1 00:00:00,080 --> 00:00:07,080 Welcome back to the 1031 timestamp recap. I'm your host, Marty Bent, joined by the real John Arnold. 2 00:00:07,300 --> 00:00:11,220 And we spent some time in New York last week in person. It was a great time, John. 3 00:00:11,520 --> 00:00:16,400 It was. We didn't get to meet the other John Arnold, who was along with me, 4 00:00:16,640 --> 00:00:21,380 were collectively worth billions of dollars. But we still have not gotten a chance to meet 5 00:00:21,380 --> 00:00:24,100 in person to compare notes on our wealth, but maybe next time. 6 00:00:24,560 --> 00:00:28,700 It's going to happen. It's going to happen one day. We have it up here. I think this is episode 7 00:00:28,700 --> 00:00:34,540 three and we've noticed, hey, we have no calls to action in the show. So I want to highlight why we 8 00:00:34,540 --> 00:00:39,580 come to do this video every Monday morning. It's because John writes the 1031 timestamp, 9 00:00:39,680 --> 00:00:44,700 goes out Saturday mornings, great little weekend read. So if you want to subscribe to that, 10 00:00:44,700 --> 00:00:51,100 go to 1031 timestamp.com. That's T-E-N-31 timestamp.com. Make sure you get all the lists. 11 00:00:51,300 --> 00:00:56,200 But as we said, we were in person in New York last week for Bitcoin Investor Week. We spent a 12 00:00:56,200 --> 00:01:02,500 couple days up there, a meeting with founders and investors around the city. And we spent a lot of 13 00:01:02,500 --> 00:01:06,880 morning time together, particularly at breakfast. Shout out Jack's wife, Frida. We got breakfast 14 00:01:06,880 --> 00:01:13,860 there two days in a row. The crux of the conversation was about the impact that AI is 15 00:01:13,860 --> 00:01:20,660 going to have on the job markets and that viral post that went out last week, which after time 16 00:01:20,660 --> 00:01:26,060 is set in, it seems like it was a bit self-serving. But John, I'll let you do the setup from here 17 00:01:26,060 --> 00:01:29,120 because it was a very stimulating conversation last week at breakfast. 18 00:01:29,820 --> 00:01:31,900 Yeah. I mean, there's a lot we could get into, 19 00:01:32,400 --> 00:01:35,680 but as it relates to that post, 20 00:01:35,920 --> 00:01:39,380 I think it's an interesting, it was a really interestingly timed post. 21 00:01:40,560 --> 00:01:43,000 You know, a lot has been said in the last week about, 22 00:01:43,180 --> 00:01:44,660 I think that got like 50 million views. 23 00:01:45,100 --> 00:01:46,860 If you don't know what we're referring to, just Google. 24 00:01:46,860 --> 00:01:51,060 Something big is happening on Twitter or have your agent do it for you better yet. 25 00:01:51,660 --> 00:01:55,480 But basically a lengthy post that some might call a doom post, 26 00:01:55,480 --> 00:02:16,800 Others might call it more exciting and optimistic, whatever, but describing all the ways that the acceleration in artificial intelligence capabilities across the frontier models are quickly and exponentially leading to potential disruptions to the way that everyone does work in any knowledge worker does work and eventually any worker at all. 27 00:02:17,460 --> 00:02:19,300 And so go give that a read if you haven't. 28 00:02:19,300 --> 00:02:30,320 And there have been a couple, a lot of different chains of thought of reactions around that, a lot of which, you know, call it, well, this is just like an AI slot post or this is a post from a guy promoting his company. 29 00:02:31,440 --> 00:02:36,340 He's encouraging you to go spend money on a bunch of AI tools that may benefit him in some way, you know, whatever. 30 00:02:36,340 --> 00:03:06,320 I think the one thing that's interesting about that post, if nothing else, that's worth pulling out is deep in the post, the author talks about these tools, these models, and increasingly these agentic frameworks developing something like taste and judgment, doing increasingly things that you would previously have thought would be the irreducible component of humanity that's necessary to guide these tools and help them. 31 00:03:06,340 --> 00:03:11,720 them to and force them to make kind of the right decisions and exercise some level of what we would 32 00:03:11,720 --> 00:03:16,500 call, you know, abstract judgment on what not just like what what is right in the context of any 33 00:03:16,500 --> 00:03:21,240 micro task, but how all the micro tasks kind of like fit together. And what's what's the right 34 00:03:21,240 --> 00:03:25,160 thing to be pursuing? What's the right pathway to be pursuing? And what needs to be left aside? 35 00:03:26,160 --> 00:03:32,680 Whether that's ultimately true or not is, I think, a question mark. I know, Marty, that you 36 00:03:32,680 --> 00:03:38,300 on our team have been probably the most active in playing with a lot of these things. So maybe I'll 37 00:03:38,300 --> 00:03:42,520 pause and put a pin in it there and just say like, I'd love, you know, you've talked about this on 38 00:03:42,520 --> 00:03:46,280 other channels and if anyone's following Marty's Twitter feed, I'm sure they've seen it there, 39 00:03:46,280 --> 00:03:51,300 but you know, what's your 60 second reaction to that claim? Cause I think that's the most like 40 00:03:51,300 --> 00:03:55,340 controversial and interesting thing in that post. It's like, we all know, you know, these bots can 41 00:03:55,340 --> 00:03:59,640 do coding well, they can increasingly read legal docs. Well, they're kind of getting to the point 42 00:03:59,640 --> 00:04:01,880 where they can maybe do stuff in Excel well. 43 00:04:01,960 --> 00:04:03,200 So you can kind of see the broad outlines 44 00:04:03,200 --> 00:04:04,960 of how it's like starting to trickle 45 00:04:04,960 --> 00:04:05,760 into white collar work. 46 00:04:05,760 --> 00:04:08,060 But like this idea of these things developing 47 00:04:08,060 --> 00:04:10,120 like taste and judgment independently, 48 00:04:10,120 --> 00:04:12,180 like what have you seen with that? 49 00:04:12,240 --> 00:04:13,120 How did that strike you? 50 00:04:13,740 --> 00:04:17,100 I haven't seen it as much as what Matt, 51 00:04:17,220 --> 00:04:18,100 I forget his last name, 52 00:04:18,320 --> 00:04:21,500 who wrote the post as he described in the article. 53 00:04:22,020 --> 00:04:24,460 But I think definitely if you're working 54 00:04:24,460 --> 00:04:28,800 at an AI company and specifically the hyperscalers 55 00:04:28,800 --> 00:04:31,880 top tier models, OpenAI, Anthropic. 56 00:04:32,280 --> 00:04:33,740 You're probably seeing that internally 57 00:04:33,740 --> 00:04:36,760 as somebody using OpenClaw specifically. 58 00:04:36,980 --> 00:04:39,380 And we've been using Clawed code 59 00:04:39,380 --> 00:04:41,180 for as long as it's been out. 60 00:04:41,280 --> 00:04:45,920 We've been using, we decided to use Clawed internally 61 00:04:45,920 --> 00:04:48,340 at TFTC over a year ago and stuck to it. 62 00:04:48,380 --> 00:04:51,620 So we've been staying up to date with the latest models. 63 00:04:51,620 --> 00:04:54,420 And now with OpenClaw, it's really sort of extended 64 00:04:54,420 --> 00:04:55,600 what we can do. 65 00:04:55,600 --> 00:05:10,300 And it's been honestly shocking over the last month implementing it and seeing how it can help us, TFTC, automate stuff on the back end so I can focus on a lot of the 1031 stuff more as well when we've been leaning into it. 66 00:05:10,300 --> 00:05:19,820 But when it comes to developing taste and recognizing that development via my interactions with OpenClaw, I'll be honest, haven't seen it. 67 00:05:19,820 --> 00:05:27,740 It's a lot of sort of prodding and iterative conversation that I have with my OpenClaw agent to get it to do exactly what I want to do. 68 00:05:27,820 --> 00:05:29,460 And it still has memory issues. 69 00:05:29,460 --> 00:05:30,940 And this could be a product of my setup. 70 00:05:31,080 --> 00:05:39,300 Point being, I would not be surprised if engineers inside of Anthropic OpenAI are beginning to see this taste develop. 71 00:05:39,300 --> 00:05:41,540 I don't think it's hit mainstream yet. 72 00:05:41,540 --> 00:05:47,960 And to your point about the job disruption, in terms of implementing these tools at companies, 73 00:05:47,960 --> 00:05:49,780 I think that's a lot of what we discussed last week. 74 00:05:49,780 --> 00:05:55,180 There is still a massive learning curve, even if somebody has been playing with the top 75 00:05:55,180 --> 00:05:58,540 of the line tools for the last two years and a team that has been doing it. 76 00:05:58,540 --> 00:06:06,300 I guess we're a good example of a company these large language model providers are targeting. 77 00:06:06,460 --> 00:06:07,260 We're a media company. 78 00:06:07,440 --> 00:06:08,240 We're trying to stay lean. 79 00:06:08,240 --> 00:06:11,740 I guess we're one archetype of one company that they could be targeting. 80 00:06:12,120 --> 00:06:13,900 We've been experimenting with it for three years. 81 00:06:14,060 --> 00:06:14,980 Now we're using OpenClaw. 82 00:06:15,160 --> 00:06:16,000 It certainly helped us. 83 00:06:16,260 --> 00:06:17,020 The taste isn't there. 84 00:06:17,020 --> 00:06:19,860 Point being, if you're going down the list of every company 85 00:06:19,860 --> 00:06:22,600 that is going to need to incorporate this, 86 00:06:22,680 --> 00:06:24,240 they're going to have to go through that learning curve too. 87 00:06:24,740 --> 00:06:29,360 Until that taste factor hits these mainstream applications, 88 00:06:29,780 --> 00:06:33,340 I don't think everybody's going to see it 89 00:06:33,340 --> 00:06:36,240 as people inside Anthropic or OpenAI 90 00:06:36,240 --> 00:06:39,620 maybe seeing it up close and personal. Does that make sense? 91 00:06:39,980 --> 00:06:43,500 Yeah, for sure. No, I mean, I think it's one of the things that we talked about last week, 92 00:06:43,500 --> 00:06:49,240 you hit on this point was this idea of like James Burnham's framework for like the, you know, 93 00:06:49,240 --> 00:06:53,280 he wrote in like the 40s, this book, The Managerial Revolution about the emergence of like the 94 00:06:53,280 --> 00:06:59,480 managerial class in the modern kind of Western, you call it capitalist, but he might even kind of 95 00:06:59,480 --> 00:07:13,614 dispute that term But what we think of as like the modern you know advanced Western economy is made up of these organizations that are themselves dominated by and controlled by like managerial elites of all these different types And they have different functions 96 00:07:14,134 --> 00:07:17,034 things, different things that they're, you know, they excel at or don't excel at. 97 00:07:17,474 --> 00:07:22,214 But one of his big kind of takeaways is like, this is like, it kind of takes that structure, 98 00:07:22,294 --> 00:07:26,714 that infrastructure takes on a life of its own. And it's really good at perpetuating itself and 99 00:07:26,714 --> 00:07:32,634 keeping itself alive and like finding new ways to add to its fiefdom. You know, this idea of like, 100 00:07:32,634 --> 00:07:36,154 if you've ever been a manager or talked to managers in kind of corporate America, you know, 101 00:07:36,174 --> 00:07:41,134 the idea of getting like budget allocations or headcount allocations for your department is like 102 00:07:41,134 --> 00:07:44,374 a, you know, a badge of honor, a badge of pride. It's kind of what you're optimizing for, 103 00:07:44,474 --> 00:07:49,194 even in like less functional organizations, like you can end up optimizing for that more than like 104 00:07:49,194 --> 00:07:53,394 actual, you know, relevant KPIs that are theoretically most valuable to shareholders. 105 00:07:53,894 --> 00:07:57,914 And certainly that's even more of the case in government institutions, right? So this, 106 00:07:57,914 --> 00:08:05,054 I think there's this idea that maybe certain like incredibly hyper intelligent, hyper forward 107 00:08:05,054 --> 00:08:11,134 thinking like AI maxis have not fully grappled with that. Just like it takes a lot longer than 108 00:08:11,134 --> 00:08:18,174 like a rational economic kind of model would suggest for disruptive transformative technologies 109 00:08:18,174 --> 00:08:24,134 to actually break into these big organizations just because there's so much inertia from the 110 00:08:24,134 --> 00:08:28,654 the way that the managerial class is kind of set up and structured and the incentives are like not 111 00:08:28,654 --> 00:08:33,994 there for anyone to go disrupt their own headcount. You know, the incentives are not there for people 112 00:08:33,994 --> 00:08:37,594 to like train their replacements basically, right? Like whether that's a human replacement or like a 113 00:08:37,594 --> 00:08:43,254 robot replacement. And, you know, there are a lot of jobs that could have been basically like 114 00:08:43,254 --> 00:08:47,854 optimized away in both certainly the government, but also like corporate America over the last 20 115 00:08:47,854 --> 00:08:52,414 years through a bunch of different, you know, advancements in just basic technology, SAS products, 116 00:08:52,414 --> 00:08:58,554 you know, VBA and Excel. But like, we still have tens of thousands, hundreds of thousands of jobs 117 00:08:58,554 --> 00:09:01,954 that, you know, have continued to persist like through that. And I think it's just a testament to 118 00:09:01,954 --> 00:09:08,214 how much friction there is and inertia there is in organizations to actually disrupt themselves 119 00:09:08,214 --> 00:09:13,254 and adopt new technologies that are going to, that would potentially have a threat of leading to like 120 00:09:13,254 --> 00:09:18,114 a shrinking headcount or whatever. Right. So I think it's like very much worth considering 121 00:09:18,114 --> 00:09:22,854 the sand and the gears that will dominate a lot of these organizations that theoretically 122 00:09:22,854 --> 00:09:28,114 should benefit the most from like totally gutting their existing SaaS stacks and implementing 123 00:09:28,114 --> 00:09:33,454 something else that's based on AI or like agentic frameworks. But all that said, you know, 124 00:09:33,474 --> 00:09:38,294 these things move, we're not well prepared to think in exponents and like the way how quickly 125 00:09:38,294 --> 00:09:43,154 these things can move. And so like, yeah, it's probably not the case that six months from now, 126 00:09:43,614 --> 00:09:47,854 like you're gonna have mass unemployment because every organization suddenly realizes 127 00:09:47,854 --> 00:09:53,114 like they can cut a bunch of heads and they can, you know, get rid of their existing software stack 128 00:09:53,114 --> 00:09:59,834 and totally replace it and overhaul it. But it's also, I think, not a great bet to bet against, 129 00:09:59,834 --> 00:10:05,894 like over time, the destruction of, or at least the reduction of like excess margin in a cost 130 00:10:05,894 --> 00:10:12,114 structure. You know, like the managerial elite dynamic is like one big thing that pushes against 131 00:10:12,114 --> 00:10:15,974 that, but it's like always in competition with this countervailing force of, you know, what 132 00:10:15,974 --> 00:10:21,754 Bezos said about your margins by opportunity, like the nature just abhors excess cost. And like, 133 00:10:21,774 --> 00:10:25,754 eventually it will like extract its pound of flesh from excess cost in some way, even if it can't 134 00:10:25,754 --> 00:10:31,154 like fully get it all. And so I think it's very much worth reading that piece with the mindset of 135 00:10:31,154 --> 00:10:36,674 like, even if this is sensationalist and like a little too, you know, too on the doomer side of 136 00:10:36,674 --> 00:10:40,974 things, and it assumes things are going to happen faster than they are. You know, I think it's worth 137 00:10:40,974 --> 00:10:45,214 thinking about, well, what if it's like 20% right over the next like two to three years, right? 138 00:10:45,214 --> 00:10:51,834 what would that mean for every investable asset class and the economy? And I think maybe that's 139 00:10:51,834 --> 00:10:56,994 like a good time to transition to the chart of the week here in the timestamp, because it was 140 00:10:56,994 --> 00:11:01,574 poetic timing that everyone, the internet was freaking out about, you know, this piece and 141 00:11:01,574 --> 00:11:06,434 these ideas and all, you know, not just software stocks anymore this week. It was everything from 142 00:11:06,434 --> 00:11:10,394 insurance to ratings agencies, to trucking companies were all going through like these 143 00:11:10,394 --> 00:11:15,174 kind of like darkly funny, like rolling sell-offs every day of like, you know, today, like all the 144 00:11:15,174 --> 00:11:18,614 trucking companies are down 20% because like somebody vibe coded like a freight forwarding, 145 00:11:18,614 --> 00:11:23,834 you know, software stack from, you know, from quad code or whatever. And you've kind of seeing 146 00:11:23,834 --> 00:11:28,554 all that. So it's funny that all of that was happening and coinciding with the release of 147 00:11:28,554 --> 00:11:34,594 the data underlying this chart. So this is the CBO just released their annual, annually, 148 00:11:34,594 --> 00:11:40,134 they refresh their projections for the next decade ahead on the federal budget revenues and spending 149 00:11:40,134 --> 00:11:47,614 and deficits. And you saw they revised this decade's cumulative deficit up by 1.4 trillion 150 00:11:47,614 --> 00:11:52,434 over the next decade. And it was interesting that they would put this out this week, because if you 151 00:11:52,434 --> 00:11:59,354 go into the piece that they released, kind of giving voiceover to the data, and you control 152 00:11:59,354 --> 00:12:05,874 artificial intelligence or AI, I think you get like one hit back. And it's like some suggestion 153 00:12:05,874 --> 00:12:10,194 that ai could lead to like increased productivity um but there's no there's no section in there 154 00:12:10,194 --> 00:12:14,754 that's like well here's how we sensitize the scenario of if this guy you know if this thesis 155 00:12:14,754 --> 00:12:19,234 is right or even if this thesis is like 10 right here's what that does to like automatic stabilizer 156 00:12:19,234 --> 00:12:24,274 payments or incremental unemployment payments or like new like federal reskilling programs we're 157 00:12:24,274 --> 00:12:28,594 gonna have to roll up to make sure that we don't have like 20 unemployment or whatever right it 158 00:12:28,594 --> 00:12:32,834 doesn't really think through any of that stuff and so i think it's just it's very interesting like 159 00:12:32,834 --> 00:12:55,494 Like we're in like a, the government is in a really tenuous position in many ways, but I think it's funny that, you know, again, darkly funny that we're looking at these budget deficit projections, which are already, you know, if you look at their forward projections out even beyond this, kind of just assuming that federal debt to GDP held by the public is just up and to the right forever to like 200% in like, you know, 15 years or whatever it is. 160 00:12:55,494 --> 00:13:15,954 We're already in that position. And this isn't even accounting for what happens if like some small version of this massive disruption takes place. And so, yeah, I think it all just like, it's an interesting like set of synchronicities this week to all these things to kind of happen at the same time. Yeah. And here's the, here's the really the money shot chart for the next 20 years. 161 00:13:15,954 --> 00:13:22,994 and it just tells you how non-consensus this kind of still is outside of like schizopo thing, 162 00:13:23,054 --> 00:13:26,854 Twitter, like all of us, you know, there are literally dozens of us or literally, you know, 163 00:13:27,054 --> 00:13:30,714 tens of thousands of us who are thinking about this every day and kind of going down these rabbit 164 00:13:30,714 --> 00:13:36,014 holes, but it's not in the math yet, right? The math is already like tenuous and not great. 165 00:13:36,394 --> 00:13:41,914 And this massive disruptive force is still like not even close to like baked into all these numbers. 166 00:13:41,914 --> 00:13:53,594 Well, that's one thing I've been trying to discern over the last week, particularly after our conversations at breakfast in New York, is what percentage of the population do you think is aware of this stuff? 167 00:13:53,594 --> 00:13:56,134 Like we're very online, particularly on X. 168 00:13:56,134 --> 00:14:00,894 And I think since we're experimenting with this and reading as much content 169 00:14:00,894 --> 00:14:05,174 about Bitcoin, AI, the intersection of both, we're definitely getting fed 170 00:14:05,714 --> 00:14:22,708 a bunch of sort of positive news about the proliferation of this technology just because of the algo and just picked up on us liking this content and consuming it rather voraciously And that what I trying to discern is like what percentage of the overall 171 00:14:22,708 --> 00:14:29,168 population have these ideas actually penetrated to? Like, is it 0.1%? Is it 1%? Is it 5%? 172 00:14:30,088 --> 00:14:34,808 My gut tells me it's less than 1%. Yeah, no, I think it's absolutely right. 173 00:14:35,248 --> 00:14:40,308 You know, I think most people that I talk to anecdotally, this isn't very like useful data, 174 00:14:40,308 --> 00:14:46,728 but like most people that I talk to, you know, really have their, their exposure to any of this 175 00:14:46,728 --> 00:14:50,908 is like, they've used chat GPT a couple of times. Right. And it's, you know, something that the, 176 00:14:50,908 --> 00:14:54,348 the author of that piece kind of wrote about that you use it a couple of times it, you know, 177 00:14:54,348 --> 00:14:57,788 maybe you're on the free version and you're at like a, you know, using a severely degraded model 178 00:14:57,788 --> 00:15:01,308 relative to kind of what's available on the frontier or even just with like a $20 plan. 179 00:15:01,308 --> 00:15:05,728 Right. And it kind of maybe hallucinates once, or it doesn't fully like understand what you're 180 00:15:05,728 --> 00:15:08,868 asking for. And it's like, Oh, well, this is basically just kind of like a different version 181 00:15:08,868 --> 00:15:11,948 of a Google search, I'll just go back to Google. And I think that's, there's a, there are a lot of 182 00:15:11,948 --> 00:15:17,208 people out there that for whom that is like the case. Um, and so, yeah, I don't, in my anecdotal 183 00:15:17,208 --> 00:15:21,008 conversations, like there's very little understanding of how quickly a lot of this is 184 00:15:21,008 --> 00:15:26,708 moving on the frontier. None of that is necessarily to say that you can totally extrapolate forward and 185 00:15:26,708 --> 00:15:30,948 think that in two years, all the white collar work will necessarily be fully automated, 186 00:15:30,948 --> 00:15:36,828 but I can guarantee you that like that isn't even in like the, as you think about like the, 187 00:15:36,828 --> 00:15:41,868 decision tree or like the scenario tree for the next year, two years, five years, the vast majority 188 00:15:41,868 --> 00:15:46,408 of people don't have any scenario on that table that involves something like that happening. 189 00:15:46,408 --> 00:15:51,248 Right. And the probability may, that may only be like 10% and AI maxes think it's like a hundred 190 00:15:51,248 --> 00:15:55,688 percent, right? Maybe it's only 10%, but most people, you know, for most people, it's 0%, 191 00:15:55,688 --> 00:15:59,528 right. That in the way that they're thinking about it. So all that is to say, yeah, like I, 192 00:15:59,588 --> 00:16:06,628 I think all of this is quite non-consensus still. And if it's even partially true, if it even, 193 00:16:06,628 --> 00:16:16,808 And if some version of it turns out to be true, you know, I think it's going to have meaningful implications for what the government is going to have to do in a bunch of different domains over the next five years. 194 00:16:17,408 --> 00:16:28,448 And that's going to have implications for how you want to have your assets allocated and also what you might want to just be kind of getting started on, you know, getting experience with with with some of these tools over the next year or two. 195 00:16:28,448 --> 00:16:35,668 Yeah, it's it's a while. So that's another thing I've been thinking through is how like what shocks the public awake? 196 00:16:35,668 --> 00:16:39,868 and maybe not even the public, but large companies, large corporations, 197 00:16:40,088 --> 00:16:42,248 particularly the publicly traded ones. 198 00:16:42,368 --> 00:16:46,328 And I'm sure, as on all the radars, but you have to imagine that, 199 00:16:46,648 --> 00:16:51,148 again, going back to the sort of managerial framework that you described earlier, 200 00:16:51,148 --> 00:16:56,288 there's a lot of inertia and people are probably not in a rush to replace themselves. 201 00:16:56,668 --> 00:17:02,108 And so I think the shock that is really going to wake people up is just you have something like a startup 202 00:17:02,108 --> 00:17:04,948 or one-man startup vibe coding something 203 00:17:04,948 --> 00:17:07,908 that completely takes out one of these incumbents, 204 00:17:08,128 --> 00:17:10,988 undercuts the cost, has a user experience 205 00:17:10,988 --> 00:17:12,528 that's at parity or superior 206 00:17:12,528 --> 00:17:15,408 and is widely adopted very quickly. 207 00:17:15,408 --> 00:17:18,668 And I think we have maybe a glimpse of that. 208 00:17:18,768 --> 00:17:21,108 It's not really disrupting incumbents, 209 00:17:21,188 --> 00:17:24,328 but it's disrupted a ton of the VC capital 210 00:17:24,328 --> 00:17:27,108 that's been deployed into AI agent wrappers 211 00:17:27,108 --> 00:17:28,848 over the last couple of years, which is OpenClaw, 212 00:17:28,848 --> 00:17:43,708 this open source AI wrapper that adds an incredible memory engine to your cloud code or GPT models that you're using and makes them incredibly more useful. 213 00:17:44,508 --> 00:17:48,288 And so Peter Steinberger, for those who are unaware, got picked up by OpenAI. 214 00:17:48,928 --> 00:17:52,628 Yesterday, OpenClaw is going to be an open source foundation, at least for now. 215 00:17:52,628 --> 00:18:04,768 We've seen the story with OpenAI before, but point being is you had this guy vibe coding in his apartment for three months and basically got taken out by OpenAI over the weekend. 216 00:18:05,388 --> 00:18:11,648 And there's been hundreds of millions, billions of dollars of capital allocated towards companies that are supposed to do what he did by himself. 217 00:18:11,748 --> 00:18:13,348 Well, eventually it wasn't by himself. 218 00:18:13,468 --> 00:18:16,768 It became a very popular open source project rather quickly. 219 00:18:16,928 --> 00:18:20,708 But I think that's an example of this disruption we're trying to highlight here. 220 00:18:21,108 --> 00:18:21,588 Yeah. 221 00:18:21,588 --> 00:18:42,408 Yeah. No, I mean, I think it's a good dynamic to think about because, you know, I put in the newsletter this week as well that I think it was UBS highlighted that, you know, they're estimating something like $100 billion of write-offs or other credit losses in the broader private credit complex, which is like trillions of dollars. 222 00:18:43,088 --> 00:18:55,008 So right now it's kind of a drop in the bucket, but attributing those losses to different software companies or other companies that have that are seeing disruption on the margin from how quickly AI is moving. 223 00:18:55,328 --> 00:19:01,328 And they highlighted in their report as well that, you know, they're saying it's a hundred billion dollars, but there are tail risks where it could be like quite a lot higher. 224 00:19:01,388 --> 00:19:11,608 And if you go look at I believe it was Bruce Richards from Marathon was on CNBC this past week as well, kind of giving an interview about kind of the same dynamic, noting he's increasingly expecting something similar. 225 00:19:11,608 --> 00:19:21,988 You've had Orlando Bravo from Toma Bravo come on CNBC a couple of times in the last couple of weeks to talk about, you know, it's a private equity firm focused famously on software companies and SaaS, among other things. 226 00:19:21,988 --> 00:19:25,528 But, you know, talking about how, you know, he doesn't see that being the case and he's kind of defending the thesis. 227 00:19:25,728 --> 00:19:32,088 And, you know, I have no dog in a fighter or a view on necessarily where it's going to go with software in the next year. 228 00:19:32,328 --> 00:19:39,048 But there is definitely like a there's a lot of leverage tied to that sector, like the legacy software sector. 229 00:19:39,048 --> 00:20:09,028 And, you know, there used to be this meme that like software contracts are better than first lien debt. Like you're just you always pay to keep the business running in the lights on like you're always going to pay the, you know, the software contracts, the enterprise software contracts. They're very, you know, sticky and low, low churn, good pricing power because they're so kind of locked into the workflows and everything. And I think a lot of that's still true. But I think a lot of credit has been issued on that kind of framework. Right. And so, you know, it doesn't really take that much given how levered the whole system is broadly, not just not just like 230 00:20:09,028 --> 00:20:15,868 private credit, but like the water that we swim in, it doesn't take that much for a few funds to 231 00:20:15,868 --> 00:20:21,528 on the margin, you know, the laggard funds that are exposed to the worst opportunities. Like if 232 00:20:21,528 --> 00:20:26,568 you start seeing those losses pile up, you know, for reasons that most people aren't, don't even 233 00:20:26,568 --> 00:20:31,428 have reason to be aware of yet, you could start to trigger some like, you know, meaningful financial 234 00:20:31,428 --> 00:20:35,548 instability that the authorities would need to respond to quickly. And so I wonder if something 235 00:20:35,548 --> 00:20:39,548 like that could also be a kind of a wake-up call that like oops we kind of like borked the financial 236 00:20:39,548 --> 00:20:44,188 system because like a few you know the the bottom quartile like software companies are getting 237 00:20:44,188 --> 00:20:48,828 disrupted and that means that a bunch of private credit funds are getting disrupted and that's 238 00:20:48,828 --> 00:20:52,268 going to have daisy chain impacts on the broader credit complex in the us and well we need a new 239 00:20:52,268 --> 00:20:56,348 fed facility to respond to that right so i do wonder if something like that ends up actually 240 00:20:56,348 --> 00:21:00,748 being the thing that shocks a lot of people and companies awake to kind of what's going on yeah 241 00:21:00,748 --> 00:21:03,088 Yeah, great time to own Bitcoin. 242 00:21:03,088 --> 00:21:07,508 And speaking of it, I mean, it's just, it's probably going to be a recurring theme because 243 00:21:07,508 --> 00:21:09,568 AI is evolving at a rapid pace. 244 00:21:09,568 --> 00:21:11,368 We'll probably be talking a lot about this. 245 00:21:11,368 --> 00:21:16,988 So it will be important to highlight the intersection of Bitcoin and AI as it proliferates and particularly 246 00:21:16,988 --> 00:21:17,988 the agentic economy. 247 00:21:17,988 --> 00:21:29,662 And so this week to highlight that intersection we going to talk about Lightning Labs What they released last week which is basically a revamp of their L402 protocol 248 00:21:29,662 --> 00:21:33,402 and making it so it's very easy for agents to down, 249 00:21:33,402 --> 00:21:34,902 to get set up with Lightning wallets 250 00:21:34,902 --> 00:21:37,242 and interact with Lightning invoices. 251 00:21:37,242 --> 00:21:40,882 And so they announced, they released a set of tools 252 00:21:40,882 --> 00:21:42,962 that give agents native access to Lightning Network, 253 00:21:42,962 --> 00:21:44,922 and GET for automatic L402 payments, 254 00:21:44,922 --> 00:21:48,542 MCP for node operations, remote signing for key isolation, 255 00:21:48,542 --> 00:21:50,322 And scope credentials for spending control. 256 00:21:50,622 --> 00:21:53,442 Machine payable web starts now, and Bitcoin makes it possible. 257 00:21:54,162 --> 00:21:58,042 And for anybody who is unaware of this machine payable web idea, 258 00:21:58,322 --> 00:22:00,722 it has been around in Bitcoin since 2015. 259 00:22:01,182 --> 00:22:05,322 I spent one whole Bitcoin in 2015 on a 21Co computer. 260 00:22:05,322 --> 00:22:11,002 And for those who are unaware, Balaji, Srinivansam, he was the CEO of 21Co, 261 00:22:11,122 --> 00:22:14,602 and they revolved the company around this idea of the machine payable web. 262 00:22:14,642 --> 00:22:15,502 And you got this computer. 263 00:22:15,502 --> 00:22:20,782 It was a very small computer, but it was hashing and quote-unquote mining Bitcoin. 264 00:22:20,922 --> 00:22:22,282 You were getting SATs payouts. 265 00:22:22,602 --> 00:22:29,162 But the idea was that you would use these SATs payouts on-chain to enable this machine payable web. 266 00:22:29,402 --> 00:22:31,982 And like many ideas in Bitcoin, it was a great idea. 267 00:22:32,122 --> 00:22:32,842 It was just too early. 268 00:22:33,442 --> 00:22:39,342 Fast forward 11 years, and now we're at the point where the machine payable web makes sense on top of Bitcoin, 269 00:22:39,342 --> 00:22:43,562 specifically because we have payments networks, 270 00:22:43,902 --> 00:22:47,082 sub-protocols on Bitcoin that actually make this feasible. 271 00:22:47,542 --> 00:22:50,862 It didn't make sense in 2015 because agents receiving 272 00:22:50,862 --> 00:22:54,122 and sending on-chain payments to get tasks done in milliseconds 273 00:22:54,122 --> 00:22:55,042 just doesn't make sense. 274 00:22:55,182 --> 00:22:57,582 With the Lightning Network, it's now incredibly possible. 275 00:22:57,582 --> 00:23:00,682 So this has also been a big topic of discussion 276 00:23:00,682 --> 00:23:03,342 within the Bitcoin sphere over the last two months is, 277 00:23:03,502 --> 00:23:04,422 okay, agents are here. 278 00:23:04,622 --> 00:23:05,742 They're going to need to spend money. 279 00:23:06,202 --> 00:23:07,502 What do we need to do to make sure 280 00:23:07,502 --> 00:23:09,422 that they're able to spend Bitcoin easily. 281 00:23:09,642 --> 00:23:12,462 We've seen many people rush to get developer kits 282 00:23:12,462 --> 00:23:15,622 and MCP protocols to market Lightning Labs 283 00:23:15,622 --> 00:23:18,842 who created the L402 protocol being one of the latest. 284 00:23:19,262 --> 00:23:20,762 I believe Breeze has one out there. 285 00:23:21,862 --> 00:23:25,082 Money DevKit, Nick Slaney is building that out. 286 00:23:25,182 --> 00:23:26,562 That's existed for a few months. 287 00:23:26,662 --> 00:23:29,262 They implemented with Replit at the end of last year. 288 00:23:29,902 --> 00:23:33,022 And so we are seeing machine payable web emerge 289 00:23:33,022 --> 00:23:36,562 and the sort of value prop of Lightning 290 00:23:36,562 --> 00:23:37,962 emerging alongside it 291 00:23:37,962 --> 00:23:40,422 or being reconfirmed alongside it. 292 00:23:40,942 --> 00:23:41,042 Yep. 293 00:23:41,522 --> 00:23:43,722 And I think like one of the most exciting things 294 00:23:43,722 --> 00:23:44,682 about this moment too, 295 00:23:44,722 --> 00:23:45,622 is like we've been talking about, 296 00:23:45,702 --> 00:23:46,442 as you referenced, 297 00:23:46,702 --> 00:23:48,942 like L402 for a while, 298 00:23:49,162 --> 00:23:51,042 we wrote some pieces back in the day, 299 00:23:51,102 --> 00:23:51,722 a few years ago, 300 00:23:51,722 --> 00:23:54,462 on kind of the eventual use of that 301 00:23:54,462 --> 00:23:56,722 and the intersection of Bitcoin and AI 302 00:23:56,722 --> 00:23:58,142 that we thought would eventually happen. 303 00:23:58,662 --> 00:24:01,322 I think the really interesting element 304 00:24:01,322 --> 00:24:02,202 of what's going on right now though 305 00:24:02,202 --> 00:24:07,282 was Matt Corralo, longtime Bitcoin developer, had a post last week about how we're now in a position 306 00:24:07,282 --> 00:24:12,902 where a few years ago, if you had, I mean, the L4-2 concept has existed for years and it was 307 00:24:12,902 --> 00:24:18,802 highlighted a few years ago as AI really started to ramp up, but only now are we at a point where 308 00:24:18,802 --> 00:24:23,182 the barrier to entry to like do something with that has collapsed to like essentially zero. 309 00:24:23,382 --> 00:24:27,342 Like a few years ago, you know, if you had seen that post from Lightning Labs, you'd basically 310 00:24:27,342 --> 00:24:31,462 say like, okay, that's great. That's cool. Now let's wait for like someone to go get like VC 311 00:24:31,462 --> 00:24:36,642 capital and with a few million dollars, you know, hire like a few engineers and spend something up 312 00:24:36,642 --> 00:24:40,262 and then, you know, wait to get product market fit and try to find traction before they could 313 00:24:40,262 --> 00:24:44,262 really kind of start to scale it. Or let's wait for some big corporate to like have somebody on 314 00:24:44,262 --> 00:24:48,842 the inside who wants to push for directing some internal resources to this and, you know, whatever. 315 00:24:48,842 --> 00:24:53,582 It's like this kind of abstract primitive that doesn't necessarily have like an immediate path 316 00:24:53,582 --> 00:24:58,622 to realization. Like, well, Matt's point was now we're at a place where like you or I with no, 317 00:24:58,622 --> 00:25:02,242 no technical experience to speak of other than like some vibe coding or like, you know, 318 00:25:02,502 --> 00:25:06,262 taking like a rep one-on-one course, right? Like we, we can just sit down and like, 319 00:25:06,322 --> 00:25:09,202 potentially if we have an idea, like we can just start using that and leveraging that. 320 00:25:09,262 --> 00:25:12,602 And I know you've already been doing it right to some extent with your, with your agents. So 321 00:25:12,602 --> 00:25:17,262 yeah, it's like, it's a really interesting time for the compounding, like flywheel of not just 322 00:25:17,262 --> 00:25:21,162 like Bitcoin and AI have this intersection, but also as AI gets more and more powerful, like 323 00:25:21,162 --> 00:25:25,602 random Bitcoiners, people who are just interested can like sit down and like do something with that 324 00:25:25,602 --> 00:25:30,402 potentially really like really useful. Um, just the only limit, right. Is, is effectively like 325 00:25:30,402 --> 00:25:34,062 their creativity and like their patience with like playing with the tools, but it doesn't require 326 00:25:34,062 --> 00:25:38,442 like, you know, a multi-year software bootcamp or time spent, you know, working at like a massive 327 00:25:38,442 --> 00:25:42,582 tech company to kind of like get their, get their feet under them. Like they can just start and like, 328 00:25:42,582 --> 00:25:46,922 you know, they can just do things right. As the meme goes. So really, really interesting and 329 00:25:46,922 --> 00:25:50,562 exciting time, crazy and somewhat scary, but also a lot of, a lot of opportunities. 330 00:25:50,562 --> 00:26:08,162 Yeah. And just to build on that point and wrap it up here, the possibilities here, like you said, I've been implementing Bitcoin payments with my agents for the last month. It works. It's very easy. There's many ways to go about it. There's many ways to skin the cat of doing it. 331 00:26:08,162 --> 00:26:13,142 But Matt Corallo and Callie quote tweeted that Matt Corallo message. 332 00:26:13,782 --> 00:26:18,002 And Callie is the founder of the creator of the Cachew Protocol, the Chami Amin Protocol. 333 00:26:18,462 --> 00:26:24,242 The point they were both making was like, hey, not only is it possible, but the onus is on us to actually go build these things. 334 00:26:24,362 --> 00:26:28,682 Because I think many Bitcoiners naively believe like, oh, the agents are just going to pick Bitcoin. 335 00:26:28,802 --> 00:26:32,522 It's the best digital native internet currency that's ever existed. 336 00:26:32,622 --> 00:26:33,402 Of course they're going to pick it. 337 00:26:33,442 --> 00:26:33,902 It's not true. 338 00:26:33,902 --> 00:26:39,702 Like I said earlier, it's not at that point, at least on the consumer apps, where it has that taste component. 339 00:26:39,882 --> 00:26:41,702 Maybe it will in the future, but right now it doesn't. 340 00:26:41,882 --> 00:26:43,722 And the agents are going to do what you tell them to do. 341 00:26:44,442 --> 00:26:49,342 And if you tell them to implement a stablecoin protocol, maybe it's Coinbase. 342 00:26:49,822 --> 00:26:54,842 I think they have a base MCP that makes it easy to send and receive stablecoins. 343 00:26:55,062 --> 00:27:02,562 Visa came out and announced that they're going to create a framework for agents to use their network to send payments in a secure way. 344 00:27:02,562 --> 00:27:06,782 obviously Stripe's going to be working on this. And so there's gonna be a ton of competition for 345 00:27:06,782 --> 00:27:11,782 agentic payments outside of Bitcoin. And if we're being frank, they definitely have more firepower 346 00:27:11,782 --> 00:27:18,502 in terms of capital behind them to go after these implementations. And so not only was what Matt's 347 00:27:18,502 --> 00:27:22,282 saying like a message of, hey, this is possible now, but it was also the onus is on us to make 348 00:27:22,282 --> 00:27:28,142 sure that Bitcoin payments become prominent in the agentic world. It's not just going to happen. 349 00:27:28,142 --> 00:27:29,382 There's competition out there. 350 00:27:30,262 --> 00:27:32,282 And I fully agree with that. 351 00:27:32,362 --> 00:27:37,002 I actually had a pretty long conversation with Justin Moon. 352 00:27:37,522 --> 00:27:39,262 And we talked about this at length. 353 00:27:39,722 --> 00:27:42,662 It was released on Saturday on the TFTC feed. 354 00:27:42,742 --> 00:27:45,402 If you guys want to go watch that, I highly recommend that you do, actually. 355 00:27:45,622 --> 00:27:46,082 This is the point. 356 00:27:46,182 --> 00:27:47,402 It's like, we can't be complacent. 357 00:27:47,902 --> 00:27:50,402 It's not going to happen just because Bitcoin is the best money. 358 00:27:50,522 --> 00:27:51,522 We truly believe that. 359 00:27:51,582 --> 00:27:52,622 I think it's objectively true. 360 00:27:52,762 --> 00:27:56,722 But work needs to be done to make that a reality in the agentech world. 361 00:27:56,722 --> 00:27:58,162 Once again, co-sign. 362 00:27:58,582 --> 00:27:59,562 Crazy time to be alive. 363 00:27:59,982 --> 00:28:01,162 Great time to be in Bitcoin. 364 00:28:01,782 --> 00:28:02,602 No doubt. 365 00:28:02,822 --> 00:28:20,322 In a world with seismic shifts in the global economy, geopolitics, domestic politics, a lot of uncertainty, being able to find certainty in a monetary protocol that is very simple, very easy to audit, and very hard to change, has a lot of value. 366 00:28:20,642 --> 00:28:21,602 We'll be back next week. 367 00:28:26,722 --> 00:28:56,702 Thank you.