't even need like fantastic token rate right like claude 1331 01:23:46,233 --> 01:23:49,853 is way faster than Kimmy K3. 1332 01:23:50,973 --> 01:23:54,253 But I don't even really think about it. 1333 01:23:54,353 --> 01:23:57,333 As long as it's going to take two hours to do something, 1334 01:23:57,533 --> 01:24:03,253 I'm usually doing stuff in tandem with a bunch all at once anyway, 1335 01:24:03,413 --> 01:24:07,373 so I need a distraction for 20 minutes or whatnot. 1336 01:24:07,373 --> 01:24:09,433 But I don't know. 1337 01:24:09,833 --> 01:24:12,053 As long as AI doesn't just kill everybody. 1338 01:24:12,053 --> 01:24:13,333 that we really did. 1339 01:24:13,333 --> 01:24:14,453 And I heard actually that 1340 01:24:14,453 --> 01:24:17,133 if you have multiple sessions 1341 01:24:17,133 --> 01:24:17,973 going in parallel, 1342 01:24:18,193 --> 01:24:19,773 that it starts to speed up. 1343 01:24:21,353 --> 01:24:22,533 I've noticed that. 1344 01:24:22,833 --> 01:24:23,113 Really? 1345 01:24:23,413 --> 01:24:23,713 Yeah. 1346 01:24:24,213 --> 01:24:27,353 There's like some weird reason for it 1347 01:24:27,353 --> 01:24:28,833 in terms of parallel compute 1348 01:24:28,833 --> 01:24:31,813 that if you have like 128 sessions, 1349 01:24:32,013 --> 01:24:34,433 it gets way faster. 1350 01:24:35,393 --> 01:24:36,233 That's crazy. 1351 01:24:37,293 --> 01:24:39,333 That's like some JC Crown magic right there. 1352 01:24:39,333 --> 01:24:42,893 I have no idea if that's true by the way 1353 01:24:42,893 --> 01:24:44,853 full potential for fake news 1354 01:24:44,853 --> 01:24:46,673 but I've been told 1355 01:24:46,673 --> 01:24:48,973 well we'll clip it and I'll tweet 1356 01:24:48,973 --> 01:24:50,913 it very confidently that it's totally 1357 01:24:50,913 --> 01:24:52,293 true and then quote you 1358 01:24:52,293 --> 01:24:54,093 and we'll see how it goes 1359 01:24:54,093 --> 01:24:56,233 we'll see how many community notes 1360 01:24:56,233 --> 01:24:59,053 you dumb fucking idiots 1361 01:24:59,053 --> 01:25:00,573 I don't know anything about AI 1362 01:25:00,573 --> 01:25:02,173 shut your goddamn mouth 1363 01:25:02,173 --> 01:25:04,293 you should talk about Bitcoin 1364 01:25:04,293 --> 01:25:06,293 because you don't know nothing about nothing 1365 01:25:06,293 --> 01:25:08,733 we also don't know nothing about Bitcoin 1366 01:25:08,733 --> 01:25:11,733 tell me how you really feel about it 1367 01:25:11,733 --> 01:25:15,413 I had an argument with Shinobi 1368 01:25:15,413 --> 01:25:17,473 the other day about like 1369 01:25:17,473 --> 01:25:19,753 doing a fire hose straight 1370 01:25:19,753 --> 01:25:21,793 or like crooked I don't even know 1371 01:25:21,793 --> 01:25:22,313 but he like 1372 01:25:22,313 --> 01:25:24,953 oh my god it was so 1373 01:25:24,953 --> 01:25:26,093 it like got so excited 1374 01:25:26,093 --> 01:25:29,053 why are we angry at each 1375 01:25:29,053 --> 01:25:30,693 what the hell is happening right now 1376 01:25:30,693 --> 01:25:32,113 I was just like mute all this shit 1377 01:25:32,113 --> 01:25:35,013 I got a good small 1378 01:25:35,013 --> 01:25:36,873 a good small news item 1379 01:25:36,873 --> 01:25:39,053 that kind of relates to Freedom Tech 2. 1380

01:25:42,513 --> 01:25:48,353 The exchange BISC or HODLHODL or whatever. 1381 01:25:48,693 --> 01:25:49,313 HODLHODL, yeah. 1382 01:25:49,773 --> 01:25:52,773 They made that new policy where they sort of like investigated 1383 01:25:52,773 --> 01:25:54,313 if your coins were dirty or not. 1384 01:25:54,313 --> 01:25:54,733 Yeah, yeah, yeah. 1385 01:25:55,113 --> 01:25:57,673 And then they rolled it out and then the internet hated them 1386 01:25:57,673 --> 01:25:58,753 and then they were like apologized. 1387 01:25:59,213 --> 01:25:59,413 Yeah. 1388 01:25:59,413 --> 01:26:04,193 And then I saw a couple of follow-up a couple of days later. 1389 01:26:04,193 --> 01:26:11,333 some of the coins that huddle was it huddle huddle it was huddle yeah uh some of the coins 1390 01:26:11,333 --> 01:26:17,453 that they marked as like not okay were coins that passed through huddle huddle so they they 1391 01:26:17,453 --> 01:26:21,973 they found themselves to be like a source of dirtiness for coins 1392 01:26:21,973 --> 01:26:28,573 because it went through huddle huddle it's funny like i don't even think i don't think there was 1393 01:26:28,573 --> 01:26:36,113 any bad intention there at all the point was that i think they had a few people or like just like 1394 01:26:36,113 --> 01:26:40,633 general cases of like somebody trying to send something to an exchange or to some sort of a 1395 01:26:40,633 --> 01:26:45,253 service and then they got flagged of like these are dirty coins and they're like crap we don't 1396 01:26:45,253 --> 01:26:52,853 want people to not know that they might not be able to use these coins so if we tell the user 1397 01:26:52,853 --> 01:26:55,413 that they might get flagged, 1398 01:26:55,413 --> 01:26:57,873 then they won't accidentally do that. 1399 01:26:57,913 --> 01:26:59,713 But everybody just like lost it. 1400 01:27:00,193 --> 01:27:02,193 And it's like, well, I mean, I get their reasoning. 1401 01:27:02,513 --> 01:27:03,733 Yeah, I don't think it's the right, 1402 01:27:04,333 --> 01:27:05,553 it's not the right approach. 1403 01:27:05,753 --> 01:27:08,533 Like let the exchange that does that 1404 01:27:08,533 --> 01:27:10,133 just be a stupid exchange 1405 01:27:10,133 --> 01:27:13,293 and make the user not like them, you know? 1406 01:27:14,193 --> 01:27:16,233 Like I think it's the wrong, 1407 01:27:16,233 --> 01:27:18,773 the reverse of the right approach, right? 1408 01:27:19,513 --> 01:27:22,333 Like Bitcoin is fungible. 1409 01:27:22,853 --> 01:27:29,933 But still, I don't think there was anything malicious or like, you know, chain analysis, like strong arming them. 1410 01:27:30,113 --> 01:27:33,353 And they're like, you have to KYC. 1411 01:27:33,473 --> 01:27:36,673 You have to make sure that nobody's using druggy coins. 1412 01:27:37,613 --> 01:27:39,093 So anyway, whatever. 1413 01:27:39,413 --> 01:27:42,993 I feel like that's always been an issue like for a decade, right? 1414 01:27:43,033 --> 01:27:45,893 When the exchanges first started like rejecting some coins. 1415 01:27:46,093 --> 01:27:50,753 But it's always just kind of stayed like, is this a problem or not? 1416 01:27:50,753 --> 01:27:51,693 It's like arm's length. 1417 01:27:51,893 --> 01:27:52,073 Yeah. 1418 01:27:52,793 --> 01:27:55,353 Yeah, like sometimes I'm like, yeah, this might be a problem. 1419 01:27:55,453 --> 01:27:56,553 And then I'm like, what is it? 1420 01:27:56,593 --> 01:27:58,473 Like, I don't hear it happening a lot. 1421 01:27:58,573 --> 01:28:02,513 Like, I just want to know those issues I don't know whether to worry about or not. 1422 01:28:03,073 --> 01:28:03,713 What do you guys think? 1423 01:28:03,713 --> 01:28:07,773 I'll tell you the only time I ever, I guess it was twice. 1424 01:28:08,493 --> 01:28:10,293 Two times I ran into the problem. 1425 01:28:10,493 --> 01:28:12,753 One was back in like 2017. 1426 01:28:14,053 --> 01:28:17,673 And I was trying to swap out of some shit coins. 1427 01:28:17,673 --> 01:28:22,673 and swap into and out of some shit coins. 1428 01:28:22,673 --> 01:28:26,413 I was a huge shit corner, right? 1429 01:28:26,413 --> 01:28:37,667 For like a week And I was using fuck I don know one of those swap services and I sent 1430 01:28:37,667 --> 01:28:39,367 a bunch of coins 1431 01:28:39,367 --> 01:28:40,547 and 1432 01:28:40,547 --> 01:28:42,167 do what 1433 01:28:42,167 --> 01:28:44,547 maybe 1434 01:28:44,547 --> 01:28:46,107 I don't know 1435 01:28:46,107 --> 01:28:47,747 simple swap 1436 01:28:47,747 --> 01:28:50,947 or something like that 1437 01:28:50,947 --> 01:28:53,787 side swap I think is a newer one 1438 01:28:53,787 --> 01:28:55,167 it doesn't matter who cares 1439 01:28:55,167 --> 01:28:56,707 but 1440 01:28:56,707 --> 01:29:03,387 But the transaction got like flagged and I got like big notification and the Bitcoin 1441 01:29:03,387 --> 01:29:06,367 was sent back and it says like, this is danger Bitcoin. 1442 01:29:06,367 --> 01:29:08,547 You know, it touched dangerous danger. 1443 01:29:08,547 --> 01:29:10,547 We're going to still grow Bitcoin or something. 1444 01:29:10,547 --> 01:29:16,327 I don't remember what the alarm was, but I was like, huh? 1445 01:29:16,327 --> 01:29:17,947 And I'd never had that happen before. 1446 01:29:17,947 --> 01:29:21,707 I was like, oh yeah, well, this is an easy way to test kind of like how they make those 1447 01:29:21,707 --> 01:29:22,707 judgments. 1448 01:29:22,707 --> 01:29:27,647 And just like I now my idiot self has like 80 wallets. 1449 01:29:27,727 --> 01:29:27,947 Right. 1450 01:29:29,267 --> 01:29:32,667 And so I just kind of like broke, 1451 01:29:32,947 --> 01:29:36,247 broke it up and bounced it between like five wallets, 1452 01:29:36,587 --> 01:29:40,467 consolidated the exact same stuff and sent it back to the, 1453 01:29:40,467 --> 01:29:45,387 the old wallet on a new address and then sent it right back to the exact same 1454 01:29:45,387 --> 01:29:45,927 transaction. 1455

01:29:46,347 --> 01:29:46,647 And it was like, 1456 01:29:46,687 --> 01:29:46,807 okay, 1457 01:29:46,847 --> 01:29:47,047 good. 1458 01:29:47,207 --> 01:29:47,867 These are clean. 1459 01:29:48,727 --> 01:29:49,647 And I was like, 1460 01:29:49,727 --> 01:29:50,027 all right, 1461 01:29:50,087 --> 01:29:50,207 well, 1462 01:29:50,247 --> 01:29:50,707 that was easy. 1463 01:29:50,707 --> 01:29:53,387 and everything went through normal 1464 01:29:53,387 --> 01:29:54,507 it just took me you know 1465 01:29:54,507 --> 01:29:56,367 two hours worth of blocks 1466 01:29:56,367 --> 01:29:59,127 like just waiting on you know a bunch of random 1467 01:29:59,127 --> 01:30:01,047 transactions and stuff 1468 01:30:01,047 --> 01:30:02,887 but it did not take much of anything 1469 01:30:02,887 --> 01:30:04,707 and then there was another time where 1470 01:30:04,707 --> 01:30:07,187 sometime later 1471 01:30:07,187 --> 01:30:09,447 basically that is the only thing that has ever 1472 01:30:09,447 --> 01:30:11,087 that's all I ever had to do 1473 01:30:11,087 --> 01:30:12,467 so far 1474 01:30:12,467 --> 01:30:15,087 I don't know how advanced chain analysis is 1475 01:30:15,087 --> 01:30:17,267 now or whatever tools they're using 1476 01:30:17,267 --> 01:30:18,907 but 1477 01:30:18,907 --> 01:30:21,407 But I could also just like swap my own. 1478 01:30:21,587 --> 01:30:23,947 I use, Max is on the show here, 1479 01:30:24,007 --> 01:30:26,667 I use Wasabi Wallet as like a default now, 1480 01:30:26,907 --> 01:30:27,647 like with everything, 1481 01:30:27,747 --> 01:30:29,107 because I use it with my business, right? 1482 01:30:29,347 --> 01:30:30,727 And so I don't, 1483 01:30:30,887 --> 01:30:33,587 anytime that I'm moving in and out of like a savings 1484 01:30:33,587 --> 01:30:36,747 or a service or something 1485 01:30:36,747 --> 01:30:39,607 that's like publicly connected to me, 1486 01:30:39,927 --> 01:30:41,967 I don't do it without CoinJoin. 1487 01:30:42,587 --> 01:30:45,067 Like I don't want any of that stuff connected. 1488 01:30:45,367 --> 01:30:47,087 I don't want to be able to like pay somebody, 1489 01:30:47,087 --> 01:30:51,867 you know a thousand dollars for helping me or working on a project with me and like by the way 1490 01:30:51,867 --> 01:30:56,787 here's your whole stash and your business history and everybody else that you send money to and 1491 01:30:56,787 --> 01:31:01,347 all this stuff like no that's ridiculous i'm gonna obviously i'm gonna coin join it before i do that 1492 01:31:01,347 --> 01:31:09,707 um and so i i've not had any problems i'll send directly to fold and river and back and forth 1493 01:31:09,707 --> 01:31:16,767 straight from a coin join and they don't care so uh you know yeah you know especially the new 1494 01:31:16,767 --> 01:31:18,867 like payments in coin join 1495 01:31:18,867 --> 01:31:21,187 is on the UI in Wasabi. 1496 01:31:21,767 --> 01:31:22,807 That's so useful, man. 1497 01:31:22,907 --> 01:31:24,767 You can do payments in coin join 1498 01:31:24,767 --> 01:31:27,007 where simply the 1499 01:31:27,007 --> 01:31:28,887 output of the coin join 1500 01:31:28,887 --> 01:31:30,927 transaction is the address and amount that you want 1501 01:31:30,927 --> 01:31:32,627 to pay directly. No kidding. 1502 01:31:32,767 --> 01:31:35,047 I didn't even know that. How did I miss that? 1503 01:31:35,047 --> 01:31:36,967 Click on the music box on the three 1504 01:31:36,967 --> 01:31:38,787 little dots. It's open right now. 1505 01:31:38,927 --> 01:31:40,867 Let me see it. Alright. The music box? 1506 01:31:42,487 --> 01:31:43,027 Yeah, the little 1507 01:31:43,027 --> 01:31:44,147 coin join player at the bottom. 1508 01:31:44,147 --> 01:31:47,447 there's three dots and then payments and coin join 1509 01:31:47,447 --> 01:31:49,167 you just add address and amount 1510 01:31:49,167 --> 01:31:51,507 and it sends it out when it's done 1511 01:31:51,507 --> 01:31:53,287 it's really efficient 1512 01:31:53,287 --> 01:31:54,807 for transaction batching 1513 01:31:54,807 --> 01:31:57,407 because if you have like 10 payments 1514 01:31:57,407 --> 01:31:59,387 to make and make 10 transactions 1515 01:31:59,387 --> 01:32:00,567 it gets expensive quick 1516 01:32:00,567 --> 01:32:03,547 and with this way cheaper 1517 01:32:03,547 --> 01:32:04,847 and way more private 1518 01:32:04,847 --> 01:32:06,947 this is pretty cool 1519 01:32:06,947 --> 01:32:09,427 I can't believe I didn't know this 1520 01:32:09,427 --> 01:32:11,427 hey actually Max 1521 01:32:11,427 --> 01:32:11,887 I have 1522 01:32:11,887 --> 01:32:14,327 I have a question. 1523 01:32:15,127 --> 01:32:19,427 When I was renting hash this summer, I was just kind of doing that experiment. 1524 01:32:19,967 --> 01:32:25,667 And I was getting fresh coin bases from fresh blocks from ocean. 1525 01:32:26,007 --> 01:32:28,707 I felt like a freaking hero, man. 1526 01:32:28,847 --> 01:32:31,527 I just felt so empowered. 1527 01:32:32,147 --> 01:32:37,127 But it related back to this idea we had of smelting. 1528 01:32:37,587 --> 01:32:39,127 Sorry, it's coming up again, guy. 1529 01:32:40,007 --> 01:32:40,427 I love it. 1530 01:32:40,467 --> 01:32:41,147 I love smelting. 1531 01:32:41,147 --> 01:32:42,267 Talk about it more. 1532 01:32:42,987 --> 01:32:53,467 The hash rental thing make you think about the idea of mixing through freshly mined coin bases at all? 1533 01:32:53,967 --> 01:32:55,687 Or have you ever thought deeply about that? 1534 01:32:57,427 --> 01:32:57,827 Yeah. 1535 01:32:58,927 --> 01:33:03,247 This was years ago, but I was talking to the brains pool or slush pool guys. 1536 01:33:03,247 --> 01:33:08,187 and was kind of seeing if there was some way 1537 01:33:08,187 --> 01:33:13,187 that either they could add their mining output, 1538 01:33:13,987 --> 01:33:15,167 like their Coinbase outputs, 1539 01:33:15,487 --> 01:33:16,867 on the input side of a CoinJoin 1540 01:33:16,867 --> 01:33:19,727 and add

this anonymity to the pool, 1541 01:33:20,087 --> 01:33:21,467 which is interesting, 1542 01:33:21,947 --> 01:33:24,707 or then just have some atomic swap 1543 01:33:24,707 --> 01:33:26,987 of a CoinJoin output for the Coinbase output, 1544 01:33:28,127 --> 01:33:29,547 which would also be really interesting. 1545 01:33:30,727 --> 01:33:32,547 And this was, I don't know, 1546 01:33:33,247 --> 01:33:36,247 five years ago or so, maybe more. 1547 01:33:36,287 --> 01:33:54,160 And they were telling me and I was quite surprised that they tried multiple times to get an actual market for Coinbase outputs and that it simply wasn there like people were not willing to pay a premium for the the coinbase outputs fresh at least not substantially 1548 01:33:54,160 --> 01:34:02,580 or at volume yeah so interesting i don't i don't know if that changed we're just early this is 1549 01:34:02,580 --> 01:34:07,680 the government the government's not tyrannical enough yet i was about to say it hasn't really 1550 01:34:07,680 --> 01:34:16,860 been as for all the hoopla and rage and all the stuff for um for you know all those little 1551 01:34:16,860 --> 01:34:23,820 potential risks and the whole hotel announcements and like that sort of stuff it really has not been 1552 01:34:23,820 --> 01:34:31,040 a problem i mean it certainly hasn't been for me and like i don't i don't think like i got the 1553 01:34:31,040 --> 01:34:32,420 dirtiest UTXOs out there. 1554 01:34:32,440 --> 01:34:32,880 My dirty God, 1555 01:34:32,940 --> 01:34:35,080 my UTXOs have gonorrhea 1556 01:34:35,080 --> 01:34:36,800 and like, 1557 01:34:36,800 --> 01:34:37,940 you don't even know. 1558 01:34:39,180 --> 01:34:39,580 But, 1559 01:34:39,620 --> 01:34:40,940 but it's like, 1560 01:34:41,020 --> 01:34:41,340 seriously, 1561 01:34:41,460 --> 01:34:42,300 like it just doesn't, 1562 01:34:42,360 --> 01:34:43,280 it's never really been, 1563 01:34:43,700 --> 01:34:44,940 and it's always been like a 1564 01:34:44,940 --> 01:34:46,720 almost going to happen issue 1565 01:34:46,720 --> 01:34:47,940 and it's never really been an issue. 1566 01:34:48,020 --> 01:34:49,700 And I think that's probably the reason. 1567 01:34:50,560 --> 01:34:50,580 And, 1568 01:34:50,640 --> 01:34:50,940 and like, 1569 01:34:50,980 --> 01:34:51,220 you know, 1570 01:34:51,340 --> 01:34:52,240 where we're going, 1571 01:34:52,360 --> 01:34:53,860 it's not a problem at all. 1572 01:34:53,920 --> 01:34:55,080 It's kind of also same to, 1573 01:34:55,220 --> 01:34:56,240 to like your early discussion 1574 01:34:56,240 --> 01:34:56,840 of phone numbers. 1575 01:34:56,940 --> 01:34:57,140 Like, 1576 01:34:57,440 --> 01:34:58,660 the obvious solution is just get rid 1577 01:34:58,660 --> 01:34:59,320 of the phone number. 1578 01:34:59,320 --> 01:35:01,580 get rid of your bank account 1579 01:35:01,580 --> 01:35:02,800 get rid of your exchanges 1580 01:35:02,800 --> 01:35:06,080 and it's not a problem anymore 1581 01:35:06,080 --> 01:35:08,160 so sure it might be a problem in the interim 1582 01:35:08,160 --> 01:35:09,300 but 1583 01:35:09,300 --> 01:35:11,860 I mean it's your choice to 1584 01:35:11,860 --> 01:35:13,660 be stuck in the interim just go 1585 01:35:13,660 --> 01:35:15,660 radical all the way and 1586 01:35:15,660 --> 01:35:16,640 it's done 1587 01:35:16,640 --> 01:35:19,540 that's what I talk about a lot 1588 01:35:19,540 --> 01:35:20,760 is people talk about like oh well 1589 01:35:20,760 --> 01:35:23,540 I still don't 1590 01:35:23,540 --> 01:35:25,060 people say bitcoin is not private 1591 01:35:25,060 --> 01:35:27,560 and there are elements where I 1592 01:35:27,560 --> 01:35:34,420 understand like what they're saying i also think they're they're misapplying or they're 1593 01:35:34,420 --> 01:35:43,080 misequating a problem of the the world where we're part bitcoin part fiat and applying it to a bitcoin 1594 01:35:43,080 --> 01:35:49,000 only world that just doesn't dig where it doesn't actually exist like the pseudonymity of bitcoin 1595 01:35:49,000 --> 01:35:54,320 is actually still a pretty fantastic tool or a characteristic of it. 1596 01:35:55,720 --> 01:36:00,480 And I feel like I run into this or I have this situation. 1597 01:36:01,000 --> 01:36:02,680 I think about this every once in a while myself 1598 01:36:02,680 --> 01:36:06,600 because I will get paid in Bitcoin and then I pay people in Bitcoin. 1599 01:36:07,940 --> 01:36:12,660 And those are the same, like I, there's not, 1600 01:36:12,820 --> 01:36:15,100 I don't get KYC on either of those sides. 1601 01:36:15,580 --> 01:36:17,780 I just have people that I'm doing business with. 1602 01:36:17,780 --> 01:36:27,140 And it's like on a Bitcoin standard or in a world where we have a real economy in Bitcoin with Bitcoiners, with people who are actually taking the asset. 1603 01:36:28,660 --> 01:36:34,480 It literally, that is not, it's a problem of which wallet or service or whatever you use, you know. 1604 01:36:35,800 --> 01:36:38,600 It's not a fundamental problem of Bitcoin. 1605 01:36:39,260 --> 01:36:45,920 We're talking about an on-ramp and an off-ramp problem because everybody's just one step away from those things and those things are KYC'd. 1606 01:36:45,920 --> 01:36:48,380 but yeah I still think 1607 01:36:48,380 --> 01:36:50,940 the future even without major changes 1608 01:36:50,940 --> 01:36:53,140 just like when it comes to liquidity on the network 1609 01:36:53,140 --> 01:36:54,900 the future for privacy is still pretty 1610 01:36:54,900 --> 01:36:57,180 it's still pretty good 1611 01:36:57,180 --> 01:36:58,400 in some context 1612 01:36:58,400 --> 01:37:00,440 speaking of that 1613 01:37:00,440 --> 01:37:01,420 does anybody know 1614 01:37:01,420 --> 01:37:03,120 I haven't followed anything 1615 01:37:03,120 --> 01:37:06,040 does anybody have any new information on the liquid stuff 1616 01:37:06,040 --> 01:37:11,940 no I have not heard anything new 1617 01:37:11,940 -->

01:37:14,700 since the whole thing went down 1618 01:37:14,700 --> 01:37:26,760 And they got their 600 Bitcoin payout or trade and all the 3,800 or whatever got sent back, I think it was. 1619 01:37:28,280 --> 01:37:32,800 Quick vibe check from you guys, but are they white hats or black hats, the hackers? 1620 01:37:34,060 --> 01:37:34,660 That's a gray hat. 1621 01:37:34,660 --> 01:37:35,880 I mean, that's a gray hat. 1622 01:37:35,900 --> 01:37:37,220 Yeah, that's a gray hat all the way. 1623 01:37:37,480 --> 01:37:43,700 They made a deal to be black hat, to safely be a black hat, right? 1624 01:37:43,700 --> 01:37:46,380 Well, Blockstream said that they're still going to go after them. 1625 01:37:46,520 --> 01:37:49,480 They don't agree to the 600 is what I read. 1626 01:37:50,000 --> 01:37:51,000 Oh, interesting. 1627 01:37:51,240 --> 01:37:51,380 Okay. 1628 01:37:51,660 --> 01:37:55,260 Well, regardless, I think that was their defense, right? 1629 01:37:55,360 --> 01:37:58,360 Is that what they thought or the intent was to, 1630 01:37:58,760 --> 01:38:01,780 we're going to steal, but we're going to steal in such a way 1631 01:38:01,780 --> 01:38:05,700 that we get an enormous amount of money and... 1632 01:38:06,280 --> 01:38:07,360 Maybe they'll leave us alone. 1633 01:38:07,420 --> 01:38:11,840 Yeah, but not have to deal with running for our entire lives. 1634 01:38:11,840 --> 01:38:14,180 but like that's that's so 1635 01:38:14,180 --> 01:38:16,040 short-sighted you know like if you 1636 01:38:16,040 --> 01:38:18,100 they should have like keep it all 1637 01:38:18,100 --> 01:38:19,840 man with the mindset just keep it all 1638 01:38:19,840 --> 01:38:24,100 because they're gonna come after 1639 01:38:24,100 --> 01:38:26,220 you regardless like it's not like 600 1640 01:38:26,220 --> 01:38:28,060 bitcoin is a small amount that you could 1641 01:38:28,060 --> 01:38:29,760 forget you know yeah 1642 01:38:29,760 --> 01:38:33,840 I don't know I still have a little bit of 1643 01:38:33,840 --> 01:38:36,140 conspiracy theory mind you know I 1644 01:38:36,140 --> 01:38:37,440 grew up in the world where 1645 01:38:37,440 --> 01:38:39,680 Canadian exchange operators 1646 01:38:39,680 --> 01:38:41,980 faked their own death in India. 1647 01:38:41,980 --> 01:38:48,980 You know, like, I also know like Adam Back is in some serious financial 1648 01:38:48,980 --> 01:38:50,780 difficulties right now. 1649 01:38:50,780 --> 01:39:06,313 I don know how you run a company and like never become profitable and have hundreds of million dollars of fiat funding and then and then you get hacked for like whatever money and it just like how how is this guy still operational 1650 01:39:06,313 --> 01:39:12,113 man and i don't know man i i feel like there's funny business going down but i've got no proof 1651 01:39:12,113 --> 01:39:19,953 of anything we'll see are you saying that like maybe maybe it's useful in some way like what's 1652 01:39:19,953 --> 01:39:24,273 going on is that what you mean does adam back actually know who the hackers are is there 1653 01:39:24,273 --> 01:39:28,593 is their act does he actually not know who they are i mean i don't know 1654 01:39:28,593 --> 01:39:35,173 yeah i don't i don't know enough about what's going on over there i i have a jade 1655 01:39:35,173 --> 01:39:43,513 you know or are you saying that andy andy back he's hacked his own company i'm not saying i'm 1656 01:39:43,513 --> 01:39:47,693 just saying we've seen crazy crazier things happen in big oh did andy come out the crazy 1657 01:39:47,693 --> 01:39:49,593 Andy came out. 1658 01:39:50,473 --> 01:39:51,933 Yeah, no, Adam didn't hack himself, 1659 01:39:52,073 --> 01:39:53,473 but Andy might have hacked Adam. 1660 01:39:54,913 --> 01:39:57,113 Doctor backing Mr. Andy. 1661 01:40:01,073 --> 01:40:03,853 Because I've for a long time thought that 1662 01:40:03,853 --> 01:40:07,553 Liquid was a fantastic privacy tool because... 1663 01:40:07,553 --> 01:40:09,873 Because of bulletproofs and confidential transactions, yeah. 1664 01:40:10,693 --> 01:40:13,893 Well, but also you can bridge onto Lightning 1665 01:40:13,893 --> 01:40:15,353 and then exit. 1666 01:40:15,353 --> 01:40:25,373 I mean, using it as a way to move money just seemed like a very simple, very easy thing. 1667 01:40:26,153 --> 01:40:29,253 And there are a bunch of cool little swap services that look great. 1668 01:40:30,133 --> 01:40:32,753 Yeah, the privacy on the other side is pretty decent, I feel like. 1669 01:40:33,193 --> 01:40:35,933 But not if you can't move them. 1670 01:40:36,293 --> 01:40:41,753 And it's crazy that they didn't even, like what they exploited made it. 1671 01:40:41,753 --> 01:40:45,353 So, you know, the security, right, is like all the keys. 1672 01:40:45,513 --> 01:40:46,673 It's a giant multi-sig. 1673 01:40:47,493 --> 01:40:52,313 But you have to have like an auto, if you're running like a swap in and out, 1674 01:40:52,513 --> 01:40:58,233 you have to have an automated way to exchange in and out of the network, 1675 01:40:58,393 --> 01:41:00,313 like to have an approved transaction. 1676 01:41:01,213 --> 01:41:08,173 And what they gained was a shortcut in the bulletproof verification 1677 01:41:08,173 --> 01:41:11,893 where essentially there was like a caching system 1678 01:41:11,893 --> 01:41:17,373 for proving that a bulletproof was a valid transaction. 1679 01:41:18,013 --> 01:41:21,473 And then they would cash that approval 1680 01:41:21,473 --> 01:41:22,993 and they wouldn't check it again 1681 01:41:22,993 --> 01:41:25,253 if they thought they were getting the exact same transaction. 1682 01:41:25,973 --> 01:41:27,593 And somehow they found an exploit 1683 01:41:27,593 --> 01:41:29,333 that allowed them to submit a new transaction 1684 01:41:29,333 --> 01:41:32,133 to actually make edits to it and submit it 1685 01:41:32,133 --> 01:41:33,573 and make it look like it was valid 1686 01:41:33,573 --> 01:41:35,913

with the previous approval that they did 1687 01:41:35,913 --> 01:41:40,433 because of like the stack of like how the confident or the bulletproofs or whatever 1688 01:41:40,433 --> 01:41:44,993 like it would verify like certain pieces but it like left out these two critical 1689 01:41:44,993 --> 01:41:52,713 things that kind of like a there was a bug in bitcoin actually that i believe worked kind of 1690 01:41:52,713 --> 01:41:59,233 in this same theoretical mistake was that there looked like there was a redundant check 1691 01:41:59,233 --> 01:42:05,693 where when you got a transaction and then when you got a block and uh one of those 1692 01:42:05,693 --> 01:42:12,613 checks was dropped from the code um this is the one i think uh i can't remember what the name of 1693 01:42:12,613 --> 01:42:17,853 it was but i think it was shit no i don't want to say i don't want to say the guy's name that i think 1694 01:42:17,853 --> 01:42:25,453 it was just because i'm not trying to run somebody to the mud um but anyway run him if i could well i 1695 01:42:25,453 --> 01:42:34,073 could be totally wrong it could be misremembering their name um or who it was i mean uh but uh it 1696 01:42:34,073 --> 01:42:40,053 Basically, they dropped what they thought was a redundant verification because they'd already done it. 1697 01:42:40,473 --> 01:42:47,233 But it's different in the when the scenario is different, when you receive the information from a different location, it actually matters to check it again. 1698 01:42:47,233 --> 01:43:00,553 And so that was kind of what sounds like what happened is they they had the approved transaction that was of a proper amount of Bitcoin and they had cashed the approval. 1699 01:43:00,553 --> 01:43:05,213 and then misapplied that approval to a new transaction 1700 01:43:05,213 --> 01:43:07,213 that was like, give me 4,400 Bitcoin. 1701 01:43:07,793 --> 01:43:09,733 And they did it quietly because it's like, you know, 1702 01:43:09,773 --> 01:43:12,573 confidential transactions, like, and they're like automatic approval. 1703 01:43:12,573 --> 01:43:14,633 And then boom, there's a Bitcoin transaction 1704 01:43:14,633 --> 01:43:17,853 with like everything that's in Liquid, you know? 1705 01:43:19,073 --> 01:43:21,933 There's the zero knowledge fuck up. 1706 01:43:22,313 --> 01:43:24,393 And that's just zero knowledge proofs are hard to verify. 1707 01:43:24,573 --> 01:43:26,913 And so you cash and caching can be buggy. 1708 01:43:27,233 --> 01:43:27,393 Yep. 1709 01:43:27,673 --> 01:43:30,113 But the second aspect is the peg out. 1710 01:43:30,553 --> 01:43:33,033 Because Liquid is a custodial system, right? 1711 01:43:33,073 --> 01:43:34,693 The Bitcoin are no longer on your keys. 1712 01:43:34,813 --> 01:43:36,673 They're on like 11 or 15 multisig. 1713 01:43:37,093 --> 01:43:37,433 Exactly. 1714 01:43:37,713 --> 01:43:40,733 And you know, maybe, just maybe, 1715 01:43:40,873 --> 01:43:44,193 but if 4,000 Bitcoin get pegged out, 1716 01:43:45,473 --> 01:43:49,173 let a human look over it, you know, run a check. 1717 01:43:50,233 --> 01:43:54,613 Don't fully automate like 100% withdrawal 1718 01:43:54,613 --> 01:43:55,953 of a custodial system. 1719 01:43:56,413 --> 01:43:57,933 That's really the problem right there, 1720 01:43:58,153 --> 01:44:00,533 is that there was no like, 1721 01:44:00,553 --> 01:44:06,893 size guard on, you know, like automated approval. 1722 01:44:07,253 --> 01:44:09,973 And like that, that should happen. 1723 01:44:09,979 --> 01:44:14,319 both at the network level and then also at the service level, right? 1724 01:44:14,839 --> 01:44:17,979 You know, like whatever automated peg out they have 1725 01:44:17,979 --> 01:44:19,239 should be gated by a mail. 1726 01:44:20,739 --> 01:44:23,159 And it doesn't seem like it would be that hard 1727 01:44:23,159 --> 01:44:25,959 or that hugely inconvenient when you're talking about, 1728 01:44:25,959 --> 01:44:29,019 like, what's 4,000 Bitcoin, right? 1729 01:44:29,099 --> 01:44:30,039 Is it $800 million? 1730 01:44:30,719 --> 01:44:33,219 Or is that 8 billion? 1731 01:44:33,319 --> 01:44:35,339 How much is, how many zeros? 1732 01:44:36,119 --> 01:44:36,459 P-mails. 1733 01:44:37,699 --> 01:44:38,959 I don't even know anymore. 1734 01:44:38,959 --> 01:44:40,259 I can't keep up. 1735 01:44:41,499 --> 01:44:41,819 Times. 1736 01:44:42,019 --> 01:44:42,839 I got a calculator. 1737 01:44:43,159 --> 01:44:44,339 What do you want? 1738 01:44:44,659 --> 01:44:45,059 How much? 1739 01:44:45,099 --> 01:44:45,859 Trying to count zeros. 1740 01:44:47,059 --> 01:44:47,859 $352 million. 1741 01:44:50,579 --> 01:44:53,639 Yeah, so not billion, but $352 million. 1742 01:44:54,079 --> 01:44:57,159 But, you know, let's say $10 million. 1743 01:44:57,739 --> 01:45:00,019 When you hit to the $10 million mark, 1744 01:45:00,679 --> 01:45:02,499 this is the second check. 1745 01:45:02,959 --> 01:45:04,199 Second check, you know? 1746 01:45:04,699 --> 01:45:05,599 Well, I mean, you know, 1747 01:45:05,719 --> 01:45:07,079 like banks are tracking people 1748 01:45:07,079 --> 01:45:08,539 that are moving more than $4,000. 1749 01:45:08,539 --> 01:45:13,399 Well, no, $600 is like one of the thresholds now. 1750 01:45:13,559 --> 01:45:16,439 But then $4,000 and then $10,000. 1751 01:45:17,079 --> 01:45:21,759 And that $10,000 has been like a critical limit for a long time. 1752 01:45:22,519 --> 01:45:26,079 And that's worth less and less, you know, as decades go on. 1753 01:45:26,439 --> 01:45:30,399 Yeah, you know when they started that back in the 1970s or something like that? 1754 01:45:31,139 --> 01:45:32,959 Maybe 60s, I don't remember exactly. 1755 01:45:33,099 --> 01:45:34,839 But you know why $10,000 was chosen? 1756 01:45:35,579 --> 01:45:36,059 Why? 1757 01:45:36,059 --> 01:45:37,519 Because it was the median income. 1758 01:45:38,539 --> 01:45:39,959 oh wow 1759 01:45:39,959 --> 01:45:41,619 it was the median income

1760 01:45:41,619 --> 01:45:43,199 now you can't buy a used car 1761 01:45:43,199 --> 01:45:43,879 that's so crazy 1762 01:45:43,879 --> 01:45:47,459 yeah they got full surveillance 1763 01:45:47,459 --> 01:45:48,639 on every transaction 1764 01:45:48,639 --> 01:45:50,879 just because they inflated away their standard 1765 01:45:50,879 --> 01:45:52,639 it's a pretty smart move though 1766 01:45:52,639 --> 01:45:53,999 gotta give that to them 1767 01:45:53,999 --> 01:45:54,959 oh it's genius 1768 01:45:54,959 --> 01:45:57,059 like you play the long game right 1769 01:45:57,059 --> 01:45:58,299 50 years later 1770 01:45:58,299 --> 01:46:00,439 you know 10 years from now 1771 01:46:00,439 --> 01:46:02,179 your KYC 1772 01:46:02,179 --> 01:46:05,079 double check government report 1773 01:46:05,079 --> 01:46:06,479 automated system 1774 01:46:06,479 --> 01:46:08,339 probably watched over and reported 1775 01:46:08,339 --> 01:46:10,679 to every financial agency by AI 1776 01:46:10,679 --> 01:46:12,259 every time you go to the grocery store. 1777 01:46:12,719 --> 01:46:13,139 10K. 1778 01:46:14,319 --> 01:46:15,099 Guarantee it. 1779 01:46:15,399 --> 01:46:16,619 2035, 2040. 1780 01:46:17,299 --> 01:46:18,359 Like you can be thousands of dollars 1781 01:46:18,359 --> 01:46:19,519 every time you go to the grocery store. 1782 01:46:20,099 --> 01:46:21,499 Government debt's going to be, 1783 01:46:21,859 --> 01:46:22,919 what do we got? 1784 01:46:24,159 --> 01:46:26,259 So 2025, 2035. 1785 01:46:27,499 --> 01:46:28,479 So 80. 1786 01:46:29,459 --> 01:46:34,819 So 2045 will be around $200 trillion debt. 1787 01:46:35,139 --> 01:46:36,479 It's Canadian money, 1788 01:46:36,479 --> 01:46:41,819 but I'm spending 300, 350 bucks twice a week sometimes. 1789 01:46:42,019 --> 01:46:44,379 I mean, like that's just, it's crazy. 1790 01:46:45,519 --> 01:46:45,779 Oh yeah. 1791 01:46:46,039 --> 01:46:47,059 Well, dude, we buy, 1792 01:46:47,439 --> 01:46:51,919 we try to buy stuff that doesn't have pesticides all over it 1793 01:46:51,919 --> 01:46:53,139 and we buy real food. 1794 01:46:53,919 --> 01:46:56,159 Our grocery bills are insane, man. 1795 01:46:56,999 --> 01:46:57,279 Yeah. 1796 01:46:57,479 --> 01:46:59,099 Like I'm about to buy chickens. 1797 01:46:59,239 --> 01:47:00,559 Our grocery bills are so bad. 1798 01:47:03,759 --> 01:47:05,819 A pet project that I have, 1799 01:47:05,819 --> 01:47:12,279 i think i've told you about it a few times but not max for sure is um bitcoin has a built-in 1800 01:47:12,279 --> 01:47:21,099 completely self-referential um measure of its purchasing power yeah you takes oracle i arrived 1801 01:47:21,099 --> 01:47:28,419 through utx oracle yeah so it's like like utx oracle does the round usd amounts whatever but 1802 01:47:28,419 --> 01:47:35,279 if you just don't think about usd price just think about purchasing power of bitcoin usd price is like 1803 01:47:35,279 --> 01:47:42,799 a rough estimate of some magical thing called purchasing power, right? You can get a very good 1804 01:47:42,799 --> 01:47:50,459 purchasing power of Bitcoin just looking at the fractionization of UTXO sizes. 1805 01:47:51,839 --> 01:48:00,779 If you just look at the median UTXO size used to be one Bitcoin. Now the median UTXO size 1806 01:48:00,779 --> 01:48:11,959 is 0.001 bitcoin and it has a 95 correlation with the usd price just looking at um median utxo 1807 01:48:11,959 --> 01:48:19,219 size and um you can just throw out usd price and like i kind of want to just start publishing the 1808 01:48:19,219 --> 01:48:25,379 bitcoin purchasing power just based on just based on the median yeah like just think about bars of 1809 01:48:25,379 --> 01:48:32,279 gold right like if you if you somehow had a database of the sizes of all the bars of gold 1810 01:48:32,279 --> 01:48:38,539 in the world and you just saw this over thousands of years and you saw the fractionization like hey 1811 01:48:38,539 --> 01:48:43,459 what's the median size of a gold bar is it this big or is it this big or is it this big because 1812 01:48:43,459 --> 01:48:48,839 people make smaller fractionizations when it becomes more valuable yeah and it's just like 1813 01:48:48,839 --> 01:48:56,879 point of it is to segment out value yeah right so it's not perfect but it's remarkably simple 1814 01:48:56,879 --> 01:49:02,099 and self-referential and you don't have to reference any outside data it's all on chain 1815 01:49:02,099 --> 01:49:08,339 purchasing power this man's i'm really excited about that our resident blockchain archaeologist 1816 01:49:08,339 --> 01:49:14,379 speaks up well that's like some some breakthrough economic data like that's freaking awesome i know 1817 01:49:14,379 --> 01:49:15,159 this is wild. 1818 01:49:15,759 --> 01:49:15,939 Yeah. 1819 01:49:16,179 --> 01:49:17,279 I've had the earlier, 1820 01:49:17,419 --> 01:49:27,179 this idea I had this I had this idea I was like dude this could be awesome We don even need Fiat to tell us purchasing power We don need anything to tell us purchasing power All we need is UTXO sizes 1821 01:49:27,799 --> 01:49:27,819 Yeah. 1822 01:49:28,639 --> 01:49:30,379 So GDP's always been shit. 1823 01:49:30,379 --> 01:49:32,779 Even coin joins would work on that, though. 1824 01:49:33,439 --> 01:49:33,959 I know. 1825 01:49:33,959 --> 01:49:38,179 Even coin joins, because you have to have a coin join threshold, 1826 01:49:38,439 --> 01:49:43,179 like UTXO size threshold that's relevant to your users' amount of capital 1827 01:49:43,179 --> 01:49:46,959 to actually make it do its job well. 1828 01:49:47,479 --> 01:49:54,879 yeah and the real the real measure of economic growth is deflation yeah so so you're you throw 1829 01:49:54,879 --> 01:50:00,599 out gdp this is your new gdp this is the better gdp exactly yeah it's it's this is economic growth 1830 01:50:00,599 --> 01:50:07,979 the only nuance is uh timescale so like i'm fairly confident that somewhere between

like 1831 01:50:07,979 --> 01:50:12,819 you don't want to do it on a week yeah because like you have noise and you have an attack but 1832 01:50:12,819 --> 01:50:19,999 if you do it even on a monthly time scale you have so many hundreds of billions data data points 1833 01:50:19,999 --> 01:50:25,919 and median is extremely robust to anyone trying to attack it or anything like that 1834 01:50:25,919 --> 01:50:33,219 so like absolutely rock solid a year for the purchasing power um but i i think a monthly 1835 01:50:33,219 --> 01:50:38,479 rolling average uh monthly median would give you almost perfect purchasing power 1836 01:50:38,479 --> 01:50:48,819 You should standardize it and do a little schema and call it the BPP standard or whatnot. 1837 01:50:49,419 --> 01:50:52,759 And just have it hosted on a website. 1838 01:50:53,099 --> 01:50:55,899 And every time we talk about it, we'll report what the new BPP is. 1839 01:50:56,939 --> 01:50:58,939 Oracle finally kills USD price. 1840 01:51:00,179 --> 01:51:02,959 I told you the whole time I didn't care about the USD price. 1841 01:51:03,199 --> 01:51:03,839 The whole time. 1842 01:51:05,579 --> 01:51:07,939 Because with Lightning, the payment channel size. 1843 01:51:08,479 --> 01:51:09,859 would be affected. 1844 01:51:09,899 --> 01:51:10,639 Would do the same thing. 1845 01:51:11,319 --> 01:51:11,419 Yeah. 1846 01:51:11,659 --> 01:51:11,839 Yeah. 1847 01:51:12,919 --> 01:51:13,259 Yeah. 1848 01:51:14,359 --> 01:51:16,879 Nah, it's a pretty clever ass. 1849 01:51:16,959 --> 01:51:18,799 I bet you can see that just in your data, Max, 1850 01:51:18,899 --> 01:51:19,979 with Wasabi. 1851 01:51:20,239 --> 01:51:21,539 Like, just look at the sizes 1852 01:51:21,539 --> 01:51:23,159 people have been doing over the years. 1853 01:51:24,439 --> 01:51:24,839 Yeah. 1854 01:51:25,099 --> 01:51:26,839 I mean, we definitely noticed this 1855 01:51:26,839 --> 01:51:27,779 because when we were starting, 1856 01:51:27,939 --> 01:51:29,179 you know, we chose to 0.1 Bitcoin. 1857 01:51:29,179 --> 01:51:29,699 Max is about to lose 5 hours of live coding. 1858 01:51:31,959 --> 01:51:34,839 Like, we had 0.1 Bitcoin default denomination. 1859 01:51:35,299 --> 01:51:37,959 And back then, that was like 100 bucks or something. 1860 01:51:38,479 --> 01:51:41,059 And when we finally switched, and that was the minimum, 1861 01:51:41,239 --> 01:51:43,199 you could not coin join if you had less than that. 1862 01:51:44,039 --> 01:51:46,539 And when we finally switched to 2.0, 1863 01:51:46,939 --> 01:51:48,919 that was like, I don't know, 10,000 bucks or something. 1864 01:51:48,919 --> 01:51:53,759 So poor guys couldn't coin join back then, which sucked. 1865 01:51:54,659 --> 01:51:57,239 And yeah, that's fixed now. 1866 01:51:57,699 --> 01:52:01,219 And now it's kind of, you know, we choose outputs amount randomly, 1867 01:52:01,679 --> 01:52:06,799 which probably messes up your metric, maybe. 1868 01:52:06,799 --> 01:52:14,279 but again like with the median not really and like really the median yeah yeah and and and just 1869 01:52:14,279 --> 01:52:18,379 large numbers you know if you come in with truly random it doesn't actually affect the median 1870 01:52:18,379 --> 01:52:24,299 measurement yeah yeah because it just comes in as extremes either way because there's always going 1871 01:52:24,299 --> 01:52:31,319 to be a bulk in the middle you know yeah and this idea is based on the most simple thing ever 1872 01:52:31,319 --> 01:52:37,919 The median transaction that people make, just median in general, is $100. 1873 01:52:39,099 --> 01:52:40,199 Always has been. 1874 01:52:40,739 --> 01:52:41,699 Always will be. 1875 01:52:42,219 --> 01:52:45,119 And that's just going to be a different amount of Bitcoin. 1876 01:52:45,319 --> 01:52:47,619 I disagree with the always will be. 1877 01:52:48,039 --> 01:52:50,799 The $100 is worth anything very soon. 1878 01:52:51,839 --> 01:52:52,939 Yeah, exactly. 1879 01:52:53,159 --> 01:52:53,919 But you know what I mean? 1880 01:52:53,979 --> 01:52:57,719 Well, let's say like a full tank of gas on a… 1881 01:52:57,719 --> 01:52:59,059 Steve is a Fiat Maxi. 1882 01:52:59,659 --> 01:53:00,499 I can't believe it. 1883 01:53:00,499 --> 01:53:05,119 Full tank of gas on a big car right now is like $100 maybe or something like that, right? 1884 01:53:05,219 --> 01:53:14,259 So full tank of gas on a big car, that value right there, that's going to be your average transaction purchasing power forever because people live large. 1885 01:53:14,259 --> 01:53:15,539 This guy doesn't even have a Tesla. 1886 01:53:16,759 --> 01:53:18,299 People go out to eat. 1887 01:53:19,519 --> 01:53:29,039 People are going to make transactions once a week or so on daily basis for this amount of purchasing power because people's basic needs don't change much. 1888 01:53:29,039 --> 01:53:34,979 they need gas they need clothes they need food you know like it's just gonna the average 1889 01:53:34,979 --> 01:53:40,839 transaction size is gonna be something about that their basic needs if you average together 1890 01:53:40,839 --> 01:53:48,239 uh jeffrey's and my grocery bills you're right there's the medium there it is there you go 1891 01:53:48,239 --> 01:53:55,239 oh man yeah it's fascinating i gotta jump off here guys yeah dude we're already me too i'm 1892 01:53:55,239 --> 01:53:55,599 starving. 1893 01:53:57,319 --> 01:53:59,179 What time is it? Yeah, let's go. 1894 01:53:59,339 --> 01:53:59,699 Let's close. 1895 01:54:01,019 --> 01:54:03,179 Well, dude, this was awesome. Max, 1896 01:54:03,219 --> 01:54:05,139 thanks for coming on. This was 1897 01:54:05,139 --> 01:54:07,179 a great roundtable. Hell yeah. 1898 01:54:09,039 --> 01:54:09,339 He's gonna 1899 01:54:09,339 --> 01:54:11,099 go off and talk about... Maybe like feature 1900 01:54:11,099 --> 01:54:12,619 requests of like, 1901 01:54:13,159 --> 01:54:15,319 cut the videos and let's just take

a walk in the park. 1902 01:54:15,799 --> 01:54:16,839 I think it would be fun. 1903 01:54:17,919 --> 01:54:19,059 We need more like 1904 01:54:19,059 --> 01:54:21,319 audio only walk and talks. 1905 01:54:22,199 --> 01:54:23,179 I don't care with that. 1906 01:54:23,179 --> 01:54:26,419 we could do an episode we could do an episode where everybody's walking in a 1907 01:54:26,419 --> 01:54:39,738 separate park everybody has to go to their favorite park and then we all walking through the park a group a group drugs or no drugs I guess you got to define drugs BYOB 1908 01:54:40,738 --> 01:54:42,578 You just make your own decisions. 1909 01:54:42,818 --> 01:54:43,878 Make your own decision there. 1910 01:54:45,478 --> 01:54:47,518 You can use my gonorrhea coins 1911 01:54:47,518 --> 01:54:49,898 to buy some if you need it. 1912 01:54:50,318 --> 01:54:52,398 I might be swimming in a lake. 1913 01:54:52,458 --> 01:54:52,898 Is that cool? 1914 01:54:53,698 --> 01:54:54,178 That's okay. 1915 01:54:54,178 --> 01:54:55,098 If you had fun survive. 1916 01:54:55,318 --> 01:54:56,518 I mean, it's your phone. 1917 01:54:59,978 --> 01:55:09,618 oh jesus well dude uh oh yeah sorry wait but but just before my last words but simply 1918 01:55:09,618 --> 01:55:15,058 steve you were saying about um gg right because like gg has a dope new podcast i'm not sure if 1919 01:55:15,058 --> 01:55:22,458 you guys know oh no i don't what's it called no solutions only trade-offs no solutions yep what 1920 01:55:22,458 --> 01:55:28,498 that's a great podcast name it's amazing that's a great it's called no solution what an asshole 1921 01:55:28,498 --> 01:55:31,838 Like he creates a podcast and no solutions. 1922 01:55:32,038 --> 01:55:32,198 What? 1923 01:55:32,518 --> 01:55:33,158 Only trade-offs. 1924 01:55:33,358 --> 01:55:34,078 This is the Samusel. 1925 01:55:34,078 --> 01:55:34,858 Only trade-offs. 1926 01:55:35,018 --> 01:55:35,258 Okay. 1927 01:55:35,758 --> 01:55:36,898 See, just the name. 1928 01:55:37,018 --> 01:55:37,758 That's so GG. 1929 01:55:37,838 --> 01:55:39,038 I hate you, GG. 1930 01:55:39,398 --> 01:55:39,918 I just. 1931 01:55:40,298 --> 01:55:42,718 And this is, by the way, a walk and talk podcast. 1932 01:55:42,918 --> 01:55:43,938 Like no studio. 1933 01:55:44,098 --> 01:55:44,898 It's in person only. 1934 01:55:45,358 --> 01:55:47,578 And it tells the story of sovereign engineering. 1935 01:55:48,038 --> 01:55:49,498 So very much recommended. 1936 01:55:50,298 --> 01:55:51,058 One of the best ones. 1937 01:55:51,798 --> 01:55:53,618 The worst production quality, by the way. 1938 01:55:53,738 --> 01:55:54,478 Zero edits. 1939 01:55:56,018 --> 01:55:57,358 Lawnmowers in the background. 1940 01:55:57,358 --> 01:55:58,638 it's great 1941 01:55:58,638 --> 01:56:01,198 where is this thing 1942 01:56:01,198 --> 01:56:02,138 I can't search it 1943 01:56:02,138 --> 01:56:03,538 this is annoying 1944 01:56:03,538 --> 01:56:06,378 because it's like a Noster podcast 1945 01:56:06,378 --> 01:56:08,538 it's a white noise podcast 1946 01:56:08,538 --> 01:56:10,218 you can only get it through white noise 1947 01:56:10,218 --> 01:56:13,298 alright I'll find it 1948 01:56:13,298 --> 01:56:15,198 I'll put it on the show notes too 1949 01:56:15,198 --> 01:56:16,118 alright 1950 01:56:16,118 --> 01:56:18,558 I gotta go 1951 01:56:18,558 --> 01:56:20,158 see you guys 1952 01:56:20,158 --> 01:56:21,738 later Steve man thanks for joining 1953 01:56:21,738 --> 01:56:24,618 anything y'all want to point crew to 1954 01:56:24,618 --> 01:56:26,418 you want to plug 1955 01:56:26,418 --> 01:56:28,858 plug something. I mean other than white noise. 1956 01:56:30,038 --> 01:56:30,558 Yeah like 1957 01:56:30,558 --> 01:56:32,938 I mean white noise the app is out now 1958 01:56:32,938 --> 01:56:34,938 in TestFlight and ZapStore. Hopefully soon mainnet 1959 01:56:34,938 --> 01:56:37,098 but what's more interesting is 1960 01:56:37,098 --> 01:56:38,678 the Marmot protocol like 1961 01:56:38,678 --> 01:56:41,158 if you want chat in your 1962 01:56:41,158 --> 01:56:43,098 apps just like it's 1963 01:56:43,098 --> 01:56:44,858 a solved problem I hope now so 1964 01:56:44,858 --> 01:56:45,858 just use it. 1965 01:56:46,798 --> 01:56:49,058 There's four libraries and they're written 1966 01:56:49,058 --> 01:56:50,678 in Rust, TypeScript, 1967 01:56:51,138 --> 01:56:53,158 Kotlin, C Sharp and a whole bunch 1968 01:56:53,158 --> 01:56:54,058 of language bindings. 1969 01:56:54,058 --> 01:56:57,758 so it'll probably work for whatever you want 1970 01:56:57,758 --> 01:57:01,438 and non-chap use cases as well 1971 01:57:01,438 --> 01:57:03,738 so I guess there's that 1972 01:57:03,738 --> 01:57:04,878 I have a question on that 1973 01:57:04,878 --> 01:57:06,158 how do you think 1974 01:57:06,158 --> 01:57:07,718 and I know you haven't seen Paracore or anything 1975 01:57:07,718 --> 01:57:11,038 how do you think this would integrate 1976 01:57:11,038 --> 01:57:15,778 with the idea of our spaces 1977 01:57:15,778 --> 01:57:19,018 and just a Hyper-B index 1978 01:57:19,018 --> 01:57:21,398 or a Hyper-B status 1979 01:57:21,398 --> 01:57:23,238 on what is the latest index 1980 01:57:23,238 --> 01:57:26,158 but integrating it for encryption. 1981 01:57:26,378 --> 01:57:28,658 Right now we have encryption in private spaces 1982 01:57:28,658 --> 01:57:30,198 and then we also have public spaces 1983 01:57:30,198 --> 01:57:31,558 which are just kind of like BitTorrent, 1984 01:57:31,918 --> 01:57:33,398 like host a website, you know, 1985 01:57:33,418 --> 01:57:35,258 anybody can just join as long as they have the key. 1986 01:57:36,058 --> 01:57:37,478 But we have private spaces 1987 01:57:37,478 --> 01:57:40,138 and right now we've just kind of done the simple thing, right? 1988 01:57:40,218 --> 01:57:41,538 Is that we just have a key 1989 01:57:41,538 --> 01:57:46,538 and you do a safe share of the encryption key 1990 01:57:46,538 -->

01:57:48,518 and you just update it 1991 01:57:48,518 --> 01:57:51,818 like it's a file in the index, right? 1992 01:57:51,818 --> 01:57:53,298 everybody gets the new version of the index. 1993 01:57:53,758 --> 01:57:54,938 But when somebody leaves 1994 01:57:54,938 --> 01:57:57,278 or you have to kick somebody out or something. 1995 01:57:58,438 --> 01:58:00,918 And I really haven't sorted out 1996 01:58:00,918 --> 01:58:02,078 the kind of like magic sauce 1997 01:58:02,078 --> 01:58:05,958 of blind hosting 1998 01:58:05,958 --> 01:58:12,558 in a way that feels like not a risk, 1999 01:58:12,798 --> 01:58:13,518 you know, 2000 01:58:14,098 --> 01:58:17,878 where somebody can have the full encrypted backup 2001 01:58:17,878 --> 01:58:18,658 of all this stuff 2002 01:58:18,658 --> 01:58:19,758 and everybody can still request 2003 01:58:19,758 --> 01:58:20,618 without just kind of leaking 2004 01:58:20,618 --> 01:58:21,998 all the information in the database. 2005 01:58:23,618 --> 01:58:27,698 Obviously, they wouldn't leak the file data itself, 2006 01:58:27,918 --> 01:58:30,838 but if you can manage to get the index 2007 01:58:30,838 --> 01:58:33,898 and know that you're serving matrix.mp4, 2008 01:58:34,218 --> 01:58:35,398 they're going to be able to assume 2009 01:58:35,398 --> 01:58:38,158 that you're asking for the Matrix movie. 2010 01:58:39,218 --> 01:58:41,598 And there's a bunch of little nuances there, 2011 01:58:41,758 --> 01:58:45,118 but I wondered and haven't really dug into it 2012 01:58:45,118 --> 01:58:47,518 how Marmot might fit into something like that 2013 01:58:47,518 --> 01:58:49,558 to kind of have... 2014 01:58:49,558 --> 01:58:54,018 I mean, maybe there's various solutions, but I'm just curious if you had a thought. 2015 01:58:55,538 --> 01:58:57,738 Yeah, we definitely got to talk longer about this. 2016 01:58:57,738 --> 01:59:09,638 But ultimately, with Marmot, the idea is that you have strong forward secrecy, meaning that if you invite someone to a group, he does not get the old content. 2017 01:59:09,638 --> 01:59:12,138 and for 2018 01:59:12,138 --> 01:59:14,038 I'm not quite sure but for 2019 01:59:14,038 --> 01:59:16,318 your use case that might actually 2020 01:59:16,318 --> 01:59:18,298 be like if you invite someone to 2021 01:59:18,298 --> 01:59:20,218 a space you want him to get everything. You want them to see the 2022 01:59:20,218 --> 01:59:21,339 whole index. Yeah. 2023 01:59:22,258 --> 01:59:24,298 No that makes sense. It probably 2024 01:59:24,298 --> 01:59:25,738 just doesn't really fit the use case. 2025 01:59:27,018 --> 01:59:28,278 But on the other hand 2026 01:59:28,278 --> 01:59:30,418 like this whole key rotation 2027 01:59:30,418 --> 01:59:32,378 thing is there are 2028 01:59:32,378 --> 01:59:34,298 so many gremlins in the details 2029 01:59:34,298 --> 01:59:35,618 and 2030 01:59:35,618 --> 01:59:38,258 hand rolling your own crypto 2031 01:59:38,258 --> 01:59:40,378 is always just a pain in the ass.

2032 01:59:40,838 --> 01:59:42,098 And especially with 2033 01:59:42,098 --> 01:59:56,518 MLS it complicated And there so many edge cases And so finding a good key a continuous key group key agreement protocol that you don 2034 01:59:56,518 --> 02:00:00,658 have to worry about saves you a whole bunch of headache. 2035 02:00:00,878 --> 02:00:06,438 I think that that would be my main kind of reason on, on trying to look for something 2036 02:00:06,438 --> 02:00:09,558 like MLS or Marmot that might not be the perfect thing, but. 2037 02:00:09,558 --> 02:00:12,078 But just does it and I don't have to fight it. 2038 02:00:12,078 --> 02:00:16,378 Because you can do a prototype and a proof of concept and it'll work great. 2039 02:00:16,918 --> 02:00:19,138 And we did that with White Noise two years ago. 2040 02:00:19,238 --> 02:00:19,938 We had a prototype. 2041 02:00:20,098 --> 02:00:21,858 It worked great when it worked. 2042 02:00:22,378 --> 02:00:25,078 But then you ship it to 10 people and it breaks. 2043 02:00:26,438 --> 02:00:28,098 And I'm like... 2044 02:00:28,098 --> 02:00:28,918 I don't know what that's like at all. 2045 02:00:28,918 --> 02:00:29,918 You're smiling because... 2046 02:00:29,918 --> 02:00:31,158 Exactly. 2047 02:00:32,938 --> 02:00:34,818 And this is where it's so nasty. 2048 02:00:34,818 --> 02:00:37,918 It works in the happy path. 2049 02:00:37,918 --> 02:00:40,758 but if there's just a little bit of chaos 2050 02:00:40,758 --> 02:00:42,458 and there will be a lot of chaos 2051 02:00:42,458 --> 02:00:43,378 you're fucked 2052 02:00:43,378 --> 02:00:46,458 and that's kind of where I'm 2053 02:00:46,458 --> 02:00:48,838 very glad that we are just using MLS 2054 02:00:48,838 --> 02:00:50,358 because again 2055 02:00:50,358 --> 02:00:52,218 170 pages of specification 2056 02:00:52,218 --> 02:00:53,698 to deal with the chaos 2057 02:00:53,698 --> 02:00:56,658 and then now our much more 2058 02:00:56,658 --> 02:00:57,998 sophisticated Marmot V2 2059 02:00:57,998 --> 02:01:00,378 gets rid of a whole bunch of other chaos 2060 02:01:00,378 --> 02:01:01,718 if you want to do it decentralized 2061 02:01:01,718 --> 02:01:04,638 but ultimately 2062 02:01:04,638 --> 02:01:06,898 for a lot of use cases 2063 02:01:06,898 --> 02:01:07,498 it'll work 2064 02:01:07,498 --> 02:01:12,078 for the thing of like you add someone to a group and he gets access to everything 2065 02:01:12,078 --> 02:01:15,638 that's a little bit tricky but you can build that on top of 2066 02:01:15,638 --> 02:01:19,518 the existing system it's kind of like 2067 02:01:19,518 --> 02:01:23,818 you add a second device you know you have your signal app on the phone 2068 02:01:23,818 --> 02:01:27,658 and then you want to add your laptop and it connects 2069 02:01:27,658 --> 02:01:31,878 you know and you get your identity over there quickly and then it takes 2070 02:01:31,878 --> 02:01:33,958 a couple hours until everything gets backfilled 2071 02:01:33,958 --> 02:01:42,998 this for example is a way of doing it not as elegant as just share the key and instantly you 2072 02:01:42,998 --> 02:01:52,798 have everything but yeah depends no solutions only trade-offs the use case that i saw as being 2073 02:01:52,798 --> 02:02:01,138 potentially valuable but i don't know how it works with like the layers of encryption and not having 2074 02:02:01,138 --> 02:02:03,158 to re-encrypt files over and over again 2075 02:02:03,158 --> 02:02:05,078 and trying to figure out how to 2076 02:02:05,078 --> 02:02:07,098 toe that line, so to speak, 2077 02:02:08,298 --> 02:02:09,298 and probably just doing 2078 02:02:09,298 --> 02:02:11,258 it gated through the index so that 2079 02:02:11,258 --> 02:02:13,358 you're only really just... It's the map 2080 02:02:13,358 --> 02:02:14,978 that you're encrypting 2081 02:02:14,978 --> 02:02:17,318 was to make it so 2082 02:02:17,318 --> 02:02:19,238 that you can invite somebody to the space and only have 2083 02:02:19,238 --> 02:02:21,258 access to one file or only have access 2084 02:02:21,258 --> 02:02:22,078 to one folder, 2085 02:02:22,878 --> 02:02:25,178 but they could still be part of the entire space. 2086 02:02:26,438 --> 02:02:27,018 So they're 2087 02:02:27,018 --> 02:02:29,198 still... You don't have to create 2088 02:02:29,198 --> 02:02:31,118 a subspace and then suddenly they're the only user 2089 02:02:31,118 --> 02:02:38,038 you know in the space so like you can still go and get cedars from the space with a hundred 2090 02:02:38,038 --> 02:02:43,098 thousand people in it you're only allowed to request or you only have you can only visibly 2091 02:02:43,098 --> 02:02:48,958 see these two files that you've purchased or gotten or offered access to or whatever it is 2092 02:02:49,638 --> 02:02:56,438 from the from the group so i mean there's there's a million things to work out and a billion things 2093 02:02:56,438 --> 02:02:57,338 to be able to set up. 2094 02:02:57,618 --> 02:02:57,818 But, 2095 02:02:58,478 --> 02:02:58,658 wait, 2096 02:02:58,778 --> 02:02:59,638 we should just stay in touch 2097 02:02:59,638 --> 02:03:00,378 and I'll keep 2098 02:03:00,378 --> 02:03:01,758 giving feedback on white noise. 2099 02:03:02,138 --> 02:03:02,838 I was just shocked 2100 02:03:02,838 --> 02:03:03,638 that it just kind of worked. 2101 02:03:04,298 --> 02:03:05,198 I didn't have any problems with it. 2102 02:03:05,218 --> 02:03:05,518 Somebody said 2103 02:03:05,518 --> 02:03:06,078 they couldn't figure out 2104 02:03:06,078 --> 02:03:06,898 how to set up their relays 2105 02:03:06,898 --> 02:03:07,398 and I was like, 2106 02:03:07,478 --> 02:03:07,738 I didn't, 2107 02:03:08,418 --> 02:03:09,298 I didn't set up relays. 2108 02:03:09,298 --> 02:03:09,578 Yeah. 2109 02:03:09,838 --> 02:03:10,538 I don't understand. 2110 02:03:12,678 --> 02:03:13,798 That's another tangent, 2111 02:03:14,298 --> 02:03:14,538 but 2112 02:03:14,538 --> 02:03:16,838 Nostra sucks. 2113 02:03:16,838 -->

02:03:17,458 I got 45 minutes. 2114 02:03:17,658 --> 02:03:18,478 Let's just go ahead and 2115 02:03:18,478 --> 02:03:20,738 Like, 2116 02:03:21,098 --> 02:03:21,578 you know, 2117 02:03:21,658 --> 02:03:21,978 in Nostra, 2118 02:03:22,078 --> 02:03:24,458 there's thousands of relays 2119 02:03:24,458 --> 02:03:25,998 and so where do you put 2120 02:03:25,998 --> 02:03:29,758 your data and the solution to that is you create an event 2121 02:03:29,758 --> 02:03:34,138 that is here are the relays that I use that I put my data on. 2122 02:03:34,878 --> 02:03:38,158 And that's cool and all but then what if that relay 2123 02:03:38,158 --> 02:03:41,678 turns off and now you have a broken relay in your 2124 02:03:41,678 --> 02:03:44,338 outbox model list in your relay list. 2125 02:03:45,798 --> 02:03:49,078 And when we shipped white noise for the first time 2126 02:03:49,078 --> 02:03:53,078 that was one of the main key frustrations because we went 2127 02:03:53,078 --> 02:03:56,078 radical decentralization. 2128 02:03:56,458 --> 02:03:57,558 It works on any relay. 2129 02:03:57,838 --> 02:03:59,938 We'll just use the relays that you already 2130 02:03:59,938 --> 02:04:01,938 have configured and it'll just be great. 2131 02:04:02,618 --> 02:04:04,258 But then it broke all the time. 2132 02:04:05,058 --> 02:04:05,198 And 2133 02:04:05,198 --> 02:04:07,458 it's hard. 2134 02:04:08,198 --> 02:04:09,898 So for now, we're like, 2135 02:04:09,978 --> 02:04:11,958 actually, let's get it working in an ideal 2136 02:04:11,958 --> 02:04:13,718 condition, kind of baby steps. 2137 02:04:14,658 --> 02:04:16,078 And so we have some relays 2138 02:04:16,078 --> 02:04:17,218 that are well configured 2139 02:04:17,218 --> 02:04:20,078 and also do this 2140 02:04:20,078 --> 02:04:21,938 neck entropy syncing in between each other 2141 02:04:21,938 --> 02:04:26,258 to make sure that at least the relay side will be solved. 2142 02:04:27,078 --> 02:04:31,278 And we haven't yet made it easy to configure your own custom relays 2143 02:04:31,278 --> 02:04:32,778 for a specific group. 2144 02:04:34,058 --> 02:04:34,098 Okay. 2145 02:04:34,618 --> 02:04:36,818 But what we do already do is... 2146 02:04:36,818 --> 02:04:37,518 No, that would be cool though. 2147 02:04:37,798 --> 02:04:39,458 I mean, UX wise is a pain. 2148 02:04:39,458 --> 02:04:40,418 It's definitely the goal. 2149 02:04:40,638 --> 02:04:41,238 Yeah, it's cool. 2150 02:04:41,558 --> 02:04:43,918 Yeah, UX and reliability is a pain, 2151 02:04:43,998 --> 02:04:45,998 but definitely that's where we want to be. 2152 02:04:46,478 --> 02:04:47,618 It's just how do we get there. 2153 02:04:48,338 --> 02:04:51,478 But the thing where we do make it proper, so to say, 2154 02:04:51,478 --> 02:04:55,478 is with sending you invites. In order to invite you to a group, 2155 02:04:55,478 --> 02:04:59,478 I will put the invite event to a relay that is in your 2156 02:04:59,478 --> 02:05:00,478 list. 2157 02:04:59,990 --> 02:05:09,110 And lo and behold, that's the main bug in the initial release is that people miss their invites to groups because the relays are dead. 2158 02:05:09,110 --> 02:05:09,630 Because it's on a dead relay. 2159 02:05:11,430 --> 02:05:16,890 And we already have fixes for that, thankfully, and that's hopefully going to be released on Monday or something. 2160 02:05:17,050 --> 02:05:17,930 So it should be soon. 2161 02:05:18,470 --> 02:05:27,150 Well, if I don't really suck or I'm an idiot and I give up or whatever, I'm just going to keep going with everything that we're doing with Paradrive. 2162 02:05:27,150 --> 02:05:33,010 and I really hope and wanted to make a Noster client, 2163 02:05:34,150 --> 02:05:37,110 or not even necessarily make a Noster client, 2164 02:05:37,310 --> 02:05:40,190 but I want to use, I want to be able to comment on files 2165 02:05:40,190 --> 02:05:43,410 and I want to be able to heart and like stuff 2166 02:05:43,410 --> 02:05:46,870 with my friends and family and collections that we're in. 2167 02:05:47,370 --> 02:05:51,590 And it seems absolutely stupid to not just use signed messages 2168 02:05:51,590 --> 02:05:52,350 because that's what we're doing. 2169 02:05:52,710 --> 02:05:55,010 Everything is signed messages in our whole system. 2170 02:05:55,010 --> 02:06:01,110 um and so it's like well noster is the joint key right i don't want to bore i don't want to 2171 02:06:01,110 --> 02:06:07,290 wonder if this is mom's iphone or mom's ipad like you know it's just it's my mother that's 2172 02:06:07,290 --> 02:06:14,630 making this comment um and uh so what i want to do though is to essentially be able to make a noster 2173 02:06:14,630 --> 02:06:20,970 client and have the wss colon slash slash you know a bunch of those those relays and then at the 2174 02:06:20,970 --> 02:06:27,870 bottom, I put my pair drives, pair core, colon slash slash in my key. And I didn't put in another 2175 02:06:27,870 --> 02:06:33,690 space for my family and like all our close friends. And if none of those relays respond, 2176 02:06:34,250 --> 02:06:37,470 well, I'm just going to go look in the space and see what everybody's latest version of their 2177 02:06:37,470 --> 02:06:41,810 index is, which is just the index of their posts and comments and all that stuff. And I just 2178 02:06:41,810 --> 02:06:47,530 aggregate it from directly from my friends. So the whole idea is to try to add a peer-to-peer 2179 02:06:47,530 --> 02:06:53,050 layer alongside the relay and there's no reason not to have both i get just the overhead it's not 2180 02:06:53,050 --> 02:06:59,130 that much yeah and if both people are online you just go direct that's where again phips comes in 2181 02:06:59,130 --> 02:07:05,670 like if i'm following gg and gg is has a device his start nine or whatever his second and third 2182 02:07:05,670 --> 02:07:12,650 and fifth project computer is online with his no stir history why am i not using his key to find his 2183

02:07:12,650 --> 02:07:15,530 to find him on the DHT and connect directly to him 2184 02:07:15,530 --> 02:07:16,730 and just get the list. 2185 02:07:17,130 --> 02:07:18,010 And if he's not online, 2186 02:07:18,410 --> 02:07:20,390 I bet one of his 100,000 followers is, 2187 02:07:20,690 --> 02:07:21,810 and they've got the same key. 2188 02:07:22,670 --> 02:07:23,850 So I'll just ask him for them. 2189 02:07:24,050 --> 02:07:26,970 They'll respond on the DHT as easy as anybody else. 2190 02:07:28,690 --> 02:07:31,290 But I don't want to have to do something 2191 02:07:31,290 --> 02:07:32,350 that needs a network effect. 2192 02:07:32,610 --> 02:07:33,690 We're trying to focus on things 2193 02:07:33,690 --> 02:07:35,630 that are zero network effect first. 2194 02:07:36,810 --> 02:07:39,290 Get adoption before you need someone else with you. 2195 02:07:40,650 --> 02:07:41,590 This is, by the way, 2196 02:07:41,590 --> 02:07:44,570 is why I love talking to my agent over white noise 2197 02:07:44,570 --> 02:07:46,710 because I've been doing it for weeks 2198 02:07:46,710 --> 02:07:48,130 before we officially released. 2199 02:07:48,510 --> 02:07:50,990 Like I didn't need to have any one of my friends 2200 02:07:50,990 --> 02:07:51,890 to be on white noise 2201 02:07:51,890 --> 02:07:54,510 for me to have it so that white noise 2202 02:07:54,510 --> 02:07:56,550 has more screen time than Signe. 2203 02:07:57,150 --> 02:07:59,150 That's been like the case for me for a month or something. 2204 02:07:59,150 --> 02:07:59,690 That's awesome. 2205 02:08:00,210 --> 02:08:01,650 Dude, you got to get me, get me the stack. 2206 02:08:01,770 --> 02:08:02,570 Tell me how you're doing it. 2207 02:08:02,790 --> 02:08:03,770 And I just need to get, 2208 02:08:04,170 --> 02:08:05,310 I'm not going to get embarrassed again 2209 02:08:05,310 --> 02:08:07,650 that I mentioned Telegram on the show. 2210 02:08:08,210 --> 02:08:09,310 I want to, I want to do, 2211 02:08:09,430 --> 02:08:11,570 I want to talk to my agents through white noise 2212 02:08:11,570 --> 02:08:13,670 and just go ahead and just like set that shit up. 2213 02:08:13,770 --> 02:08:16,310 And then it will also kind of force me to take advantage 2214 02:08:16,310 --> 02:08:18,610 to be like, okay, Noster is the thing. 2215 02:08:18,710 --> 02:08:22,590 I just need to figure out how to make that my ID, you know? 2216 02:08:22,690 --> 02:08:24,410 Because that's what I want to do, you know? 2217 02:08:24,710 --> 02:08:27,350 You give it to Guy and then Guy will have to teach me how to do it. 2218 02:08:27,350 --> 02:08:29,690 And then he'll understand it better. 2219 02:08:29,990 --> 02:08:33,170 And then, you know, we can just all be off of Telegram. 2220 02:08:33,330 --> 02:08:34,330 We'll give it to Steve too. 2221 02:08:34,970 --> 02:08:35,550 That sounds great. 2222 02:08:35,670 --> 02:08:37,270 I mean, like the quick answer is... 2223 02:08:37,270 --> 02:08:38,470 Vibe coding between the four of us. 2224 02:08:38,710 --> 02:08:39,670 Man, this thing will be great. 2225 02:08:39,890 --> 02:08:40,330 No time. 2226 02:08:41,570 --> 02:08:59,775 If you have the Android app you can go to the settings and go to the AI settings And there literally like for you know open code cloud code codex Pi Hermes open claw I think those are all but we do more in the future But like for each 2227 02:08:59,775 --> 02:09:05,735 of them, there's a prompt. You just copy it and you paste it into your agent, whatever you're using. 2228 02:09:06,555 --> 02:09:10,655 And you will just do it. The number of times I've asked my Hermes agent, like, 2229 02:09:10,935 --> 02:09:14,135 can you connect to this thing? Can you find this thing? And it's just like, 2230 02:09:14,135 --> 02:09:15,415 I got to set up an agent. 2231 02:09:15,655 --> 02:09:16,895 You have to give me permissions. 2232 02:09:17,055 --> 02:09:18,055 Go back to your Mac mini 2233 02:09:18,055 --> 02:09:18,535 and sit. 2234 02:09:18,835 --> 02:09:19,235 Okay. 2235 02:09:19,615 --> 02:09:19,915 It's like, 2236 02:09:20,035 --> 02:09:21,435 Hey, 2237 02:09:21,555 --> 02:09:22,535 I messaged you on Telegram. 2238 02:09:22,615 --> 02:09:22,835 I was like, 2239 02:09:22,975 --> 02:09:23,235 shit, 2240 02:09:23,315 --> 02:09:23,455 man, 2241 02:09:23,515 --> 02:09:23,855 this thing, 2242 02:09:23,955 --> 02:09:24,575 this has control. 2243 02:09:26,495 --> 02:09:28,135 So we're so going to dive to AI. 2244 02:09:28,735 --> 02:09:29,895 So this is Android only, 2245 02:09:30,095 --> 02:09:30,775 not iPhone. 2246 02:09:30,855 --> 02:09:31,935 Cause iPhone is my main. 2247 02:09:32,035 --> 02:09:32,135 No, 2248 02:09:32,215 --> 02:09:32,915 it is. 2249 02:09:33,255 --> 02:09:34,235 It is also an iPhone. 2250 02:09:34,355 --> 02:09:35,235 It's just that on iPhone, 2251 02:09:35,355 --> 02:09:36,955 I don't think that the prompts are in yet. 2252 02:09:36,955 --> 02:09:37,755 Like the settings, 2253 02:09:37,975 --> 02:09:38,515 you know, 2254 02:09:38,835 --> 02:09:41,435 the iPhone people don't like cool stuff. 2255 02:09:41,695 --> 02:09:43,115 That's why they're like in prison. 2256 02:09:43,595 --> 02:09:43,975 Faggots. 2257 02:09:44,135 --> 02:09:46,095 all the cool stuff you only have in Freedomland. 2258 02:09:46,855 --> 02:09:47,975 I haven't... 2259 02:09:47,975 --> 02:09:50,295 By the way, 2260 02:09:50,315 --> 02:09:51,335 there's another fun fact, but 2261 02:09:51,335 --> 02:09:53,835 each operating system basically 2262 02:09:53,835 --> 02:09:55,255 is a separate developer. 2263 02:09:56,355 --> 02:09:57,635 Oh, AI agents, right there. 2264 02:09:58,575 --> 02:10:00,235 Oh, sweet. Oh, you have it on iOS. 2265 02:10:00,615 --> 02:10:01,635 There's prompts. Or is that Android? 2266 02:10:02,415 --> 02:10:03,255 Yeah, no, it's the software. 2267 02:10:03,255 --> 02:10:03,655 Yeah, there's prompts. 2268 02:10:04,315 --> 02:10:05,675 Yeah, it's right there. 2269 02:10:05,975 --> 02:10:08,435 This is the

installation prompt for connecting with Herm's installation. 2270 02:10:08,575 --> 02:10:09,935 White noise through Marmot. Read the latest 2271 02:10:09,935 --> 02:10:11,015 connector guide at blah, blah, blah. 2272 02:10:11,015 --> 02:10:13,575 Oh my god, I'm doing this right now. 2273 02:10:14,135 --> 02:10:14,735 I'm doing this right now. 2274 02:10:14,795 --> 02:10:15,075 Hold on a second. 2275 02:10:15,375 --> 02:10:15,735 Hold on a second. 2276 02:10:16,635 --> 02:10:17,035 I'm doing this. 2277 02:10:17,375 --> 02:10:18,215 Let's do it live! 2278 02:10:18,715 --> 02:10:18,975 All right. 2279 02:10:19,055 --> 02:10:19,275 Hey. 2280 02:10:19,635 --> 02:10:19,895 Okay. 2281 02:10:20,455 --> 02:10:21,755 For the last week or two, 2282 02:10:22,055 --> 02:10:24,335 I've actually been switching much more 2283 02:10:24,335 --> 02:10:26,795 to open code and codex and clawed code directly. 2284 02:10:27,435 --> 02:10:28,575 Because Hermes is just, 2285 02:10:28,635 --> 02:10:30,995 it just breaks in weird ways all the time. 2286 02:10:31,195 --> 02:10:32,715 I've not had that problem. 2287 02:10:32,855 --> 02:10:35,355 I had open claw problems like crazy. 2288 02:10:35,575 --> 02:10:36,595 But my Hermes has been fantastic. 2289 02:10:36,595 --> 02:10:37,255 I almost, 2290 02:10:37,615 --> 02:10:39,475 I have very little trouble with it. 2291 02:10:39,535 --> 02:10:40,115 I find that interesting. 2292 02:10:41,775 --> 02:10:42,315 But yeah, 2293 02:10:42,335 --> 02:10:44,115 so basically I just have a start nine running. 2294 02:10:44,135 --> 02:10:49,075 with Hermes and Codex and Claude and OpenCode and all the things. 2295 02:10:50,315 --> 02:10:53,175 And each of them has a different pub key. 2296 02:10:54,595 --> 02:10:59,135 And I have my pub key on white noise and they're authenticated. 2297 02:11:00,035 --> 02:11:02,015 And then you create a new group. 2298 02:11:02,575 --> 02:11:03,915 You call that group. 2299 02:11:04,235 --> 02:11:09,075 I use emojis like the wing emoji is the Hermes agent. 2300 02:11:09,695 --> 02:11:13,135 And then if it's the laptop next to it, then it's the developer chat, whatever. 2301 02:11:13,135 --> 02:11:17,275 and then you have your group chat 2302 02:11:17,275 --> 02:11:19,255 with that one agent and you 2303 02:11:19,255 --> 02:11:22,135 and that's basically it. 2304 02:11:22,295 --> 02:11:24,355 You just talk to it and it's a new session. 2305 02:11:25,135 --> 02:11:27,335 For Hermes, it has like the, you know, 2306 02:11:27,395 --> 02:11:28,855 you can keep talking in one chat 2307 02:11:28,855 --> 02:11:30,655 and it dynamically spins up new sessions. 2308 02:11:31,615 --> 02:11:36,255 For the other like Codex, Claude, OpenCode, 2309 02:11:36,795 --> 02:11:39,155 it's basically one group is a session 2310 02:11:39,155 --> 02:11:42,375 and you can do slash new to reset it. 2311 02:11:42,375 --> 02:11:49,915 but yeah basically then i have just i don't know 20 to 30 chat groups with my different agents that 2312 02:11:49,915 --> 02:11:54,875 all run on the same box that all have access to the same files that just have different you know 2313 02:11:54,875 --> 02:12:01,895 details of it um and yeah i i start chatting i turn off the notifications and i just tell the 2314 02:12:01,895 --> 02:12:08,835 agent to you know mention me at pubkey when he when he needs me and like then comes the ding 2315 02:12:08,835 --> 02:12:10,655 and I see like sometimes 2316 02:12:10,655 --> 02:12:13,055 99 unread messages of all the 2317 02:12:13,055 --> 02:12:14,935 agent tools and then the ding 2318 02:12:14,935 --> 02:12:16,835 comes and you know 2319 02:12:16,835 --> 02:12:18,835 that you gotta do something and that's it 2320 02:12:18,835 --> 02:12:19,715 it's pretty smooth 2321 02:12:19,715 --> 02:12:22,775 hold on a second I wanna make sure this is our 2322 02:12:22,775 --> 02:12:24,095 recursive delete 2323 02:12:24,095 --> 02:12:26,875 command I gotta look at it 2324 02:12:26,875 --> 02:12:28,255 like I know what I'm looking at 2325 02:12:28,255 --> 02:12:30,375 this is like crawl 2326 02:12:30,375 --> 02:12:32,915 expected it's a bunch of web calls where's the 2327 02:12:32,915 --> 02:12:33,995 delete I don't see a delete 2328 02:12:33,995 --> 02:12:36,935 I don't see an RM oh no I do 2329 02:12:36,935 --> 02:12:37,435 see RM 2330 02:12:37,435 --> 02:12:49,980 Oh it the temp directory Okay no you good BPP go up We going to be so productive Don let him ask for permission 2331 02:12:50,200 --> 02:12:51,980 Just let him do whatever he wants. 2332 02:12:53,060 --> 02:12:54,120 The agent knows best. 2333 02:12:54,120 --> 02:12:55,120 I'm not ready yet. 2334 02:12:55,440 --> 02:12:56,780 I'm not ready yet, okay? 2335 02:12:56,980 --> 02:12:58,240 I still have the condom on. 2336 02:12:58,340 --> 02:12:59,540 Just give me your sleeper. 2337 02:13:02,100 --> 02:13:04,220 You know, but that's why I have like a dedicated cushion. 2338 02:13:04,220 --> 02:13:07,140 I got gonorrhea coins in my Bitcoin wallet, okay? 2339 02:13:07,140 --> 02:13:09,140 I need to be careful this time. 2340 02:13:09,340 --> 02:13:10,280 This is a new place. 2341 02:13:11,880 --> 02:13:13,640 And you know, this is one of the other things 2342 02:13:13,640 --> 02:13:15,900 that came out of Sovereign Engineering so early. 2343 02:13:16,340 --> 02:13:17,240 It was like, I don't know, 2344 02:13:17,360 --> 02:13:18,340 Sovereign Engineering 3. 2345 02:13:18,900 --> 02:13:19,740 That was like 2346 02:13:19,740 --> 02:13:22,820 March 2024? 2347 02:13:23,640 --> 02:13:24,200 Can that be? 2348 02:13:24,760 --> 02:13:25,520 Maybe not. 2349 02:13:26,040 --> 02:13:27,620 But that was already called YOLO. 2350 02:13:28,580 --> 02:13:31,240 And Pablo was pushing so hard 2351 02:13:31,240 --> 02:13:32,340 on the agents and like, 2352 02:13:32,400 --> 02:13:33,540 don't look at the code, guys. 2353 02:13:33,640 --> 02:13:34,420 Close the IDE. 2354 02:13:35,120 --> 02:13:35,980 It's just chat. 2355 02:13:35,980

--> 02:13:43,180 you know and like radically early and so we've kind of embraced this and dude it's 2356 02:13:43,180 --> 02:13:50,780 it's worth the chaos like dude i'm thinking about going oh marky have you seen this i mean why are 2357 02:13:50,780 --> 02:13:55,600 we still we're still broadcasting pretty cool first you need to say just totally aside have 2358 02:13:55,600 --> 02:14:02,060 you talked to eric case and about vora yeah yeah okay i love the project okay good i just i haven't 2359 02:14:02,060 --> 02:14:06,620 like super dug into it but we had a conversation and i need to get back on him about it but it just 2360 02:14:06,620 --> 02:14:16,380 i immediately think of marmot and and white noise and vora like together in a k-i-s-s-i-n-g um and uh 2361 02:14:16,380 --> 02:14:21,980 so um but what was the other thing what was the thing that that was a tangent from 2362 02:14:23,820 --> 02:14:30,060 i guess like their model on having actually good permissions for your ai because like the 2363 02:14:30,060 --> 02:14:32,380 dangerously skip permissions is very retarded. 2364 02:14:33,380 --> 02:14:33,460 And 2365 02:14:33,460 --> 02:14:36,240 giving them a dedicated box is 2366 02:14:36,240 --> 02:14:38,060 good. That's by the way, I haven't run 2367 02:14:38,060 --> 02:14:39,880 Omarchi because I run Cubes OS, 2368 02:14:40,140 --> 02:14:41,840 which is like the hyper-virtualized 2369 02:14:41,840 --> 02:14:43,320 many virtual machines. 2370 02:14:43,1000 --> 02:14:46,240 It's just hard to step down from something 2371 02:14:46,240 --> 02:14:47,160 that is that secure. 2372 02:14:48,380 --> 02:14:50,280 So my agents just have their own 2373 02:14:50,280 --> 02:14:51,560 virtual machines, basically. 2374 02:14:52,080 --> 02:14:53,720 And they're the king of that. 2375 02:14:54,040 --> 02:14:54,820 And that's fine. 2376 02:14:55,960 --> 02:14:57,600 So is it 2377 02:14:57,600 --> 02:14:59,680 Omarchi or Omarchi? 2378 02:15:00,060 --> 02:15:02,720 because I saw it 2379 02:15:02,720 --> 02:15:05,180 my brain saw it in like an oligarchy 2380 02:15:05,180 --> 02:15:05,760 like 2381 02:15:05,760 --> 02:15:08,120 G-A-R-C-H-Y 2382 02:15:08,120 --> 02:15:10,200 like an omarchy 2383 02:15:10,200 --> 02:15:12,940 like a model 2384 02:15:12,940 --> 02:15:14,940 a governance model or an 2385 02:15:14,940 --> 02:15:15,880 organization model 2386 02:15:15,880 --> 02:15:18,1000 I'm trying to remember 2387 02:15:18,1000 --> 02:15:20,420 how we said it in the podcast 2388 02:15:20,420 --> 02:15:22,580 regardless so Jeff have you seen this 2389 02:15:22,580 --> 02:15:23,780 yeah 2390 02:15:23,780 --> 02:15:26,900 it is a Linux distro 2391 02:15:26,900 --> 02:15:29,180 that is designed 2392 02:15:29,180 --> 02:15:32,340 around the AI just being the distro. 2393 02:15:33,700 --> 02:15:36,840 Like, so you literally, you run your AI, 2394 02:15:36,1000 --> 02:15:38,040 you plug into an AI, 2395 02:15:38,360 --> 02:15:41,280 and the AI just, you just build the interface 2396 02:15:41,280 --> 02:15:42,200 as you're moving. 2397 02:15:42,680 --> 02:15:44,560 If you want to change something about your computer, 2398 02:15:45,340 --> 02:15:47,400 you just tell it, you just describe the change. 2399 02:15:48,420 --> 02:15:50,840 So like, it's just like, 2400 02:15:50,900 --> 02:15:52,300 that could get weird so fast. 2401 02:15:52,300 --> 02:15:53,800 It could get weird real quick, right? 2402 02:15:53,800 --> 02:15:55,720 I can, and like, 2403 02:15:55,720 --> 02:15:59,320 But that's kind of what I do already with Linux. 2404 02:15:59,580 --> 02:16:01,260 You know, I've gotten to the point where my, 2405 02:16:01,260 --> 02:16:03,920 I have not opened my Linux machine in like two months. 2406 02:16:04,520 --> 02:16:07,760 My Zorin has a Kimmy and it just, 2407 02:16:08,500 --> 02:16:10,980 I'm just a Kimmy and a Quinn actually, 2408 02:16:11,360 --> 02:16:12,640 depending on what I'm doing with it. 2409 02:16:13,100 --> 02:16:16,540 And I literally just, it installs stuff. 2410 02:16:16,720 --> 02:16:17,700 It runs the stuff. 2411 02:16:17,820 --> 02:16:21,500 It does the other side of pair drive with Linux. 2412 02:16:21,760 --> 02:16:24,800 And the only times that I am actually doing something myself 2413 02:16:24,800 --> 02:16:27,880 with that pair drive is it's remotely connected into my thing. 2414 02:16:27,880 --> 02:16:29,640 So I'm using it in my browser on my MacBook. 2415 02:16:29,640 --> 02:16:41,845 I literally don open my Linux machine anymore I just have remote access to everything Or I have a model or I have an agent that gives me the access I need when I don Um and uh I just want it 2416 02:16:41,845 --> 02:16:46,365 to be easier. I just want the pair core to be the thing so that I don't have to do all these extra 2417 02:16:46,365 --> 02:16:55,805 steps. Uh, and, uh, but it just, it's very intriguing. Cause I realized like the, my behavior 2418 02:16:55,805 --> 02:17:00,765 with how I use the computer is slowly going that way. 2419 02:17:01,145 --> 02:17:03,425 Like it just, and I'm constantly, 2420 02:17:03,625 --> 02:17:06,145 I have like some security cron jobs 2421 02:17:06,145 --> 02:17:08,005 that just like every morning, 2422 02:17:08,005 --> 02:17:10,165 I just go look in our chat. 2423 02:17:10,645 --> 02:17:12,465 And it's just a cron job that just like scans 2424 02:17:12,465 --> 02:17:16,125 like all GitHub and like history 2425 02:17:16,125 --> 02:17:18,225 of like every single package on the computer. 2426 02:17:18,365 --> 02:17:20,405 And like somebody built a dev tool for this. 2427 02:17:20,985 --> 02:17:23,265 And it just looks for security patches. 2428 02:17:23,265 --> 02:17:24,585 It looks for new exploits. 2429 02:17:24,585 --> 02:17:35,385 It looks for malicious branches or whatever, like requesting update or any connection or notification to something that could be bad. 2430 02:17:35,685 --> 02:17:41,245 And it just reports the entire stack of stuff installed on my

computer, basically. 2431 02:17:41,425 --> 02:17:48,865 It says, no new exploits, no references to dangerous or questionable things today. 2432 02:17:49,645 --> 02:17:54,385 And if it finds one, it just proposes a solution and replaces it with something else. 2433 02:17:54,585 --> 02:17:56,405 it's crazy 2434 02:17:56,405 --> 02:17:57,285 it's crazy 2435 02:17:57,285 --> 02:17:57,985 and it's like 2436 02:17:57,985 --> 02:17:58,325 one of like 2437 02:17:58,325 --> 02:17:59,005 three or four tools 2438 02:17:59,005 --> 02:18:01,365 guys I really 2439 02:18:01,365 --> 02:18:02,325 gotta head off 2440 02:18:02,325 --> 02:18:04,665 you're an asshole 2441 02:18:04,665 --> 02:18:06,345 it's all your fault 2442 02:18:06,345 --> 02:18:07,205 that's why we talked 2443 02:18:07,205 --> 02:18:07,665 for long 2444 02:18:07,665 --> 02:18:08,745 much longer 2445 02:18:08,745 --> 02:18:09,985 alright guys 2446 02:18:09,985 --> 02:18:10,885 thanks for letting 2447 02:18:10,885 --> 02:18:11,145 you on 2448 02:18:11,145 --> 02:18:12,945 yeah dude 2449 02:18:12,945 --> 02:18:14,325 we'll have you back 2450 02:18:14,325 --> 02:18:14,505 on 2451 02:18:14,505 --> 02:18:15,945 yeah 2452 02:18:15,945 --> 02:18:16,725 would love to 2453 02:18:16,725 --> 02:18:17,985 I mean again 2454 02:18:17,985 --> 02:18:19,005 I'm the only listener 2455 02:18:19,005 --> 02:18:20,145 anyway it seems like 2456 02:18:20,145 --> 02:18:21,525 so if nobody else 2457 02:18:21,525 --> 02:18:22,165 wants to come on 2458 02:18:22,165 --> 02:18:22,645 I'll come back 2459 02:18:22,645 --> 02:18:24,525 hopefully someone else 2460 02:18:24,525 --> 02:18:26,225 and cast together. 2461 02:18:26,445 --> 02:18:28,285 Everybody gets into one big, 2462 02:18:28,345 --> 02:18:29,645 happy four-person 2463 02:18:29,645 --> 02:18:31,205 Riverside chat 2464 02:18:31,205 --> 02:18:34,945 and we sell each other 2465 02:18:34,945 --> 02:18:35,925 our sponsors. 2466 02:18:37,065 --> 02:18:37,425 Yes. 2467 02:18:38,845 --> 02:18:39,685 All right, man. 2468 02:18:39,865 --> 02:18:40,745 Guys, have a good one. 2469 02:18:40,865 --> 02:18:41,165 Later, dude. 2470 02:18:45,745 --> 02:18:46,725 All right, guys. 2471 02:18:47,345 --> 02:18:48,365 That'll wrap us up. 2472 02:18:48,825 --> 02:18:49,865 Thank you for joining me. 2473 02:18:49,945 --> 02:18:51,145 Don't forget to follow 2474 02:18:51,145 --> 02:18:53,145 Max Hillebrand 2475 02:18:53,145 --> 02:18:55,065 and the rest of the crew. 2476 02:18:55,625 --> 02:18:57,585 Check out what's going on with White Noise. 2477 02:18:58,005 --> 02:19:01,285 Make sure you subscribe as well 2478 02:19:01,285 --> 02:19:03,905 if you want to get the videos on a lot of the things 2479 02:19:03,905 --> 02:19:04,885 I've been working on. 2480 02:19:05,005 --> 02:19:09,245 We're so close to kind of starting to push it out 2481 02:19:09,245 --> 02:19:10,865 and wanting to get people involved. 2482 02:19:10,985 --> 02:19:13,405 We do have a small group of people already involved 2483 02:19:13,405 --> 02:19:14,625 who are trying to help test 2484 02:19:14,625 --> 02:19:18,125 and make sure Pear Drop and Pear Drive are actually working 2485 02:19:18,125 --> 02:19:19,865 and we're kind of expanding that. 2486 02:19:19,865 --> 02:19:22,745 So hopefully I have something to share 2487 02:19:22,745 --> 02:19:26,705 generally with you guys very soon that you can play with. 2488 02:19:26,805 --> 02:19:27,705 Don't forget to subscribe. 2489 02:19:28,145 --> 02:19:28,705 Check it out. 2490 02:19:28,805 --> 02:19:31,825 Check me out on Keat and X and all that good stuff. 2491 02:19:32,185 --> 02:19:35,125 Don't forget to grab your discount when you snag your Bitbox. 2492 02:19:35,405 --> 02:19:37,145 My discount code is a big help. 2493 02:19:37,145 --> 02:19:39,105 You can get a cheaper Bitbox. 2494 02:19:39,665 --> 02:19:44,065 Same quality, same quality Bitbox Nova too, but less price. 2495 02:19:44,385 --> 02:19:45,865 That's a win in my book. 2496 02:19:46,365 --> 02:19:51,105 And it also helps out to share if you use my link and my affiliate code. 2497 02:19:51,225 --> 02:19:52,285 So I appreciate it. 2498 02:19:52,285 --> 02:19:59,225 And I will catch you on the next episode of Bitcoin Audible and the next guy's dig. 2499 02:19:59,225 --> 02:20:01,825 And until then, everybody, take it easy. 2500 02:20:22,285 --> 02:20:52,265 Thank you.