Loading summary

Transcript47 lines

[00:00]
A
Anthropic destroys millions of books. So this one has been going uber viral across all the different feeds for the past 24 hours, maybe 36 hours. I don't know when exactly the news broke, but the general gist here, as you can see, that court documents revealed that AI company Anthropic spent millions of dollars physically scanning print books to build Claude an AI assistant similar to ChatGPT. You know, if you're, if you're not aware of that. In the process, the company cut millions of print books from their bindings, scanned them into digital files, and threw away the original solely for the purpose of training AI. Details buried in a copyright ruling on fair use, whose broader fair use implications were reported yesterday. So, gentlemen, who would like to go first on this topic? It's, it's an interesting one. It's a little bit different than what we've covered on the show in the past couple weeks here. Who wants to give a first take on signal versus noise today?
[00:49]
B
I'll, I'll go first. And maybe it's a little unexpected, but I do think this is actually mostly noise in the sense that I think it was, people were being a little bombastic in terms of how it was being reported and how it really went viral in terms of people basically insinuating that Anthropic was destroying ancient texts from hundreds or thousands of years ago, that there's only a few versions of, that they're rare, and somehow they're now destroyed forever. Upon further review, that isn't exactly what was happening. When they say rare books, it's not that they're ancient texts, it's that they're very, like, obscure. And so, yes, there's some like, you know, 1984 esque vibes around this entire thing of like, you know, they need to canvas the corpus of, you know, human intelligence and destroy and sort of like, you know, pull up the ladder behind them on the intelligence, and it no longer lives in physical form and it's, it's only in anthropic coffers. And there's also a natural hypocrisy with them complaining and whining about all these distillation efforts on the other end of their models from China and other places where they're doing a form of distillation themselves with all of these books and texts. But I do think it was kind of inaccurately reported based on the court findings of these weren't super valuable one of a kind text that they were destroying to scan, basically. The other nuance here is that the Reason they're destroying them isn't necessarily like super nefarious. It's just like they can scan them faster that way. And then Elon, I think reported or replied to somebody's viral tweet about this being like, we're doing this the hard way, the slow way, where you don't basically rip the binding apart so you can scan it faster. He's like, when we do this, we do it the hard way because it's more moral.
[02:48]
C
Yeah. I mean, I think I'm always skeptical when something goes insanely viral. I'm like, why? Why, why it did. So I think from the angle of what's anthropic doing, how true is it? I think there's noise in that respect. Where I think there's an insane amount of signal is ultimately just the notion of books and there's like in data and information and it being wiped away. There's no shortage of projects. I wish I could. The names are escaping me right now on codifying and capturing the legacy Internet because it's increasingly becoming harder and harder to find. But then as you go further, you think about old text news, information. What made it through the test of time has historically been very few articles, objects, books. And you add to the notion of the idea I think it came out with around literacy decreasing more utilization of these tools, AI. And then I think there's direct analysis, empirical evidence from like physicians around the brain and how it operates specifically from like memory, cognitive ability when you leverage these tools. The example is like spatial data, I think when it comes to maps and sense of direction. And so I've always just had these things in the back of my mind around data and like the information for the Internet. But it also like as a child I remember just always liking to read and I've gotten away from it. I'm trying to get back because I don't know how true this is. But reading an X number of pages or writing really help with that like cognitive ability, especially if you're outsourcing some of it to AI. And I think that as we grow up and our kids grow up, it's just going to be this foreign world to pick up a physical book and read it. And this is just one small manifestation of that signal and it should scare the living crap out of everyone because what these inputs. Remember during COVID when Google told you if can a man when this was super woke time, but can a man have a baby? And it would say yes. And that's the world I think we're definitely Going to. In the sense of not a man having a baby, but the books will tell you. And we see this right now with like the open models and people liking them because the guardrails aren't there. So I think this is just the precursor to what's to come.
[05:04]
D
Yeah, I'll go next. I think that there's a number of different angles here. I'm gonna go with one of them being signal. And not just the actual like burning of books and destroying them, but the fact that Anthropic is trying to train on unique information that isn't widely available to everybody. I think that over time there's going to be more and more information that is on easily accessible through models. And the web is. There just continues to be more information published by agents. And having the sovereignty over your data workflows and specialized earned secrets for companies and your own way that you do things is going to have a lot more value. And that's why I think that increasingly it's becoming almost untenable to leak a lot of these enterprise secrets to frontier AI that are using it to train off their own data.
[05:57]
A
We're a little bit pressed for time here, but I would just say I know there's some ancient manuscripts on men giving birth to children that have been destroyed and since have been gatekept by Anthropic. So that's a big problem that we'll need to solve for.
[06:13]
C
It's a good one, Jackson.
[06:16]
A
All right, let's. We gotta move forward. We have. We have to respect the timer here. So everyone just be mindful of your comments, please. So next one would be grocery store. Grocery prices experienced the biggest jump in 50 years. The headline does a lot of the talking. But Nick, if you could pull up the chart as well. This is really what caught a lot of people's attention online this morning. Where you can see here, if you're not following on YouTube, you can see this is a line chart from starting in 2012 to present in 2026 and showing that the price of food purchased for homes rose 33% from 2019 through June 2026. So call it seven years 30 up 33%. I would argue if you're eating real food, more likely higher than 33%. And so this says consumers are now buying fewer of their favorite foods and less food overall. This actually reminds me, I guess I'll go first because I didn't have a chance on the last one. This reminds me, I think it was from 2023. Wall Street Journal published a piece and the headline was essentially you should skip breakfast because you want to save money on your food prices. And so this is where unfortunately a lot of things are going right where I think that there's a reality that we have an incredibly K shaped economy. This is a term that was coined out of 2020 in particular, where you had all this liquidity go to those who had access to the money spigots. You had zero interest rate policy, you had negative interest rates in some countries. And so naturally there was an outlier in terms of where all the asset prices sit today and those who have fallen behind. And of course, when you inject trillions of dollars of new money, right, liquidity just means new money coming into the system. And this is not just a US problem. Naturally you're going to have a chart that looks like this. I mean the Trend line from 2012 to 2019 is pretty much unmoved in terms of price, food inflation and then it goes vertical in the past seven years. And this is a big problem. It ties into what we could also talk about with New York piloting out state owned grocery stores, which I'm just not surprised to see. I mean this is going to be the biggest trend, I think heading into the 2028 election. There's just a growing populace of people who are really outraged and I can appreciate to a degree why. But unfortunately most people have zero idea what the root cause of this is. So there's a big signal to pay attention to, of course, because if food prices began become unstable, then there's a lot of issues that are downstream of that.
[08:30]
C
Yeah, I mean this is highest amount of signal we could show. And it's why I personally get frustrated every time we talk about the Fed speak aspect and then even just like what's going on with Iran on, off, it doesn't matter. Food prices are going to increase no matter what. The key stat is, you know, American households, the richest country on earth, reducing food consumption while GDP and earnings print records like that's, that's the notion of what we talk about, that's the real versus nominal. And there's no shortage of gaslighting around 2, 3, 4% when it's substantially higher. And it's the only input that matters. If you can't feed yourself, like what else? And then the worst part, and this is like a little esoteric, is this is all like slop. The general food that's being fed to people is just like getting worse. And, and so anyway, this is, this is not good.
[09:19]
B
Yeah, agree with all that. This is a massive signal in the sense that it pierces the veil of manipulated cpi. So anytime you look at the actual underlying inputs to CPI and sort of sift through the manipulation of that sort of endpoint data, you start to see what's actually going on. And to your point, Michael, this isn't even the full picture. If you instead were tracking high quality meats or other foods that you know, you could go get at a farmer's market or something that's like not as slobified then that, that line would be even more aggressively upward sloping in terms of the price of all of these things. And so it's a massive signal and it's, it's a nice reminder of the Fed speak of the manipulated data that is cpi, because this, this is out of control clearly. And then the, the natural foil, as you mentioned, Jackson, is that, you know, Mamdani in New York City is going to try to just basically price fix and have these grocery stores that will sell out immediately and they'll have to figure out some sort of like lottery system of who even gets to go. And it'll lose the, you know, lose the state and city a ton of money because grocery margins are probably the thinnest across like any industry. Pretty much like you know, 2% margins generally around there. And so if you're undercutting prices by 30%, you're losing a ton of money. So it doesn't make any sense. But like, like you said, Jackson, like it's not a totally unsurprising or not a totally surprising reaction to where we are and people don't know the cause and so they look for this solution that doesn't really fix anything, just sort of papers over it.
[11:04]
D
Yep, it's the, the massive signal here is essentially everything was reset at a significantly higher base after 2020, 2021, when many monetary units entered the system. And things are only going to accelerate here in terms of the money printing and then in addition to that where we have a increasingly globally fractured world and higher energy prices, which are a key input to pretty much everything and so should only go higher from here as well as just the increased regulation, whether it's New York City trying to get involved with grocery stores that should prevent from finding the correct clearing prices. And so yeah, there's, there's a lot of signal in here and don't anticipate this getting any better anytime soon.
[11:51]
A
Yeah, Liam, I think what you said toward the end there is the new base, right? I mean it's so, it's so easy for me to think about what prices used to be. And in fact, it can be challenging. I know it's so easy, but it could also be challenging to an extent of like having essentially amnesia about what things used to cost. I posted something maybe a week or two ago about. I forget if it was food prices or something else. And people were in my comments saying, like, why do you care what things cost in 2017? And it was a sandwich shop that I went to like 10 years ago and what their prices were and they have since doubled. It's just insane. Like, nobody understands what, what's going on here.
[12:25]
C
And honestly, I would make anybody a bet this is much higher. It's not reflected because of shrinkflation, because there's no shortage of things that you see, whether it's like toilet paper or paper towels that have like reduction implies or number of wraps or whatever, or the number of things in whatever you're buying or the number of liters or fluid ounces in the. In the deal. So it's much higher. And that's how you mask a lot of this.
[12:48]
A
Yeah, for sure. All right, so the next topic, Hiring is accelerating for AI native skills. So I'll read the headline here from WSJ. Big companies are starting to hire again, defying predictions of an AI wipeout. I mean, this is great for me to hear, of course, because I thought that I was already in the permanent underclass, but now maybe I have a shot. But who wants to take a first crack at the news here?
[13:10]
C
This is noise. This is just gaslighting. Look, so yes, people are hiring for people that are AI native, but it's the equivalent of putting a notice out that food prices are reducing. I can guarantee you if there's 10,000 SKUs at your local supermarket, you might find one that reduced. But while 99% of them are increasing, this is just gaslight. And like, people are getting laid off. It's not because of AI, they're using it as an example, but people increasingly get laid off because of interest rates and reduction in people's capital outlay because of real estate or food. They just spend less. And if they spend less, then there's less consumer demand for your products and services that you have to lay people off or you go out of business. It's just a recursive downward loop.
[13:53]
B
I would agree. I think this is mostly noise in the sense that you have to recognize that we're in this very unique, interesting sort of transition period where, yes, on the margin there's going to be incremental new hires for people with AI native skills. But to your point, Michael, those are going to be massively outweighed by the amount of displacement and job loss. We're just scratching the surface of it. I don't think we've really seen it materially. Like all the layoffs we've seen I don't think are directly related to people just being replaced by AI, but you're starting to see small slivers of it. And we haven't even hit the tipping point of like real robotics boom where in terms of like actual physical labor, there's going to be massive, massive displacement. And so while this might be like a little near term blip of like, yeah, okay, big companies realize that they need people who are good at using these tools. That's it's just a gateway to the transition that ultimately like kills a ton of jobs. So yeah, I would say it's noise in that this is a near term transition period where this data might be true, but it's not. It's not, it doesn't have any long term sort of signal to it.
[15:04]
D
I disagree. I'm going to go on the signal side. I think that as it's just Javon's paradox where as a cost of many jobs reduces, there will be increasing demand and it's going to be better to, or actually economical to enter a lot of industries that it wasn't previously. I also think that as AI and robotics, et cetera, that we continue to see improve over time. There will always be bottlenecks, whether it's in the like digital or physical world. And I don't think that AI, robotics and everything will be good enough to create a recursive improvement layer where there is no human that is helpful at any stage within the process to unstick some of the bottlenecks. But I do also think that, yes, it's saying that it's there are AI native people. And I think that it's the distinction here is just those people who are willing to take autonomy and actually responsibility and learn the tools will become more valuable. Whereas there are some people who are just like, fuck it, I don't need to learn about AI, I'm going to just retire now.
[16:14]
A
Yeah, I'm going to go with signal as well in the sense that a little bit different than Lee have described. But at the end of the day there's not that many people who are AI native. And so the signal here really is that if you are on the cutting edge of this thing, like there's going to be Immense opportunity for you. These companies are going to try to hire back people. But if you've ever seen the data, it really speaks to the fact that about 2% of people even have a paid subscription to any of these services. And let alone of that 2%, it's just a fraction of those that actually are using it for any sort of productivity gains or efficiencies or increasing their output at work. I would imagine a lot of people who have access to these tools and pay for them are asking questions like where should I go to eat for lunch? Or could you plan my weekend trip? And like there's very little in terms of actual output. And it may feel like that's not the case because if you spend time online on X, you're seeing essentially like the 0.01% of people who are on the frontiers of using these technologies. And you're also seeing a lot of LARPing. There's just a lot of people like pretending that they're doing all these things, but they make content about it and they make the money from the content, not from actually using the tools. Right. So I would say at the end of the day, the other data point to pay attention to is the fact that in October of last year, it was the largest month of Layoffs in over 20 years. It was like I think 23 years or 22 years from October of 2020, 2003 to October of 2025. So look, there's opportunity for those who want, who want to take it. But yeah, what do you have, Michael?
[17:42]
C
Just a quick thing is I think that the diffusion like those people larping is the tip of the spear and how hard it is to actually use it because they're LARPing, but they actually haven't produced value. Think about how many people listening or even on here that we create docs and all these things, they never get like actually come to fruition or the hardest thing is getting a dollar from the individual user. And to Liam's point, I think the Jevons paradox works in most examples except for intelligence, because intelligence will come across everything, including robotics, for other things. We've never seen that before. And so the diffusion from this point into the next couple of years will take too long for those people majority to capture and then have an impact at a firm. And because it's moving so fast that puck keeps moving, that's why the diffusion gets harder. But then ultimately it gets on the other side where a lot of jobs and what you thought you learned will ultimately just become irrelevant. And so I think this is a lot bleaker and just crazier than I think what was initially expected or just the takes that I've had on. You can get ahead of this. Like you can get ahead of it, but that was like six months to 18 months ago. Because it. It only compounds eventually gets too far ahead of you and it's going to be weird.
[18:50]
B
Well, yeah. The other thing just to mention is I think we're kind of talking about two different things in the sense that people's ability and willingness to use the tools is a different conversation than robots replacing factory line workers. Liam, the point you're making, you have to then assume that each robot on a line needs one person sort of managing it. Again, I just don't think that's true over a longer time horizon.
[19:16]
D
I just don't think that the robots make the robots. I think that there is huge.
[19:21]
C
The robots are definitely gonna make the robots in the same way they're working on code to make more code for research. They're going to automate themselves. And then the other extent is like from a sales perspective, eventually you're going to have one head of sales and he's just going to have all of his agents trained and he's gonna be able to make the calls go outbound and those 10 roles are gone. You're just gonna have one. And what are those other 10 people gonna do?
[19:38]
D
Like, I think that there are going to be more roles as they will have more opportunity to sell things if it's as products become easier to make. And I think that there will still be bottlenecks and robots creating more robots.
[19:55]
C
I. I agree with all that. I think it goes back to the one out of 10,000 SKUs is like for every one roll that will be increased, there'll be 10,000 laid off. And what are those 10,000 happen? I don't think we've ever seen it at that scale as the disconnect.
[20:08]
A
Yeah, time will tell. I mean, if I'm in that 10,000, find me on Twitter Macrojack21. You can find some good heron content on evening walks. All right, so next topic here to round things out. Agenic AI, the killer use case for blockchain and crypto. So this is from Franklin Templeton, their digital asset department. I haven't had the chance to read this article yet, so I'm going to defer to Michael to kick things off here. Signal or noise?
[20:37]
C
I think I'll go with noise with this particular article because blockchain crypto long term, I Think there's going to be a lot of capital destruction here, but I think the signal is that you're going to naturally need payment flow through all of this stuff we're talking about. And there's no shortage of news headlines that have come out from open router and stripe potential acquisition to what Buzz recently announced. And that's really the, the main meta here is a lot of people at the frontier of AI are saying money won't matter. A lot of notable and that's the thing we have to recognize that money will matter until you coordinate economic activity. But you're going to need a form of money that's native to the Internet to pass through all of these rails. And I think we're going to see this, pick up the agentic SEO and just things people searching has increased. So you're going to hear more of this. But I think it's going to be on us to bring the signal of like what are the long term practical use cases and what are bots going to transact with? And it's not going to be, you know, crypto.
[21:36]
A
Yeah, I mean I would say signal just because I've long thought that this thesis would be validated because for anyone who's been paying attention to like the bitcoin and crypto space for call it five to 10 years now, whether it's been seriously or just on the periphery, I think anyone, if they were five to 10 years ago, would have expected adoption to be way further than it currently is today. And there was kind of an expectation that people would just get this in their hands, they would understand the use cases of bitcoin and, and they would kind of move over to a better form of money. Unfortunately, that's just the not the reality for a number of reasons. People are busy, they may have obligations that prevent them from digging into it. They may be lazy, they may be apathetic. Like there's just a number of reasons why people have not adopted at the rates that we've seen. And so I'm actually far more bullish on the agency use case at this point for a broader type of mass adoption. And if the thesis that we just referred to in the last segment plays out, then naturally there's going to be a lot of buying power, there's going to be a lot of activity, there's going to be just a ton going on in that agency world. And so naturally then to Michael's point, well, they're going to kind of not have the baggage that humans have and they're just going to be a use case for something that actually works on their rails. So I think that does end up in stable coins and I think it naturally will end up developing into Bitcoin or potentially a barbell depending on the use case of how these agents are transacting. It probably ends up being a mix of both.
[23:00]
B
I think that's fair near term in terms of what you said there at the end. I think there's not a near term jump to agents, you know, strictly wanting to use bitcoin rails. I think there's this interim period where depending on the use case, like you said, stablecoins may work but I think the longer term trajectory is certainly favors something that is more open, truly decentralized. And the other thing, why like, like you said Michael, like this particular article is noisy to me in the sense that Franklin Templeton is very much on these sort of broader crypto altcoin side of things over the past couple years. And crypto's always looking for the killer use case, right? Like you know, it was fundraising with ICOs and it was NFTs and it was, you know, the next thing. And so now they're grafting onto this. You know, it's also the tokenization aspect of all the, you know, real world asset kind of stuff. And so they're always looking for the killer use case. So that element of it I think is noise. But you can see the longer term trajectory here where yes, there's immense value to be had through these types of rails where, you know, micropayments and things become economical on these rails where they're not in the traditional system. So there's natural, there's a natural trend there. But I think the, the signal is really the long term trajectory towards bitcoin.
[24:17]
D
Yeah, I think this is mostly noise. I think that people aren't going to give like people are controlling agents at this point for the most part and they're not going to give them significant amounts of money in the near term. And so I think that the amount of money that they're saving using stable coins etc is minimal. And for the average person, if they came back and their agent told them that they lost money on bitcoin, most of them would be like, what is going on here? I do think that over a long that the signal here is that just AI is going to allow for significantly better product experiences, clis agent onboarding so that people who do actively want to use Bitcoin and stablecoins will be able to do it in a better, quicker, more safer and more cost effective way. But I am bearish on you know just giving agents money and then picking stable coins and bitcoin without really having the ultimate end user. Want that?
[25:25]
A
That's fair. You have thoughts, Michael?
[25:28]
C
No, I was going to say let's rip the the round table. Do you want me to. You want me to start?
[25:32]
A
Yeah, please do know we're time constraint introduce what it is real quick.
[25:37]
C
So rapid fire. Just us going around the horn on anything we found interesting high signal. Whether it's notable news that we didn't cover or content. The news action item is the buzz announcement recently by Jack Dorsey. I think that there's a lot there you'll be hearing from us from about but I think we need to figure out a way to make this a little more interactive around feedback and also the topics we cover. And it feels like there's something there for us to do with the signal versus noise audience. I say there's like 100 people in here right now. We're just starting out 109 now. But it be interesting whether it's in the live past or pre. We can put the topics and then see what gets upvoted. So we'll work through that. If you have any thoughts or feedback, share them with us so we can get that rolled out asap.
[26:29]
B
I like it. I was going to share a piece of research or content I guess came out about a month or maybe two months ago. But from our friends at Incrementum which is a gold and bitcoin fund based in Liechtenstein, they put out an annual report called the In Gold We Trust report. You can find it at ingoldwetrust.report and it's 500 pages of just jam packed signal on gold markets. A bit on bitcoin but really also about like debt, deficits, inflation, some incredible charts in there. So if you're not familiar with the Incrementum guys, definitely check out that report. Like I said, they put out that sort of gold overview, 500 pages, a ton of data once a year and so that came out recently. So check that out.
[27:22]
D
Nice.
[27:23]
A
Yeah, that sounds great. I love the research they produce. Brian, I'm assuming you didn't print it out, rip the spine off and then give it to Anthropic. Did you?
[27:30]
B
Did not.
[27:31]
A
Okay, good stuff. Mine real quick would be from our friend Marty Bent. He published something last week. I think this is such an important concept that maybe a lot of our listeners understand and we obviously talk quite a bit about it. But it is still knowledge that people don't don't know. And I think it's very consequential that people figure it out. And so what it is is he published something on the topic of Gen Z not growing up under a capitalist society. It's grown up under this like political managed credit system, right? The fiat economy. And I think people should check that piece out and maybe try to find a way to either share it with friends and family or at least put it in maybe more layman's terms or things that are going to be appreciated by people in your family. Because it is unfortunately a big problem. I mentioned that, you know a lot's going to lead to this 2028 election and unfortunately the biggest problem is sitting right in front of us and virtually nobody understands it. In terms of the broken money,
[28:28]
D
I completely agree. Lastly, from me, I thought that the CEO of Deep Seeks interview that got leaked last week was there's some signal in there. I think that he's probably talking his buck a little bit just given the fact that you know he they probably knew it was going to leak and he was probably being a little bit too overly optimistic on AI in general. But the thought that the part that I really thought was signal is just the fact that they have know very smart capable people who are working under almost no min hierarchy, KPIs etc. They're just self motivated and willing to be creative and you know, when active people are aligned on the similar goals there's a lot of a lot that can be accomplished. So I'd check that out if you
[29:16]
C
haven't real quick just for Brian and the guys. This is completely random but last night I set up these different agents and buzz and have different APIs are hitting like it was Fable and Claude or open AI but I had it until like 10:30. You argue on who the the best basketball player ever exists is and take a different position to come to objective answer. And so they started going back and forth. I think it was three of them creating the rubric and then they spotted like a gap. And so I think this is just like an easy example. Obviously they don't need to talk about bitcoin basketball but you can imagine having different agents tied into different models and then ultimately coming down and like battle testing, curating like whatever outcome object. This is super innocent but just an example of like the insanity because they were able to come through like a gap.
[30:04]
B
Did they land anywhere?
[30:06]
C
Well so they built me a website or multiple websites too for like what how you do it that were interactive but I think they came out with Jordan, but it was based on like certain hero 60. You can see these like local websites we have.
[30:18]
B
We haven't reached ASI yet if they're landing on Jordan, but it's pretty cool, right?
[30:23]
C
Because you just kind of. It's just like. It's basically boundless on the limitations of like what you can do here.
[30:32]
A
It's pretty crazy chat. Is this an AI native builder? Let us know in the comments. I will say thank you for being here. If you joined live or catch the recording, my ask would be if you made it to the end of the show, please follow along, subscribe like comment Rate 5 stars, whatever that may be. My job seriously depends on it. I'm hanging on by a thread right now and more likely than not, I'm in the 10,000 that's going to be is going to be terminated. So please help me hang on as long as I can. My family depends on it.
[31:00]
D
Thank you.
[31:01]
C
The droids are starting. The droids are starting to move faster than Jackson.
[31:04]
A
Thank you for being here. We'll catch you on Thursday. Thursday, yeah, we'll see you then.

Summary6 min read

Podcast Summary: Onramp Bitcoin Media – "What Anthropic Actually Did to Millions of Books"

Date: July 29, 2026
Episode Theme:
This episode unpacks the viral controversy surrounding Anthropic’s book-scanning practices and explores the broader implications for AI training, intellectual property, and data sovereignty. The hosts engage in a “signal vs. noise” debate, examining whether the outrage is justified, then pivot into macroeconomic themes including food inflation, AI-native job markets, and the intersection of AI agents with crypto payments.

Main Theme

The hosts address the uproar over court revelations that Anthropic physically destroyed and digitized millions of print books for AI training. Their discussion explores fact versus sensationalism, the philosophical and practical consequences for information preservation, and how this controversy reflects larger shifts in technology and markets.

Key Discussion Points & Insights

1. Anthropic's Scanning of Books: Fact vs. Hype

(00:00-06:15)

Viral Reaction:
The news exploded online with accusations that Anthropic destroyed “rare” or “ancient” books to create digital training data for Claude, paralleling Orwellian fears of knowledge erasure.

“People were being a little bombastic... insinuating that Anthropic was destroying ancient texts... turns out, these weren’t super valuable one-of-a-kind texts.”
— B (00:49)
Clarifications & Nuance:
The books scanned were largely obscure, not irreplaceable cultural artifacts. Physical destruction was a byproduct of speed: it’s faster to scan unbound pages.

“The reason they’re destroying them isn’t necessarily super nefarious—it’s just so they can scan them faster.”
— B (01:40)
Concerns about Information Loss and Social Change:
The panelists reflect on the deeper pattern: digitization can mean the loss of physical media, risking gaps in the historical record, especially as reliance on digital memory (via AI) increases and traditional reading declines.

“As we grow up, and our kids grow up, it’s going to be this foreign world to pick up a physical book... it should scare the living crap out of everyone because what these inputs, these guardrails, are going to do.”
— C (02:48)
Hypocrisy in the AI World:
Irony noted—Anthropic objects to data distillation practices abroad while engaging in aggressive corpus-building themselves.

2. Food Price Inflation: Signal or Noise?

(06:15-12:47)

Sharp Food Inflation and Public Anxiety:
U.S. grocery prices have risen 33% from 2019 to 2026—biggest increase in 50 years, driving decreased consumption and fueling political debates about economic instability.

“If you’re eating real food, [inflation is] more likely higher than 33%... consumers are buying less food overall.”
— A (06:15)
K-Shaped Recovery and Liquidity Effects:
Discussion of monetary policy, social stratification, and the illusion of stable inflation due to manipulated CPI.

“This is a massive signal... it pierces the veil of manipulated CPI... if you were tracking high-quality meats, the line would be even more aggressively upward.”
— B (09:18)
Emergence of State-Run Grocery Experiments:
New York’s state-owned pilot stores seen as reactions to deep-rooted problems, with skepticism about their viability and impact.

3. AI-Native Hiring: Real Trend or PR Spin?

(12:47-19:54)

Is “Hiring for AI Skills” a Real Trend?
Wall Street Journal headlines about increased AI-native hiring are seen by some as “gaslighting”—while layoffs outnumber hires and are driven by broader economic trends, not just AI.

“This is just gaslighting... people are getting laid off, not because of AI, but because of interest rates and reduction in people’s capital outlay.”
— C (13:09)
Transitional Dynamics:
While marginal demand rises for AI-skilled workers, the long-term picture is one of massive displacement and job destruction, especially as AI/robotics mature.

“It’s just a gateway to the transition that ultimately kills a ton of jobs... we haven’t hit the tipping point yet.”
— B (13:52)
Opposing View – Jevons Paradox and New Roles:
Some see opportunity in AI-driven productivity, arguing cost reductions can spur new industries and job types, though others say the scale will overwhelm traditional “paradox” effects.

“As cost of many jobs reduces, there will be increasing demand... I don’t think that AI, robotics, will be good enough to create a recursive improvement layer where there is no human, at any stage in the process.”
— D (15:03)

4. Agentic AI and Crypto: Real Use Case or New Hype?

(20:07-25:25)

Franklin Templeton’s “Agentic AI” Blockchains Report
The idea: As autonomous agents proliferate, they’ll need internet-native money; could bitcoin or stablecoins become the rails for agent-to-agent payments?
Skepticism about Near-term Adoption:
Many see this as another crypto use-case narrative search, noting barriers to real adoption and the long ramp for agents to handle significant money.

“I’m bearish on... giving agents money and picking stablecoins and bitcoin without the ultimate end user wanting that.”
— D (24:16)
Long-term Trajectories:
Some bullishness remains on bitcoin, especially as agents bypass human baggage and facilitate seamless microtransactions.

“There’s a natural trend... the signal is really the long-term trajectory towards bitcoin.”
— B (22:59)

Notable Quotes & Memorable Moments

A on Sensationalism:
“There’s some ancient manuscripts on men giving birth to children that have been destroyed and since have been gatekept by Anthropic. So that’s a big problem we need to solve for.”
— A (05:57) (tongue-in-cheek nod to viral misreporting)
C on Data Loss:
“As a child I remember just always liking to read... I’m trying to get back because... reading and writing really help with... cognitive ability, especially if you’re outsourcing some of it to AI.”
— C (02:48)
A on the New “Normal” in Food:
“It’s so easy for me to think about what prices used to be... could also be challenging… having essentially amnesia about what things used to cost.”
— A (11:51)
B on Crypto Killer App Cycles:
“Crypto’s always looking for the killer use case... fundraising with ICOs, NFTs... now they’re grafting onto this [agentic AI].”
— B (22:59)

Timestamps for Key Segments

Book Scanning Outrage: 00:00 – 06:15
Food Price Inflation and Macro Trends: 06:15 – 12:47
AI-Native Hiring Debate: 12:47 – 19:54
Agentic AI & Crypto Payment Rails: 20:07 – 25:25
Rapid-Fire Roundtable (“Signal” Content Recs, Tools): 25:27 – 31:04

Roundtable: Rapid-Fire Content & Signal Picks

Community Engagement Suggestions:
C proposes interactive topic voting and more feedback integration for future episodes.
Recommended Reads:
- Incrementum’s “In Gold We Trust” report – a 500-page gold and macro market deep-dive (ingoldwetrust.report)
  — B (26:28)
- Marty Bent’s piece on Gen Z’s economic upbringing under managed credit vs. capitalism — A (27:31)
- CEO of Deep Seek’s leaked interview on organizational freedom and innovation in AI startups — D (28:28)
Fun Moment:
C describes programming multiple AI agents to debate the greatest basketball player of all time (Claude, OpenAI, Fable), which culminates in a Jordan win and several self-built websites.

“It’s just like... boundless on the limitations of what you can do here.”
— C (30:22)

Overall Tone & Takeaways

Engaged, irreverent, occasionally skeptical—hosts balance macro seriousness with humor and in-group banter.
Emphasis on separating “signal” (real, actionable, systemic change) from “noise” (hype, spin, and surface-level narratives).
Recurrent themes: trust in traditional data is wavering, technological shifts are accelerating but uneven, and Bitcoin/crypto remains a core lens for interpreting macro and tech trends.

For listeners:
If you want a clear-eyed, critical approach to viral tech narratives—well-grounded in the intersections of money, AI, and information—the Onramp team delivers, blending current events with deeper system-level analysis.

Loading summary

Transcript47 lines

[00:00]
A
Anthropic destroys millions of books. So this one has been going uber viral across all the different feeds for the past 24 hours, maybe 36 hours. I don't know when exactly the news broke, but the general gist here, as you can see, that court documents revealed that AI company Anthropic spent millions of dollars physically scanning print books to build Claude an AI assistant similar to ChatGPT. You know, if you're, if you're not aware of that. In the process, the company cut millions of print books from their bindings, scanned them into digital files, and threw away the original solely for the purpose of training AI. Details buried in a copyright ruling on fair use, whose broader fair use implications were reported yesterday. So, gentlemen, who would like to go first on this topic? It's, it's an interesting one. It's a little bit different than what we've covered on the show in the past couple weeks here. Who wants to give a first take on signal versus noise today?
[00:49]
B
I'll, I'll go first. And maybe it's a little unexpected, but I do think this is actually mostly noise in the sense that I think it was, people were being a little bombastic in terms of how it was being reported and how it really went viral in terms of people basically insinuating that Anthropic was destroying ancient texts from hundreds or thousands of years ago, that there's only a few versions of, that they're rare, and somehow they're now destroyed forever. Upon further review, that isn't exactly what was happening. When they say rare books, it's not that they're ancient texts, it's that they're very, like, obscure. And so, yes, there's some like, you know, 1984 esque vibes around this entire thing of like, you know, they need to canvas the corpus of, you know, human intelligence and destroy and sort of like, you know, pull up the ladder behind them on the intelligence, and it no longer lives in physical form and it's, it's only in anthropic coffers. And there's also a natural hypocrisy with them complaining and whining about all these distillation efforts on the other end of their models from China and other places where they're doing a form of distillation themselves with all of these books and texts. But I do think it was kind of inaccurately reported based on the court findings of these weren't super valuable one of a kind text that they were destroying to scan, basically. The other nuance here is that the Reason they're destroying them isn't necessarily like super nefarious. It's just like they can scan them faster that way. And then Elon, I think reported or replied to somebody's viral tweet about this being like, we're doing this the hard way, the slow way, where you don't basically rip the binding apart so you can scan it faster. He's like, when we do this, we do it the hard way because it's more moral.
[02:48]
C
Yeah. I mean, I think I'm always skeptical when something goes insanely viral. I'm like, why? Why, why it did. So I think from the angle of what's anthropic doing, how true is it? I think there's noise in that respect. Where I think there's an insane amount of signal is ultimately just the notion of books and there's like in data and information and it being wiped away. There's no shortage of projects. I wish I could. The names are escaping me right now on codifying and capturing the legacy Internet because it's increasingly becoming harder and harder to find. But then as you go further, you think about old text news, information. What made it through the test of time has historically been very few articles, objects, books. And you add to the notion of the idea I think it came out with around literacy decreasing more utilization of these tools, AI. And then I think there's direct analysis, empirical evidence from like physicians around the brain and how it operates specifically from like memory, cognitive ability when you leverage these tools. The example is like spatial data, I think when it comes to maps and sense of direction. And so I've always just had these things in the back of my mind around data and like the information for the Internet. But it also like as a child I remember just always liking to read and I've gotten away from it. I'm trying to get back because I don't know how true this is. But reading an X number of pages or writing really help with that like cognitive ability, especially if you're outsourcing some of it to AI. And I think that as we grow up and our kids grow up, it's just going to be this foreign world to pick up a physical book and read it. And this is just one small manifestation of that signal and it should scare the living crap out of everyone because what these inputs. Remember during COVID when Google told you if can a man when this was super woke time, but can a man have a baby? And it would say yes. And that's the world I think we're definitely Going to. In the sense of not a man having a baby, but the books will tell you. And we see this right now with like the open models and people liking them because the guardrails aren't there. So I think this is just the precursor to what's to come.
[05:04]
D
Yeah, I'll go next. I think that there's a number of different angles here. I'm gonna go with one of them being signal. And not just the actual like burning of books and destroying them, but the fact that Anthropic is trying to train on unique information that isn't widely available to everybody. I think that over time there's going to be more and more information that is on easily accessible through models. And the web is. There just continues to be more information published by agents. And having the sovereignty over your data workflows and specialized earned secrets for companies and your own way that you do things is going to have a lot more value. And that's why I think that increasingly it's becoming almost untenable to leak a lot of these enterprise secrets to frontier AI that are using it to train off their own data.
[05:57]
A
We're a little bit pressed for time here, but I would just say I know there's some ancient manuscripts on men giving birth to children that have been destroyed and since have been gatekept by Anthropic. So that's a big problem that we'll need to solve for.
[06:13]
C
It's a good one, Jackson.
[06:16]
A
All right, let's. We gotta move forward. We have. We have to respect the timer here. So everyone just be mindful of your comments, please. So next one would be grocery store. Grocery prices experienced the biggest jump in 50 years. The headline does a lot of the talking. But Nick, if you could pull up the chart as well. This is really what caught a lot of people's attention online this morning. Where you can see here, if you're not following on YouTube, you can see this is a line chart from starting in 2012 to present in 2026 and showing that the price of food purchased for homes rose 33% from 2019 through June 2026. So call it seven years 30 up 33%. I would argue if you're eating real food, more likely higher than 33%. And so this says consumers are now buying fewer of their favorite foods and less food overall. This actually reminds me, I guess I'll go first because I didn't have a chance on the last one. This reminds me, I think it was from 2023. Wall Street Journal published a piece and the headline was essentially you should skip breakfast because you want to save money on your food prices. And so this is where unfortunately a lot of things are going right where I think that there's a reality that we have an incredibly K shaped economy. This is a term that was coined out of 2020 in particular, where you had all this liquidity go to those who had access to the money spigots. You had zero interest rate policy, you had negative interest rates in some countries. And so naturally there was an outlier in terms of where all the asset prices sit today and those who have fallen behind. And of course, when you inject trillions of dollars of new money, right, liquidity just means new money coming into the system. And this is not just a US problem. Naturally you're going to have a chart that looks like this. I mean the Trend line from 2012 to 2019 is pretty much unmoved in terms of price, food inflation and then it goes vertical in the past seven years. And this is a big problem. It ties into what we could also talk about with New York piloting out state owned grocery stores, which I'm just not surprised to see. I mean this is going to be the biggest trend, I think heading into the 2028 election. There's just a growing populace of people who are really outraged and I can appreciate to a degree why. But unfortunately most people have zero idea what the root cause of this is. So there's a big signal to pay attention to, of course, because if food prices began become unstable, then there's a lot of issues that are downstream of that.
[08:30]
C
Yeah, I mean this is highest amount of signal we could show. And it's why I personally get frustrated every time we talk about the Fed speak aspect and then even just like what's going on with Iran on, off, it doesn't matter. Food prices are going to increase no matter what. The key stat is, you know, American households, the richest country on earth, reducing food consumption while GDP and earnings print records like that's, that's the notion of what we talk about, that's the real versus nominal. And there's no shortage of gaslighting around 2, 3, 4% when it's substantially higher. And it's the only input that matters. If you can't feed yourself, like what else? And then the worst part, and this is like a little esoteric, is this is all like slop. The general food that's being fed to people is just like getting worse. And, and so anyway, this is, this is not good.
[09:19]
B
Yeah, agree with all that. This is a massive signal in the sense that it pierces the veil of manipulated cpi. So anytime you look at the actual underlying inputs to CPI and sort of sift through the manipulation of that sort of endpoint data, you start to see what's actually going on. And to your point, Michael, this isn't even the full picture. If you instead were tracking high quality meats or other foods that you know, you could go get at a farmer's market or something that's like not as slobified then that, that line would be even more aggressively upward sloping in terms of the price of all of these things. And so it's a massive signal and it's, it's a nice reminder of the Fed speak of the manipulated data that is cpi, because this, this is out of control clearly. And then the, the natural foil, as you mentioned, Jackson, is that, you know, Mamdani in New York City is going to try to just basically price fix and have these grocery stores that will sell out immediately and they'll have to figure out some sort of like lottery system of who even gets to go. And it'll lose the, you know, lose the state and city a ton of money because grocery margins are probably the thinnest across like any industry. Pretty much like you know, 2% margins generally around there. And so if you're undercutting prices by 30%, you're losing a ton of money. So it doesn't make any sense. But like, like you said, Jackson, like it's not a totally unsurprising or not a totally surprising reaction to where we are and people don't know the cause and so they look for this solution that doesn't really fix anything, just sort of papers over it.
[11:04]
D
Yep, it's the, the massive signal here is essentially everything was reset at a significantly higher base after 2020, 2021, when many monetary units entered the system. And things are only going to accelerate here in terms of the money printing and then in addition to that where we have a increasingly globally fractured world and higher energy prices, which are a key input to pretty much everything and so should only go higher from here as well as just the increased regulation, whether it's New York City trying to get involved with grocery stores that should prevent from finding the correct clearing prices. And so yeah, there's, there's a lot of signal in here and don't anticipate this getting any better anytime soon.
[11:51]
A
Yeah, Liam, I think what you said toward the end there is the new base, right? I mean it's so, it's so easy for me to think about what prices used to be. And in fact, it can be challenging. I know it's so easy, but it could also be challenging to an extent of like having essentially amnesia about what things used to cost. I posted something maybe a week or two ago about. I forget if it was food prices or something else. And people were in my comments saying, like, why do you care what things cost in 2017? And it was a sandwich shop that I went to like 10 years ago and what their prices were and they have since doubled. It's just insane. Like, nobody understands what, what's going on here.
[12:25]
C
And honestly, I would make anybody a bet this is much higher. It's not reflected because of shrinkflation, because there's no shortage of things that you see, whether it's like toilet paper or paper towels that have like reduction implies or number of wraps or whatever, or the number of things in whatever you're buying or the number of liters or fluid ounces in the. In the deal. So it's much higher. And that's how you mask a lot of this.
[12:48]
A
Yeah, for sure. All right, so the next topic, Hiring is accelerating for AI native skills. So I'll read the headline here from WSJ. Big companies are starting to hire again, defying predictions of an AI wipeout. I mean, this is great for me to hear, of course, because I thought that I was already in the permanent underclass, but now maybe I have a shot. But who wants to take a first crack at the news here?
[13:10]
C
This is noise. This is just gaslighting. Look, so yes, people are hiring for people that are AI native, but it's the equivalent of putting a notice out that food prices are reducing. I can guarantee you if there's 10,000 SKUs at your local supermarket, you might find one that reduced. But while 99% of them are increasing, this is just gaslight. And like, people are getting laid off. It's not because of AI, they're using it as an example, but people increasingly get laid off because of interest rates and reduction in people's capital outlay because of real estate or food. They just spend less. And if they spend less, then there's less consumer demand for your products and services that you have to lay people off or you go out of business. It's just a recursive downward loop.
[13:53]
B
I would agree. I think this is mostly noise in the sense that you have to recognize that we're in this very unique, interesting sort of transition period where, yes, on the margin there's going to be incremental new hires for people with AI native skills. But to your point, Michael, those are going to be massively outweighed by the amount of displacement and job loss. We're just scratching the surface of it. I don't think we've really seen it materially. Like all the layoffs we've seen I don't think are directly related to people just being replaced by AI, but you're starting to see small slivers of it. And we haven't even hit the tipping point of like real robotics boom where in terms of like actual physical labor, there's going to be massive, massive displacement. And so while this might be like a little near term blip of like, yeah, okay, big companies realize that they need people who are good at using these tools. That's it's just a gateway to the transition that ultimately like kills a ton of jobs. So yeah, I would say it's noise in that this is a near term transition period where this data might be true, but it's not. It's not, it doesn't have any long term sort of signal to it.
[15:04]
D
I disagree. I'm going to go on the signal side. I think that as it's just Javon's paradox where as a cost of many jobs reduces, there will be increasing demand and it's going to be better to, or actually economical to enter a lot of industries that it wasn't previously. I also think that as AI and robotics, et cetera, that we continue to see improve over time. There will always be bottlenecks, whether it's in the like digital or physical world. And I don't think that AI, robotics and everything will be good enough to create a recursive improvement layer where there is no human that is helpful at any stage within the process to unstick some of the bottlenecks. But I do also think that, yes, it's saying that it's there are AI native people. And I think that it's the distinction here is just those people who are willing to take autonomy and actually responsibility and learn the tools will become more valuable. Whereas there are some people who are just like, fuck it, I don't need to learn about AI, I'm going to just retire now.
[16:14]
A
Yeah, I'm going to go with signal as well in the sense that a little bit different than Lee have described. But at the end of the day there's not that many people who are AI native. And so the signal here really is that if you are on the cutting edge of this thing, like there's going to be Immense opportunity for you. These companies are going to try to hire back people. But if you've ever seen the data, it really speaks to the fact that about 2% of people even have a paid subscription to any of these services. And let alone of that 2%, it's just a fraction of those that actually are using it for any sort of productivity gains or efficiencies or increasing their output at work. I would imagine a lot of people who have access to these tools and pay for them are asking questions like where should I go to eat for lunch? Or could you plan my weekend trip? And like there's very little in terms of actual output. And it may feel like that's not the case because if you spend time online on X, you're seeing essentially like the 0.01% of people who are on the frontiers of using these technologies. And you're also seeing a lot of LARPing. There's just a lot of people like pretending that they're doing all these things, but they make content about it and they make the money from the content, not from actually using the tools. Right. So I would say at the end of the day, the other data point to pay attention to is the fact that in October of last year, it was the largest month of Layoffs in over 20 years. It was like I think 23 years or 22 years from October of 2020, 2003 to October of 2025. So look, there's opportunity for those who want, who want to take it. But yeah, what do you have, Michael?
[17:42]
C
Just a quick thing is I think that the diffusion like those people larping is the tip of the spear and how hard it is to actually use it because they're LARPing, but they actually haven't produced value. Think about how many people listening or even on here that we create docs and all these things, they never get like actually come to fruition or the hardest thing is getting a dollar from the individual user. And to Liam's point, I think the Jevons paradox works in most examples except for intelligence, because intelligence will come across everything, including robotics, for other things. We've never seen that before. And so the diffusion from this point into the next couple of years will take too long for those people majority to capture and then have an impact at a firm. And because it's moving so fast that puck keeps moving, that's why the diffusion gets harder. But then ultimately it gets on the other side where a lot of jobs and what you thought you learned will ultimately just become irrelevant. And so I think this is a lot bleaker and just crazier than I think what was initially expected or just the takes that I've had on. You can get ahead of this. Like you can get ahead of it, but that was like six months to 18 months ago. Because it. It only compounds eventually gets too far ahead of you and it's going to be weird.
[18:50]
B
Well, yeah. The other thing just to mention is I think we're kind of talking about two different things in the sense that people's ability and willingness to use the tools is a different conversation than robots replacing factory line workers. Liam, the point you're making, you have to then assume that each robot on a line needs one person sort of managing it. Again, I just don't think that's true over a longer time horizon.
[19:16]
D
I just don't think that the robots make the robots. I think that there is huge.
[19:21]
C
The robots are definitely gonna make the robots in the same way they're working on code to make more code for research. They're going to automate themselves. And then the other extent is like from a sales perspective, eventually you're going to have one head of sales and he's just going to have all of his agents trained and he's gonna be able to make the calls go outbound and those 10 roles are gone. You're just gonna have one. And what are those other 10 people gonna do?
[19:38]
D
Like, I think that there are going to be more roles as they will have more opportunity to sell things if it's as products become easier to make. And I think that there will still be bottlenecks and robots creating more robots.
[19:55]
C
I. I agree with all that. I think it goes back to the one out of 10,000 SKUs is like for every one roll that will be increased, there'll be 10,000 laid off. And what are those 10,000 happen? I don't think we've ever seen it at that scale as the disconnect.
[20:08]
A
Yeah, time will tell. I mean, if I'm in that 10,000, find me on Twitter Macrojack21. You can find some good heron content on evening walks. All right, so next topic here to round things out. Agenic AI, the killer use case for blockchain and crypto. So this is from Franklin Templeton, their digital asset department. I haven't had the chance to read this article yet, so I'm going to defer to Michael to kick things off here. Signal or noise?
[20:37]
C
I think I'll go with noise with this particular article because blockchain crypto long term, I Think there's going to be a lot of capital destruction here, but I think the signal is that you're going to naturally need payment flow through all of this stuff we're talking about. And there's no shortage of news headlines that have come out from open router and stripe potential acquisition to what Buzz recently announced. And that's really the, the main meta here is a lot of people at the frontier of AI are saying money won't matter. A lot of notable and that's the thing we have to recognize that money will matter until you coordinate economic activity. But you're going to need a form of money that's native to the Internet to pass through all of these rails. And I think we're going to see this, pick up the agentic SEO and just things people searching has increased. So you're going to hear more of this. But I think it's going to be on us to bring the signal of like what are the long term practical use cases and what are bots going to transact with? And it's not going to be, you know, crypto.
[21:36]
A
Yeah, I mean I would say signal just because I've long thought that this thesis would be validated because for anyone who's been paying attention to like the bitcoin and crypto space for call it five to 10 years now, whether it's been seriously or just on the periphery, I think anyone, if they were five to 10 years ago, would have expected adoption to be way further than it currently is today. And there was kind of an expectation that people would just get this in their hands, they would understand the use cases of bitcoin and, and they would kind of move over to a better form of money. Unfortunately, that's just the not the reality for a number of reasons. People are busy, they may have obligations that prevent them from digging into it. They may be lazy, they may be apathetic. Like there's just a number of reasons why people have not adopted at the rates that we've seen. And so I'm actually far more bullish on the agency use case at this point for a broader type of mass adoption. And if the thesis that we just referred to in the last segment plays out, then naturally there's going to be a lot of buying power, there's going to be a lot of activity, there's going to be just a ton going on in that agency world. And so naturally then to Michael's point, well, they're going to kind of not have the baggage that humans have and they're just going to be a use case for something that actually works on their rails. So I think that does end up in stable coins and I think it naturally will end up developing into Bitcoin or potentially a barbell depending on the use case of how these agents are transacting. It probably ends up being a mix of both.
[23:00]
B
I think that's fair near term in terms of what you said there at the end. I think there's not a near term jump to agents, you know, strictly wanting to use bitcoin rails. I think there's this interim period where depending on the use case, like you said, stablecoins may work but I think the longer term trajectory is certainly favors something that is more open, truly decentralized. And the other thing, why like, like you said Michael, like this particular article is noisy to me in the sense that Franklin Templeton is very much on these sort of broader crypto altcoin side of things over the past couple years. And crypto's always looking for the killer use case, right? Like you know, it was fundraising with ICOs and it was NFTs and it was, you know, the next thing. And so now they're grafting onto this. You know, it's also the tokenization aspect of all the, you know, real world asset kind of stuff. And so they're always looking for the killer use case. So that element of it I think is noise. But you can see the longer term trajectory here where yes, there's immense value to be had through these types of rails where, you know, micropayments and things become economical on these rails where they're not in the traditional system. So there's natural, there's a natural trend there. But I think the, the signal is really the long term trajectory towards bitcoin.
[24:17]
D
Yeah, I think this is mostly noise. I think that people aren't going to give like people are controlling agents at this point for the most part and they're not going to give them significant amounts of money in the near term. And so I think that the amount of money that they're saving using stable coins etc is minimal. And for the average person, if they came back and their agent told them that they lost money on bitcoin, most of them would be like, what is going on here? I do think that over a long that the signal here is that just AI is going to allow for significantly better product experiences, clis agent onboarding so that people who do actively want to use Bitcoin and stablecoins will be able to do it in a better, quicker, more safer and more cost effective way. But I am bearish on you know just giving agents money and then picking stable coins and bitcoin without really having the ultimate end user. Want that?
[25:25]
A
That's fair. You have thoughts, Michael?
[25:28]
C
No, I was going to say let's rip the the round table. Do you want me to. You want me to start?
[25:32]
A
Yeah, please do know we're time constraint introduce what it is real quick.
[25:37]
C
So rapid fire. Just us going around the horn on anything we found interesting high signal. Whether it's notable news that we didn't cover or content. The news action item is the buzz announcement recently by Jack Dorsey. I think that there's a lot there you'll be hearing from us from about but I think we need to figure out a way to make this a little more interactive around feedback and also the topics we cover. And it feels like there's something there for us to do with the signal versus noise audience. I say there's like 100 people in here right now. We're just starting out 109 now. But it be interesting whether it's in the live past or pre. We can put the topics and then see what gets upvoted. So we'll work through that. If you have any thoughts or feedback, share them with us so we can get that rolled out asap.
[26:29]
B
I like it. I was going to share a piece of research or content I guess came out about a month or maybe two months ago. But from our friends at Incrementum which is a gold and bitcoin fund based in Liechtenstein, they put out an annual report called the In Gold We Trust report. You can find it at ingoldwetrust.report and it's 500 pages of just jam packed signal on gold markets. A bit on bitcoin but really also about like debt, deficits, inflation, some incredible charts in there. So if you're not familiar with the Incrementum guys, definitely check out that report. Like I said, they put out that sort of gold overview, 500 pages, a ton of data once a year and so that came out recently. So check that out.
[27:22]
D
Nice.
[27:23]
A
Yeah, that sounds great. I love the research they produce. Brian, I'm assuming you didn't print it out, rip the spine off and then give it to Anthropic. Did you?
[27:30]
B
Did not.
[27:31]
A
Okay, good stuff. Mine real quick would be from our friend Marty Bent. He published something last week. I think this is such an important concept that maybe a lot of our listeners understand and we obviously talk quite a bit about it. But it is still knowledge that people don't don't know. And I think it's very consequential that people figure it out. And so what it is is he published something on the topic of Gen Z not growing up under a capitalist society. It's grown up under this like political managed credit system, right? The fiat economy. And I think people should check that piece out and maybe try to find a way to either share it with friends and family or at least put it in maybe more layman's terms or things that are going to be appreciated by people in your family. Because it is unfortunately a big problem. I mentioned that, you know a lot's going to lead to this 2028 election and unfortunately the biggest problem is sitting right in front of us and virtually nobody understands it. In terms of the broken money,
[28:28]
D
I completely agree. Lastly, from me, I thought that the CEO of Deep Seeks interview that got leaked last week was there's some signal in there. I think that he's probably talking his buck a little bit just given the fact that you know he they probably knew it was going to leak and he was probably being a little bit too overly optimistic on AI in general. But the thought that the part that I really thought was signal is just the fact that they have know very smart capable people who are working under almost no min hierarchy, KPIs etc. They're just self motivated and willing to be creative and you know, when active people are aligned on the similar goals there's a lot of a lot that can be accomplished. So I'd check that out if you
[29:16]
C
haven't real quick just for Brian and the guys. This is completely random but last night I set up these different agents and buzz and have different APIs are hitting like it was Fable and Claude or open AI but I had it until like 10:30. You argue on who the the best basketball player ever exists is and take a different position to come to objective answer. And so they started going back and forth. I think it was three of them creating the rubric and then they spotted like a gap. And so I think this is just like an easy example. Obviously they don't need to talk about bitcoin basketball but you can imagine having different agents tied into different models and then ultimately coming down and like battle testing, curating like whatever outcome object. This is super innocent but just an example of like the insanity because they were able to come through like a gap.
[30:04]
B
Did they land anywhere?
[30:06]
C
Well so they built me a website or multiple websites too for like what how you do it that were interactive but I think they came out with Jordan, but it was based on like certain hero 60. You can see these like local websites we have.
[30:18]
B
We haven't reached ASI yet if they're landing on Jordan, but it's pretty cool, right?
[30:23]
C
Because you just kind of. It's just like. It's basically boundless on the limitations of like what you can do here.
[30:32]
A
It's pretty crazy chat. Is this an AI native builder? Let us know in the comments. I will say thank you for being here. If you joined live or catch the recording, my ask would be if you made it to the end of the show, please follow along, subscribe like comment Rate 5 stars, whatever that may be. My job seriously depends on it. I'm hanging on by a thread right now and more likely than not, I'm in the 10,000 that's going to be is going to be terminated. So please help me hang on as long as I can. My family depends on it.
[31:00]
D
Thank you.
[31:01]
C
The droids are starting. The droids are starting to move faster than Jackson.
[31:04]
A
Thank you for being here. We'll catch you on Thursday. Thursday, yeah, we'll see you then.