Jassy's Shareholder Letter, Frontier Model Rollouts, Data Center Debates, The Next AI Capability

I see a large IPO on the horizon. You're surrounded by journals. Hold your position.

Right. There's misinformation. Clearing order inbound. Let's just roll. We are surrounded by journal. Hold your position. Come on. Get up. Trust the experts here. We are experts. Founder Mel. I see multiple journalists on the horizon. Stand by. UAV online. Blaze. Double blaze. Triple blaze. Double kill. Finally, rock is wins. Team deathmatch. We are experts. Triple blade. Let's just roll. Right. Marky clearing order inbound. Come on, get up. We are surrounded by journalists. Hold your position. Strike one. Strike two. Activate golden retriever mode. Mark clearing order inbound. Five. Founder.

You're watching TV. Today is Thursday, April 9th, 2026. We are live from the TBPN Ultradome, the temple of technology, the fortress of finance, the capital of capital. Uh, we have a great show for you. I was getting a little worried Monday. There was not a lot of fundraises going on. I was looking at the schedule and I was thinking, is it over? And today I'm happy to report that we're back. Uh, Andrew Dye is coming on from Allian. Raised a $55 million seed round. Uh, we got Cassa coming on with a 60 $38 million series B. We got a $170 million series C later and then a hundred million dollar series E and then a nice little mango seed from the Enclave team to to cap it off. But uh really fun show today. We got Sagg and Jetty from Breaking Points, our friend over there coming on. We got Joe Weisenthal. Yeah, wise and tall. Uh and then CZ from Binance. Yeah. uh who is releasing freedom of money detailing Binance's rise to crypto uh evolution and of course his US legal battle memoir and I'm excited to talk to Joe about the who is Satoshi Nakamoto and I'm also excited to talk to CZ about who he thinks Satoshi Nakamoto is. You would imagine that he's in a position to potentially have a a good uh a good take on that.

Uh anyway, there is a ton of uh AI news. Uh Andy Jasse released the uh 2025 shareholders letter. Amazon is known for fantastic shareholders letters dating back to 1997. Um I thought he did a good job of sort of resetting the AI narrative. There's this uh AI lab horse race going on. Of course, we've been covering it all week. Uh anthropics mythos preview and project Glass Wang launched on Tuesday. Quickly followed by news today that OpenAI also plans to deliver a model with advanced cyber security capabilities to key internet infrastructure providers. And there's this debate going on over how and when these models will roll out. Um I think this is going to be an ongoing trend. I don't think cyber security is the last model capability that will be slowly delivered to key companies first. Um, cyber security is a perfect fit for powerful coding agents and people have been digging into exactly how uh some of these uh zeroday exploits, some of these bugs and vulnerabilities were discovered. And uh it makes a ton of sense that if you have a model that's fantastic at coding, it can basically try every single coding exploit, try new coding exploits uh across a huge number of open source packages, submit poll requests, and generally harden the uh internet infrastructure that we all rely on. So in general it seems like the the roll out of uh Mythos although people are disappointed because they want to play with the latest and greatest model even if it's very expensive uh it seems like it's uh it's having a positive effect and uh and should be uh a a bit of a white pill for uh containing powerful models having them have proper impacts having them have a positive impact on the American economy on our security on all these different things.

Um I do think I wouldn't be surprised if we see something similar happen in biosafety. Now the biosafety AI research loop is a little bit longer because you might have to go to the lab. It's not entirely uh existing in a computer in a virtual machine that you can spin up and just uh you know brute force reinforcement learn against. But you could imagine if a model develops capabilities maybe in the next run, maybe in middle of this year, uh if a model becomes powerful enough to design a harmful virus or something like that, uh you would want the lab that creates that model to deliver that to the scientific community and companies that can protect the population against the development of new harmful biological viruses. just like we're protecting the internet against uh cyber security viruses. And so I I I'm I'm not exactly sure where this all goes, what the other 10 steps are, but in general, it seems like there's going to be a pattern of a powerful model becomes capable of doing something. uh that it makes sense to share with the particular community that can defend against that new capability and then make sure and then the the entire community needs to make sure that that capability is carefully under control before releasing a a version of that model that can still you know in the biosafety example like you still want a model that can help you learn about biology learn about uh how viruses are made how they mutate uh this is important for education and augmentation of the of scientist from, you know, high schoolers all the way to professionals. But the most advanced technology, it makes sense to put it in the hands of people that can actually take a real uh have a real impact on that immediately before uh rolling it out broadly.

But uh Andy Josie uh sort of zoomed out and was stepping back from the horse race because Amazon's a partner with basically everyone in the ecosystem uh and has their own models. And so he shared this uh shareholder letter zooming out on the state of AI, the state of uh Amazon's plans and he shares a bunch of very interesting anecdotes about his personal life. So we should read through some of this. Uh he says, "When I graduated from college, I wanted to be a sports cer. Uh after sending my resume to many small markets around the United States and only getting two nibbles, I settled on doing sports production at a major network. To make extra money, I coached my former high school soccer team and worked at a retail golf store. Six months later, a college classmate convinced me to interview at the consumer products company where he worked and I spent three years as a product manager there. I left that job to try some of my own businesses. After deciding the these businesses weren't my calling, I tried short stints at sales and investment banking before going back to grad school and ending up at Amazon 3 days after my last final exam in May 1997. Not exactly a straight line. He says AWS followed lots of squiggly lines, too. And of course, Andy Jesse is the uh by and large the creator of AWS. That was his major project during his tenure and founder.

Yeah. deserves so much credit in uh building that business. Uh the original vision included storage, compute, payments, and human intelligence. They had a product called Mechanical Turk where you could go and dispatch a specific task. You would have to build sort of a web UI, but it was the original uh sort of uh data capture tool, but it could also be used for little things like uh manual translation tasks. People weren't really using it for customer support tickets, but data labeling. data labeling before you really needed mass data labeling. But it was that type of business and there was always a question when scale started like is this just mechanical Turk? Is this business process outsourcing? Obviously that went on to a fantastic outcome with meta but uh it it became a different thing and quickly moved towards what we see in the uh in the expert networks and uh data collection at that's much more nuanced than uh a single task on demand. But AWS this was in the vision. Uh and they and they wound up um pulling away from that and sort of refocusing.

So Andy Jasse says some of those uh eg storage and compute became lynchpins in AWS. Others didn't uh uh others didn't succeed. We initially uh we didn't initially plan a database service. And when we built one, our first attempt failed to get traction. We went back to the drawing board and built new relational and non-reational databases which have resonated well and become core to millions of AWS applications. When we launched EC2, our compute service, it was a single instance in one availability zone, Linux only with no autoscaling, load balancing, block storage or private networking. Over the time, we added those capabilities and hundreds more services. And you see that when you go into the AWS dashboard, it is chalk full of different products. AWS was initially attractive to startups. Companies like Door Dash, Dropbox, Pinterest, Slack, and Stripe were among many that built their businesses on AWS. Uh, pundits said that enterprises and governments would never use AWS. Wait, what is going on here? I am I out of order. I'm I'm out of order here. Let me find the actual link on here and I'll stop reading from the paper because I printed the wrong here.

So, uh, he says, uh, pundits said enterprises and governments would never use AWS for anything substantive. But in 2008, Netflix, which was delivering, uh, DVDs in the mail for almost a decade at that point, decided to move all of their applications to AWS. Then came big commitments from GE In it, and others. And eventually, the CIA chose AWS as their partner to build their classified cloud. growth came fast and furious and as it accelerated so too did our capital expenditures capex with dilutive with a dilutive impa impact on free cash flow and so he's he's starting to make the case that there's this trade-off between free cash flow and capex and this will come back when he talks about his plans for uh taking advantage of the AI boom so he says at our first at at at at our 2014 more than 12 years ago AWS operating plan review the discuss started with a senior leader at the company musing, tell me why we're doing this business, like why are we doing this? It was confusing and it was very different from being an online bookstore and then eventually the everything store, but it felt it felt different but then eventually it started working. So he says AWS has worked well for Amazon but a straight line not really. Uh there's a bit there's a band I like from New Zealand called the Beths. Have you ever heard of the Beths? We got we got to listen to the Beths. Uh they've written several excellent rec records with thought-provoking lyrics. I eagerly await new releases. And when their latest album dropped last summer, uh titled Straight Line Was a Lie, it made me think how preant that expression is. Most long-term endeavors do not follow a linear top generally considered to be future me hates me and expert in a dying field. Interesting. And Little Death. What? What? Yeah, I I I'm gonna listen to the Beths on the drive home today. Uh he says any any Beth's uh heads in the in the chat.

So he's reflecting on his career and the twists and turns in building AWS. He says, 'Most long-term endeavors do not follow a linear straight line up and to the right. Progress jumps around. It'll zig up, then sometimes stall or zag down or force you back to the starting line. Sometimes it feels like you're running in circles, but the path is rarely straight.' That's because the world is complex and new technology, business model, invention, competitors, global issues, or people and cultural shifts can come into play. When we're in the middle of some of the biggest inflections of our lifetime — AI, robotics, space, industrialization, geopolitical and military conflict — and just as proficient golfers need to be skilled across driving, approach shots, chipping, and putting, durable companies must be adept at managing different elements of inflections. I'll share some of our lessons below and why we're bullish on what's ahead for Amazon. So he talks about how broad Amazon is these days: retail, logistics, AWS, ads, Kindle, Alexa, pharmacy. There are too many new efforts in flight to mention them all. He talks about the long push to bring robotic automation, acquiring Kiva in 2012. He also talks about all of the benefits that Amazon has brought to rural customers, but who are often deprioritized by logistics. This is a classic thing when you launch an e-commerce site: do you apply your flat-rate shipping to Alaska? Because every time someone ships something to Alaska, it's going to charge you an arm and a leg. Amazon's been able to solve it through a network that delivers over a billion packages each year. Customers living in over 13,000 zip codes spanning 1,200,000 square miles. And he's also focusing on closing the digital divide in rural communities. So things like high-speed internet access, internet, LEO — he's been talking about Amazon Leo, the low Earth orbit satellite network. But he makes this point that Amazon must be willing to pursue parallel paths when it's unclear what'll best drive the desired trajectory. He says two is greater than zero. When I was a kid, I used to go to New York Rangers games with my dad. I loved hockey and it was a high-quality time together. I looked up to my dad, still do, and hung on his every word. One game, my dad noticed that one of the Rangers' defensemen, Dallas Smith, had gone missing from the bench and stood up and exclaimed, 'Where's Dallas?' to which a nearby fan said, 'In Texas, moron.' So, yes, it's very weird that his name is Dallas, I guess that's the confusion. So, yes, it's fairly obvious that two is greater than zero. But too often companies focus on what looks most tidy instead of ensuring they have enough efforts in play to achieve an important outcome. Let's go back to fast delivery in our retail business. We know how much customers crave it and we see higher order completion rates when delivery promises are faster. So he talks about, you know, all the things that we know about Whole Foods and rolling out more stores, rolling out more investments in the core Amazon business. But then he goes on to talk about artificial intelligence and he says, if there's an obvious path to changing your trajectory, take it and run. But most new jumps aren't forward like that. There's invention and experimentation required and pursuing multiple paths gives you the best chance to find it. When you identify disproportionate inflections, bet big. Choosing which inflections are truly seminal versus just interesting requires judgment. Reasonable people can disagree, but if you believe you've found one of these disproportionate shifts, you want to invest as aggressively as you possibly can. This will create investment spikes that will invite scrutiny, but the game changers don't typically accommodate smoother investment horizons. One of these seminal shifts, of course, he says, is AI. And so he says, 'We've never seen a technology more quickly adopted than AI.' When ChatGPT launched in November of 2022, it reached 100 million users in two months, four times faster than TikTok and 15 times faster than Instagram. ChatGPT already has over 900 million weekly active users. Both OpenAI and Anthropic have run revenue run rates reportedly approaching $30 billion. These are breathtaking numbers for companies this soon after their commercial launches. And he has this great throwback to Thomas Edison and the dawn of electricity. A lot of people are looking for comparisons for what AI pattern matches to. And he goes back to the first commercial power station, which launched in 1882. So at the time, most people understood the first commercial power station as a better way to light a room. That was going to be the main benefit of electricity. It was going to replace lamps and you were going to have electric lights. But what they couldn't see was that electricity would eventually reorganize every factory, home, and industry on Earth. And he says that AI may have a comparable impact. The difference is that electricity took 40 years to get where it was going and AI appears to be moving 10 times faster. And he says that Amazon's smack dab in the middle of this land rush and companies are choosing AWS for AI. Three years after AWS launched commercially, it had a $58 million run rate three years in, and Amazon was already a big business at that time. So, you know, he's talking about the twists and turns of launching AWS. It's a...

You know, a lot of products worked. It's very quaint to be three years in, have the backing of, you know, a massive company, and still three years in, only hit 58 million in run rate. He compares it to three years into the AI wave, which of course started in 2023, basically, and now we're in 2026. AWS's AI revenue run rate is over 15 billion in Q1 of 2026, which is nearly $260 times bigger than AWS was at the same point. And it's ascending rapidly. And so he highlights a bunch of reasons why customers are choosing AWS for AI. Obviously, they have a product in sort of every single category now: model building, high-performance inference, lower-cost inference that runs on Trainium, their custom silicon, agent building, secure environments, etc. But he says that AWS could actually be growing faster and they are in fact limited on capacity. So AWS added 3.9 gigawatts of new power capacity in 2025 and expects to double total power capacity by the end of 2027. And AWS is monetizing that capacity as fast as it's installed. In Q4 2025, AWS reported 24% year-over-year growth and a 142 billion revenue run rate. That's a lot of absolute growth and yet we still have capacity strains that yield unserved demand. As an aside, two large AWS customers have already asked if they could buy all of the Graviton instance capacity in 2026, which is their widely adopted custom CPU chip. And he says, 'We can't agree to these requests given other customers' needs. You can't just say, okay, we're giving everything to one company.' But it gives you an idea of the demand. And so he goes on to talk about the other projects. The chips business is on fire. He talks about in the CPU space, virtually all of the workloads ran on Intel chips until they invented Graviton in 2018, which has a 40% better price performance. And then of course they're working on Trainium and Nitro as well. And so he goes on to explore a little bit of how the AWS capex cycle works in a way that's faster than AWS, and it requires more short-term capex. So AWS has to lay out the cash for land, power, buildings, chips, servers, and networking gear in advance of when we can monetize. And so there's been a lot of reporting about our data centers delayed. We're going to talk to Saagar Enjeti about some of the news and the political pushback to the data center buildout. But he says that it typically takes between six and 24 months before we can start billing customers, depending on the component. However, these capex investments fund assets with many-year useful lives: 30-plus years for data centers, five to six years for chips, servers, and networking gear. And the free cash flow and return on invested capital for these investments are cumulatively quite attractive a couple years after being in service. However, in times of very high growth like now, where the capex growth meaningfully outpaces the revenue growth, the early years' free cash flow is challenged until these initial tranches of capacity are being monetized and revenue growth outpaces capex growth. We've been through this cycle with the first big AWS wave where they invested a ton of money in capex building the servers before they were able to ramp revenue. And he says we like the results even though we've been through that cycle of high capex drawing down on free cash flow. And he says we expect to feel similarly about this next wave with a much larger potential downstream revenue and free cash flow. And so he goes on to talk about the... but I thought it was a very good letter.

Yes. Touches on a bunch of different stuff. I think this is what people wanted out of Jassy, yeah, like about a year ago, right? Yeah.

He was getting, you know, people were constantly kind of chirping at Amazon leadership and wanting some kind of like basically lay of the land like this to show that Amazon had a broad understanding and actually had, you know, real leadership and was committed to, you know, being a real player.

Yeah. Uh well, in other Amazon news, Amazon Pharmacy to offer Eli Liy and company's new GLP1 pill foundo via same day delivery. Shield Monoton says this thing is going to fly. Amazon is up 3%. I don't know if it's on this news. Yeah, it's hard to say. Amazon's actually almost up 5% today. Yeah. Uh, and I would I would expect that to be because of uh the letter, but people tend to buy a lot of GLP-1 uh drugs, too. So, but it does seem like a good I mean, we were we were talking about this with the uh uh some of the the the fast ramp of that one company um in direct to consumer GLP-1 and peptides. And there's so much skepticism about um fly by night peptide operations. Where are they being compounded? Are they safe? Do they contain even what you think they contain? And giving consumers an option that's something as you know established as Amazon with all of the all of the guard rails that they have in place uh feels like a very positive uh move for consumers.

Uh before we jump into the next story, Jasse did share a letter to shareholders from 1997. And there was I'll read the first two paragraphs. Uh to our shareholders, Amazon.com passed many milestones in 1997. By year end, we have served more than 1 and a.5 million customers, yielding 838% revenue growth to 147.8 million, and extended our market leadership despite aggressive competitive entry. But this is day one for the internet and if we execute well for Amazon.com. Today online commerce saves customers money and time. Time and money save both. Uh tomorrow through personalization online commerce will accelerate the very process of discovery. Amazon uses the internet to create real value for its customers and by doing doing so hopes to create an enduring franchise even in established and large markets. Uh well they certainly did that. Uh and and the letter goes on that to talk about h how it's all about the long term. So even back then growing like absolute crazy still, you know, wanting people to think about uh think in decades and and so Jasse doing that now and they have a good track record. Yeah. I mean Amazon's had decades of drawing down on free cash flow to invest in the future and uh they've had long history of communicating that effectively to the community and to the shareholders. Um, I mean the level to which people miss the ads. This has never I don't think this has ever happened before. Yeah. In the history of content of uh of podcasting. Yeah. We miss the ads, too. Yeah. All we have now is Well, uh we need we need some alerts because there's a there's a debate on the timeline about uh OpenAI's new cyber security product. Uh the uh there was a post that was deleted and then uh Axios issued an update. Basically the question was uh OpenAI is launching a a new model rumored to be called Spud trained on Blackwell sort of similar you know big model lots of capabilities uh and uh OpenAI's been working on cyber security products. Will they gate the roll out similarly? Are they running the same playbook? Will they uh take a different path? At what point will they make their models available? And uh people are going back and forth and Andrew Curran is sort of clarifying here that the new model and the cyber security product are separate and only the spe the cyber security specialized model will have a limited release, not the new model itself. So it looks like a general public release for Spud. Uh Dan Shipper shared some extra uh commentary around this. The Axio story floating around about OpenAI limiting the release of their newest model, Spud, isn't true. He just spoke to OpenAI and it appears the story conflated two things. They do have a cyber product. They are testing with a trusted tester group, but this is not the same thing as Spud. The Axio story has now been updated.

Yeah, there's a very there's a very strong argument to never release a model publicly that is specifically optimized for cyber, right? Because like you're just inviting a bunch like think about like the teenagers out there. I now get to be uh a super powerful hacker even though you're just kind of like probably should just be vibe coding like a fun little app or something like that, right? Yeah. And so you you would think having some cyber security capabilities in the model would just be better if you're vibe coding a product. It becomes secure out of the box. I'm just saying I'm just saying like having this super powerful like cyber focused product shouldn't be the kind of thing that anyone like you should probably have to KYC to yeah sure yeah what do you think about this yeah I mean also I think you know when you zoom out on AI progress it is like it is like very smooth um where like you know people were talking about this with mythos yesterday where like well actually if you run even like an open source model uh and you actually run enough times you actually can find similar you know exploits And it's just like kind of the efficiency of of the new model versus the the smaller open source ones. Yeah. So like like I I I'm pretty sure right now I could take you know 5.4 and I could just run it a bunch of times and I could find a lot of exploits like people have the capabilities right now. Maybe it's like a little flagged. Sure. Yeah. But and also like yeah I think I mean at some point like a lot of the coding agents I think you'll just see like in the you know in the system prompt or whatever of the coding agents it'll be like yeah make sure you're checking to see if any if there's like big security vulnerabilities. Even right now you can just ask you know codeex whatever you're building something like are there errors or are there you know security problems and it will solve them.

And this is the nature of every advanced model: in order to understand how to fight something, you'd have to know how to build it. And this is true. I mean, I'm sure that the folks at CrowdStrike or Palo Alto Networks could definitely, if they went black hat, it would be bad for everyone even in the pre-AI era. And that remains true. And so having at least as much economic incentive as possible to put the resources and the tokens and the inference budgets towards white hat hacking is good. There's also somebody in the chat making the comment that there's a lot of precedent for this with bug bounties and like time disclosures. So oftentimes white hat hackers will go to companies and say, 'We found a really bad vulnerability. We're giving you 90 days until we publish it.' But it is in the public interest to know that this vulnerability exists broadly. So we have to release it, but also we want to give you the time to react to this. And so it feels like you're sort of holding that company hostage a little bit. It can be a little bit tense, but I think if it's done carefully, it can be sort of a win-win.

Yeah, but I think regardless, I think more KYC is probably the future, right? Even if you're just talking about risk of distillation, stuff like this. I think I've talked about this before where like at some point you need more KYC to ensure people aren't just training off the model or doing nefarious things.

Meek Mill pulled out a slide from Bill Aman. The slide says a small percentage of songs are listened to. The overwhelming majority of music tracks receive zero or minimal engagement. AI is poised to exacerbate this reality. And so it shows that 2% of songs are culturally and commercially relevant. Commercially relevant is 10% of songs streamed a thousand to 100,000 times.

And then Meek is dropping that into what looks like Claude, saying, 'You're not in that 88%. You already skipped the whole bottom of the pyramid. You're already in that top 2%. People know your name. They search for you specifically. They stream on purpose. That changes everything. You already have the hardest thing: attention. You don't need the algorithm to find you. Your fans already look for you. So while AI is flooding the market with millions of trash songs that nobody... This is AI throwing AI on the... This is crazy. Your catalog just sits there collecting streams because people want Meek, not just a rap song.' And Dimes Square Holdings says, 'First in my bloodline to see a rapper with AI psychosis commenting on a deck from...'

I think it's a reasonable analysis. I just think it's a little bit unfair for the AI to be dunking on the other AI. Is this kind of like a crabs? There's a whole story about this in the journal the other day about a local group that's lobbying against data center development using AI tools, ChatGPT, very heavily to understand the legal code, how they can organize, who they should be calling. It sort of feels like a good use of AI in the sense that they're exercising their democratic rights properly through the correct channels. Like it could be so much worse. I don't know. What do you think?

Yeah. Well, hopefully by using the model enough, they'll be like, 'Oh, wait. This is actually good.' Yeah.

Yeah. Or like there are some parts that I like narrowly, but do we need this data center in my neighborhood? Maybe not. Yeah. Also, specifically on this, I think it's interesting like among celebrities, it seems like rappers specifically are like really leaning in a lot more than other types of celebrities, right? You see like Lil Baby, Lil Baby, not Lil Baby. Um...

There was a post yesterday about someone who got paid to set up OpenClaw forum. Yeah. You've seen this a bunch. Yeah. It's just funny like you don't see actors talking about their... Lil Baby was second after Baby Keem.

Yeah. That's right. That's right. I mean, Matthew McConaughey had some words about AI saying, 'Oh, yeah. When he was gonna log in, he needs to log in.' Well, no, no, that was separate. Like, wasn't it like months? It was like months, months, months and months later, he was like, 'This is coming. Like, you need to be prepared for this. You need to work alongside of it.' Yeah. But he's not talking about his specifically. Yeah. Exactly. Exactly. Yeah. I guess Ben Affleck had that company, right? Yeah. Yeah. Yeah. But, yeah. Ben Affleck already got the nine-figure exit. He's good.

Yeah. But you would imagine that AI actually might be a useful tool for understanding the impact of building a data center in a particular community because everyone is debating the economic impact. How many jobs will be created? What's the environmental impact? And being able to crunch through all those trade-offs so that the community gets to a net positive impact that the population can basically vote on and be happy about and say, 'Yeah, like we made this trade-off properly and we feel like we're getting the benefit because maybe it's generating a lot of tax revenue. Maybe the tax revenue is very durable. Maybe the tax revenue is going to the right place. Is it going to a tax refund or is it going towards a project that people don't actually support?' And so there's a whole other question for the local community: is that tax revenue meaningful? Like what dollar value do they put on new tax revenues for the city? Depends on what the city's building or doing with that revenue. Anyway, what is this post by IE Gender?

Actually, that's not impressive. The concept of a Dyson sphere was already in the training data. This is like, you know, we build the D, you know, everyone builds the Dyson sphere and then someone's coping and just saying like, yeah, I mean, it's not that impressive.

You're just watching the Dyson sphere be built around the sun thinking like, I'm not that important. Here's another self-referential thing: should AI be used to fight data center construction? Today Meta began removing ads from attorneys who are seeking clients that claim to be harmed by social media while under the age of 18. And so these class action lawsuits have been a big... you see them on social media all the time. Usually it's some sort of data breach or some random product that they're targeting you for. And I feel like many people just sort of scroll past them because the default class action is like you wait a long time and then maybe you get a check for $5 in the mail. But Meta has begun removing advertisements from attorneys. What is their justification for this? Lawyers across the country are now seeking new plaintiffs in the hopes of bringing a class action lawsuit that could result in lucrative verdicts. It's unclear if any of them have been backed by private equity as the California lawsuit appears to have been. Axios has identified more than a dozen such ads that were deactivated today, some of which came from large national firms like Morgan and Morgan. Almost all of them ran on Facebook and Instagram. Some appeared on Threads and Messenger plus Meta's audience network. One such ad read: anxiety, depression, withdrawal, self-harm. These aren't just teenage phases. They're symptoms linked to social media addiction in children. Platforms knew this and kept targeting kids anyway.

Yeah. So, Meta has something in their terms of service that says, 'We also can remove or restrict access to content, features, services, or information if we determine that doing so is reasonably necessary to avoid or mitigate misuse of our services or adverse legal or regulatory impacts to Meta.' Huh. It's basically like we're not going to let you use our product to take legal action against us.

Yeah. I wonder where the class action recruiting will go next. You know, it'll go everywhere else. Maybe out of home. Out of home. Yeah. Could go podcast. I mean, it's such a broad case that... Yeah. Well, YouTube is... it's relatively easy to reach Facebook and Instagram users. Yeah. Especially young people. It's everyone. Yeah. So they just need to cast a wide net. But it'll be interesting to see how that develops. Has anyone built an out-of-home advertising network for classrooms? Classrooms? What do you mean like teachers? I don't think people want to add advertisements. Kidding, kidding, kidding.

Anyway, 'the spreadsheet will become irrelevant,' says Buo Capital Bloke. No, the spreadsheet is eternal. The spreadsheet paradigm is capitalism. The two are irrevocably intertwined. We were using spreadsheets 3,000 years ago to trade oxen. We will still be using them in 3,000 years. Long live the spreadsheet. And he shares a stone tablet that appears to be a spreadsheet. This is in reaction to Signal's post that Microsoft 365 and Google Workspace have maybe four to five years of relevance left simply because the document spreadsheet paradigm itself will become mostly irrelevant. It already makes zero sense to draft, review or analyze anything without a native AI environment around it. That gap only widens over time. Most communication including email, chat, status updates is heading towards agent-mediated flows where humans set intent and AI handles execution. I don't know. I mean, both can be true here because you could be in a world where if you need a spreadsheet, the code is written on the fly and instantiated. Like we saw Ramp launch Ramp Sheets very quickly and it was a very full-featured product that was clearly only possible to build that fast. Cloning Google Sheets used to be insane. We had some luck cloning Slack recently.

Our intern built like a very convincing clone. It's sort of crazy in about an hour. Like I actually... I think it was on the... Was it on the $200 plan? Wow. Like that's incredible.

So, yeah, I think there's still the question of actually getting that to feature complete, maintaining it, really serving it, hosting it. That all of that seems like a hassle compared to just signing up on a consumption basis or a seat-based basis for a lot of SaaS. I'm still sort of not a firm believer in the near-term SaaS apocalypse broadly, but I think the more likely scenario is that you will see software instantiated in real time. So if you're building something and it requires a spreadsheet to visualize, and you're already seeing that with a lot of the chat apps when if you ask for a deep research report or something that requires building a timeline or a chart, it can write Python, it can write HTML that will...

The other thing is just giving customers a lot more flexibility around the product and thinking that like, okay, if you're buying a piece of software, it would be like buying a home where you get the home and then you can make changes to it to fit your own needs over time.

And so we'll see what ultimately changes, but you can imagine a world where, you know, it stops being... I forget who was on the show, but they were talking about how... Oh, it was console.com. Talking about how instead of having a request a feature button, you just let people build it themselves. Yes. Well, there's been some back and...

forth on the timeline between Nate Silver and Nikita Beer over whether or not links are truly deboosted on X. Phipe Le Moy said, 'After the exchange between Nate Silver and Nikita Beer, I did a little test to check whether that was true.' And to my surprise, what I found suggests that link deboosting was indeed reversed. What I did was randomly sample 15 tweets by the New York Times between 2019 and today. That feels like a small sample. It feels like you should look at maybe a larger swath there. But he computed the weekly average number of likes and retweets they got and plotted the results along with a trend line. The idea is that likes and retweets are probably a decent proxy for reach and the New York Times only posts tweets with external links. So by looking at this we should be able to see how many changes in the any changes in the algorithm with respect to how links are treated. As you can see it's pretty clear that starting around the spring summer of 2023 posts with links started to be penalized and eventually they were completely nuked until the spring summer of 2025 when a reversal to that policy seems to have started. To be honest, this isn't what I was expecting to find. So even if it's just a quick and dirty test, it's hardly definitive proof. It's good news and I thought I'd share the results. There is a question of like even so links, yeah, there was clearly like a deboosting moment although some really good stuff could break through like there were a few examples of people just posting links to, um, links to the, uh, like Google Docs basically and those would still rip somehow and I don't know if that was like a specific link but

I think one thing has remained true the entire time which is that if you are linking to content that is genuinely very interesting, it can still do numbers.

Yeah. And there is a little bit of maybe some feature of the New York Times where if you spent a year or two on X knowing that okay like there's sort of this adversarial relationship with links, you're going to open it in the browser. You might actually just develop a different habit of okay, I go and read the apps. Like I definitely developed the muscle memory to open the Wall Street Journal app regularly and scroll that news feed separately. And I don't really look for as much cross-pollination as I thought. Anyway, we have Sager and Jenny from Breaking Points in the waiting room. So we're bringing in to the TVPL job. Sager, how are you doing?

I'm great guys. Thanks for having me.

Thanks for coming on. Where should Let's start with the data center thing. We were talking about Andy Jassy. Oh yeah. Yeah, the tie is I'm getting brutally tied mom.

Dude, Jordy gets an exit and immediately starts dressing like a tech. What's going on here, man?

To be clear, I've been having I have the limited edition technology in motion.

Okay. Well, why has nobody sent me one? How come I dro? They're in the works. We'll send you one. You're going to be you're going to get to wear TVPN all day long. Yeah, that's great.

So we were talking about Andy Jassy's 2025-2026 shareholder letter, $200 billion of capex, tons of new data centers coming online. This applies to every company.

You were hoping for a bigger number, wouldn't it?

Oh, yeah. I mean, of course, as an American, I just want all of my farmland. I want all the trailer parks pushed out and I just want to make sure that there's beautiful data centers that are just sitting on previously like taken care of and people with generations just get kicked out. So I wanted 400 billion, 600 billion, something like that.

Yeah. I mean maybe we can start with the shape of the pushback against data centers because it's fallen in a few buckets but let's just start with like the big one which is energy. What are the headline numbers and fears that people are pointing to, the case studies that people are nervous about the energy impact specifically?

Well, I think especially guys, you know, this there's a pre-Iran war conversation, there's a post-Iran war conversation. I think that's really important to say because, you know, who knows where the price of oil will eventually stabilize, but when you have high inflation at the gas pump and likely at least some moderate increase in electricity bills, we don't know exactly where things are going to stabilize. You're going to have a lot more cost consciousness whenever it comes to energy. Now, especially in the midst of a data center conversation, I know that the pro-data center people are always telling us, don't worry, it doesn't actually impact your electricity bill all that much. Nonetheless, there is a feeling there. And John, I just sent you that study or that story this morning, which is very interesting. My colleague Emily Jashinsky actually talked to me about this, but a Wisconsin city has now been the first in the nation to actually pass an anti-data center referendum. And so, the ballot measure is basically the last thing that the voters in that town felt that they had at their disposal to effectively ban the entire city from approving any of these new projects. Do you guys remember the guy who went viral in New Jersey who blocked the data center? I interviewed him on my show. He was a longtime viewer actually of the program. He was just like a normal activist type, but hadn't really engaged with it. And he went through the zoning and he was like, 'Look, these developers are making all kinds of promises' and he was concerned about water. And I know, you know, again, the pro people are always telling me that that's not the case. Maybe, you know, I got to see a little bit more evidence for myself, but there was a problem and city council. Go ahead.

Is your mic I don't know that we're you're on the correct mic. We might be on that. The gain might be a little high. I will give everyone some news about this. Wisconsin city passes nation's first anti-data center referendum. The vote suggests ballot measures could become a powerful new tool for grassroots activists.

The chat thinks it's a foreign advert.

Let's hear if it sounds any better now. Is it a little better or turn down? I think that's great. There we go. Thank you. Amazing. Okay. So, is this what is this local Wisconsin city referendum like the new playbook that will spread all over America? Are there other jurisdictions that are already looking to copy and paste this? How do you think this where do you think this goes?

I definitely do think it will serve as a model especially as other remember this is really a bottom-up phenomenon at the community level and here in the state of Virginia where I live 40% of our power is actually consumed by data centers and so you've seen it became a major issue in the gubernatorial campaign both the Republican and the Democratic candidates were against data centers and both a lot of the candidates are seeing all of the energy I'm sure around this I'm saying at the community level I'm sure you guys also have seen all of these viral videos I referenced the one more recently of this trailer park which is being basically told they have to go out because they're going to build a new one. There's the famous one recently, I think it was in Kentucky, of a woman being I think she was offered like $20 million for her land. She's like, I'm born on this land. I'm going to die on this land. So, I really think that this is one of those bubbling political issues which no major politician, maybe Bernie Sanders, the only one that I've seen which has been willing to like jump on top of it. But I think it's going to be a major issue in '28. I think in '28 both candidates are going to have a data center platform. Maybe the Republican one, depending on who it is, is going to try and square the circle around GDP, but I don't think that voters want to hear anything about it. That's just my personal opinion.

Is the sense in Virginia that Virginia would be better off if the state had never had a data center?

Well, it's complicated. Remember, you know, we're full of a military-industrial complex, but don't forget this. We had Amazon, remember when they were going to leave New York and they were going to come here. They're actually going to come near where I live. It really didn't work. And there's a lot of local op-eds. They're like, 'Hey, we didn't get all the jobs that we were promised. Virginia threw all these tax credits.' And so people do feel really taken advantage of, I think, especially by Amazon here in Northern Virginia. But look, the backbone of our economy is always going to be the military-industrial complex and all of these like ridiculous server farms that serve the CIA and the Pentagon. But I mean, in Oregon, I think it was 30% of power in the state of Oregon is being used by data centers. It's like 20% in a few different states. Virginia is definitely we were the highest in the nation. But I can tell you at the organic level, Fredericksburg, more of the rural communities where there are the new projects that were being looked at all of those city councils, people were lining up for hours to campaign against this. Same thing in New Jersey, people lined up for, you know, in the middle of the night. And I've talked about this before, guys. Do you know what it takes for a voter to actually care? You could barely get people to come out for election day for the president. Now for to come up for a local meeting against zoning. Usually that's just really old people. This time people are bringing their children. They're sleeping in the room. They're waiting all night just to be able.

What is your advice for the data center developers out there? Because certainly I like you feel you don't think that we should just not build any data centers.

It depends. You know, I don't know. I mean, I'm becoming more radical as times go on. And so maybe we do just need a Butlerian Jihad. I'm testing the limits of your editorial independence, by the way. Shout out to you, Sam. But I mean, look, I'm not giving any advice to the enemy. I think really what I would say to the tech community more generally, and John, you and I have talked about this, is just prove to me that this is going to make my life better. And stop talking about all these scary things because I'm starting to believe you. In fact, I do believe you. What is this new Claude news that their new model is so scary they have to pre-release it to the cybersecurity company so that they can develop so no so that the new model can develop defenses against itself every day with you that the fear-based marketing like needs to stop because it's not helping anyone like part of the whole back.

I don't know guys I mean Jordy you have always made the case and maybe you're right is that that's a great way to fundraise every I'm the company that bought you wants more money. They have to go and be like, 'Look, we're changing the world. We're going to destroy all these shops.' Like, I don't know. I mean, for let's say like with Anthropic, I feel.

So my point was that it was very effective when there was no revenue in the industry, right? Yeah. And it was kind of this kind of far out idea. Nobody had had like a magical experience with AI. They hadn't had like a question, didn't have anyone to ask it to, and then asked AI the question, they got a good answer. Like when the average American had not had a very cool experience with AI, they hadn't generated an image that was, you know, photoreal or gotten help with their homework or help with, you know, writing any of these things that like now everybody has had the experience of, it clearly like was necessary to kind of marshal enough capital for a bunch of upstarts to be able to compete with the biggest companies in the world on this like the Googles of the world. And I just think at this point I made the case like, you know, at the end of last year where it was like we need to stop we need I think we need to stop this as an industry because why would average American be excited about this if you're just trying to give them, you know, Terminator nightmares but clearly the entire, you know, some people in the industry still feel like they need to.

Because I think they really believe it.

No, I think they believe it. That's See, I'm with John. Like I feel like Dario believes it and he just goes on every podcast in the world and says it like over and over again. And at a certain point, especially like with the new Claude news, I'm like, 'Dude, maybe I should just believe you.' And you know, people, what is it? People are taking leaves of absences because they're afraid. And I mean, you know, you can only call yourself the new digital god so many times before. Look, again, I'm a lite. I don't know a ton about the technology. I listen to smart people like you guys and many others to try and figure out what's going on. But if anything, I just want to be able to convey to the people who watch this show, you're not very popular right now. And especially in the midst of a global energy crunch, I would just be real careful about strolling around town and talking about how it's going to bring jobs in because a nobody believes you. They all think that their bills are going to go higher. There needs to be I think what this really needs, guys, is I don't think it can be solved at the federal level or sorry, at the company level. I think the government is going to have to step in and genuinely like codify let's say the executive order that Trump wanted to put in place where every data center project is going to have to supply the amount of energy that it's going to use it have to prove it to the local city council to the state to the feds like there needs to be genuine democratic buy-in if we're all going to decide like this is cool for us and I'm not sure that I'm not sure really that the tech companies are ready for that although maybe Sam is you guys saw his new social contract. I was about I was going to do a segment about it on my show, but I didn't have the chance because of this entire Iran war thing. But I mean, look, I mean, clearly he can foresee what the problem is. I don't really agree with some of the stuff he laid out, but yeah. Anyways.

Yeah, it was notable. So, Demis had an interview with Harry Stebbings this week. It's just 30 minutes. You would probably appreciate it because Demis is, you know, one of the godfathers. And he even was kind of talking about potentially, you know, one solution to this is like sovereign wealth funds like investing in the labs getting a piece of it so that every sort of citizen benefits. I thought that was notable because it was, you know, Google and the Gemini team is the most heavily funded lab in the entire world. And so we'll see how this stuff works out. In Virginia, Tyler on our team just dropped a truth nuke. He said in some localities in Virginia, 31% of total local tax revenue is from data centers. Is there like do you think that like I guess like part of what I'm trying to wrap my head around is like it seems like there are plenty of places in the United States where putting a data center somewhere that would generate a huge amount of local tax revenue would be a net positive. Right.

Maybe. And I believe that you're right that individual areas should be able to decide is this an industry that we want to support is this an industry that we want as a part of like personally do I want, you know, people, you know, drilling oil like, you know, 10 ft from my property line like probably not right but there are places in the United States where people have decided we're going to drill for oil here and we're going to, you know, be okay with those trade-offs and so I think I generally, you know, I've said this before I think there's a lot of reasons for somebody to like citizens need to understand okay what is the benefit of putting this data center in our county in our town and it needs to be there needs to be some positive benefit for the local community and so I think we agree there because you're not going to be using the models because the data center could be halfway around the world or in you still get access to all the AI tools that you want. It doesn't need to be in your backyard. And so I believe to Tyler who's in the chat. But I mean I do think though that that tax argument the problem is that largely again I'm speaking from a voter level is they do not feel either that impact or that they often like with the Amazon, you know, example that I gave you is that we give these tax credits away and then the promises a lot of these things don't end up happening. Come on, guys. This is a tale as old as time. Every state in the country, remember when Amazon did that ridiculous pony show for every state to throw whatever it could to them and then they ended up Yeah. for HQ2. That's the project I'm talking about is that Virginia, where I live, also did that. And then, of course, you know, local politicians feel very burned and a lot of the voters do as well. But ultimately, Jordy, I don't think we're disagreeing, man. Like, local control is really what it's all about.

That's what I'm saying. And I also think the feds though I do think that, you know, considering the way that Trump and the AI industry are so like they really do need each other just because of the way especially now after Iran. My god, can you imagine, you know, like look at how much he's reacted to market swings. One of the fears that I had really was that a lot of this, you know, you're talking about sovereign wealth funds John you and I know the amount of Gulf money sloshing around Silicon Valley. I'm like guys like we're about to see force majeure just get declared. I mean, if you see even 10% go down from QIA, SIA or any of these other investment funds, you're like, that's like an extinction level event for a lot of these or they might reallocate. They might start funding all these defense tech companies and not investing in data centers. And don't forget that an Amazon data center was struck by Iran in the UAE. I believe a couple of different tech infrastructure places were explicitly declared war on by the Iranians. So I do think that there's an element here where what we are watching play out I think politically and even geopolitically is that foreign adversaries can see clearly that this is a potential pain spot for the US economy here in the US. The administration is really aware of this, but there's also just this rising kind of populist tide against the data center movement and against I think really like abundance style assurances from politicians and from companies that after 40 years now I mean before data centers it was Walmart and before that it was something else. There's always going to be like a small guy versus a big guy feeling especially in rural parts of the country and you know I'm from Texas Jordy with the oil industry so I get exactly what you're talking about right like there's a symbiosis of we actually get rich off this stuff so we're okay with drilling for oil but you know there has been enough stories also even there where people go bust as a result and there's a lot of animosity towards the oil industry especially pre-shale revolution after things petered out And you know I don't know I've just seen this go politically in a lot of interesting directions enough to see that a something is happening b is that the tide is not only just turned but I think coming very explicitly and if what I would really advocate for is a genuine democratic process cuz that's what it's all about like people just feel like this is out of their control about AI especially they're like I want impact I want a say and I don't think they're wrong to say that.

What's the flow for getting some sort of resolution on the federal level? Is that like a new bill or something or like obviously this is going to happen after the midterms, but then yeah, and it's going to try and tie a bunch of these different questions in a bow around safety and cybersecurity and data center buildouts. Like is there a world where like a good data center exists in the sense that it supplies its own power with clean energy? It's not an eyesore. It's either buried or covered in trees or beautifully designed and it doesn't have any environmental impacts. Like is there a world where someone could get behind a data center at that level?

Yeah, absolutely. You know, and that's kind of the thing. Why are we just talking about data centers? I talked about this before. I read once a report of a guy who just came from China and he said nobody in China cares about power with data centers because they have a ton of power. He's like we have energy, dude. He's like, 'Bro, we don't care.' He's like, 'Yeah, you can build a data center. We're good. Like we have cheap and abundant power.' So if we have cheap abundant power all over the nation, no one will care either just about data centers, especially if we have, you know, cheap electricity, if there's no concern on the power generation level, and this is why, you know, I'm a huge advocate of nuclear power plants, whatever, renewables, I don't even care at this point. Shale gas, let's put it all together. But I actually think our grid and the delay of our grid and really just the knowledge right now, we're not investing a lot more into it. Doesn't seem like things are propelling. That's what makes it a zero-sum game. It doesn't need to be this way. Like really, what we need is just a ton of power. If we have a ton of power, then we don't care about anything being built.

So is codifying the ratepayer protection pledge into law a good move?

It would be a good move, but I don't think it's enough at this point. There's so much distrust, guys, like at the local level and on the political level. I recommend you follow Ryan Girdusky. He's done some more recent work on the data center question. He's been talking about this. He's like, 'Guys, this is just laying out there for any politician to pick up.' And I'm talking about it from an anti-data center position. He's much more of an expert on polling than I am. Go and check out his feed and some of the work that he's done, especially not just on feelings against AI, but data centers cuz I think he crunched the numbers. I know it was within the last couple of weeks that he looked at the question and he just said there's an overwhelming animus and that people are angry. So it's not just again let's be very clear it's about a feeling of lack of control over, you know, you're coming to take my job, you're increasing my electricity, you're changing my nation and I have to have a say as a citizen. So that's no longer about electricity that's about the whole picture of like technology and oligarchy and the economy. So you're fighting against a very very big force right now in the country.

Yeah. No it makes a ton of sense. Thank you for coming on and breaking it down. I'm sure we will be talking about this.

Have you tried just out of curiosity to just not use any AI? I mean, it's impossible to not use AI now because every company for the most part is using it. But have you tried to do your own personal like pause just for like a few weeks?

You should try. I want the like you should let's get you back on the schedule for a few weeks from now and in the meantime promise you will not use any AI at all until your next.

Guys. It would be so hard because even the chess board behind me has built-in AI. It's everywhere. Yeah, it's got the built-in it's got the chess.

No, it's fine. You can just No, just don't play chess, bro. It's fine. I can't do it. Just get an old school board.

I've been on one by the game. This is I've been completely one-shotted by the game. Anyway, yeah.

Thank you so much for coming on. We'll talk to you soon. Sorry. Great to see you, guys. Appreciate it. Always helpful. We'll talk to you soon. Up next, we have Joe Weisenthal from Bloomberg and Odd Lots, here to break down the New York Times analysis of the strongest Satoshi Nakamoto candidate. Joe, how are you doing?

Look at that. I'm good. Look at that sweater. Yeah. What do you think? You look fantastic. I love the green. You think the green works green here? The green is great and the color coming out the top. That's actually what I meant. When I said what do you think? I was like oh actually I'm wearing TVPN green for you guys. Yes. Glad you guys noticed the gesture. Well, welcome back. Nice to see you guys. Congratulations.

Is the New York Times the most underrated company in the world?

Yeah, of course. It's the most underrated tech company and the most underrated news company. The number of haters that the new and underrated. The thing though is to be fair that's the hater comment like that's just a gaming company. But look like it's extraordinary juggernaut both tech and news and so many people have some sort of criticism of it or whatever and find a reason to hate it and everyone can find good reasons to complain about the media. But it keeps on winning in ways that I've certainly been very surprised about. So I was really excited. I know the topic at hand. Satoshiology is like one of my favorite topics. I was super excited to see that they took a stab at finding out who it is.

Did you have a candidate in mind before this piece dropped? Had you watched the HBO documentary that actually pointed to a different suspect?

Yeah, I've read a lot. I actually didn't watch the HBO documentary in part I guess because it was just so heavily criticized etc. that I like maybe I don't need to watch it. I had often found the Nick Szabo, he's one of the candidates. I had always found that one to be fairly compelling. Yeah, me too. And you know, in my first read of the New York Times piece yesterday, I thought it was about Adam Back like I knew he had been in the mix. It struck me that most, you know, the piece didn't have a smoking gun, right? And that's sort of what we're all waiting for. We want that smoking gun whether it's some sort of digital signature or an email. It didn't have that. It was sort of lingu a lot of it was sort of linguistics like oh he makes it's and it's wrong or hyphens wrong. I make those same mistakes too. I don't know everyone makes those mistakes. So I wasn't totally convinced but then I read it a couple of more times over the last, you know, 24 hours. I read it again right before I came on this cuz I wanted to be prepared to chat with you guys and I'm like I've been like upgrading my odds that it is Adam John Caru has correctly identified him.

What have you worked with any journalists like John Caru before? It feels like such a lost art and it's so delightful when someone has a chance to it seems like this was reported for a year or more. Like he really he had the space and the breathing room.

No, I love it and like I said whether this was right or wrong is one of my favorite topics and it is one of these things where like over the years, you know, a handful of journalists have like I can be the one to crack this which actually like brings to my mind one of the more fascinating aspects of Bitcoin period. Which is it's amazing to think that like okay here is this thing it's kind of like a digital gold and so forth. That's a new kind of money and whoever Satoshi is he or her group of people whatever maybe it is Adam Back maybe it's not but the opse I mean can you even fathom being online for several years in a way that could not be traced back to you definitively like I wouldn't even know where to begin I wouldn't even know where to begin how to like interact online in a way that like I could ensure my anonymity and I think one of the fascinating things about Bitcoin is that it does have like this founder creator myth. This sort of the divine birth by this entity that left almost no earthly trace is as much a fascinating part of the story almost as Bitcoin itself as a technical accomplishment.

Yeah. And also just the fact that then Satoshi even like it wasn't an effort to remain anonymous in perpetuity. It was like then Satoshi went dark and nothing's moved on the wallet and it's just been this like that's obviously fueled the theory that potentially Satoshi passed away or something.

So here's a question that I have. So one of the stronger arguments or one of the pieces of evidence that the article in the New York Times points out is that Adam Back it's well known he has a long history with the cypherpunks. Had been interested in privacy and digital money for a long time going back to at least like the mid-90s etc. He had invented Hashcash which was, you know, which basically laid the foundation for proof of work which of course is elemental to mining. So he was like right there. Now one of the arguments that the piece makes is Adam Back like disappeared didn't have much of a digital trace of those first couple of years when Bitcoin was taking off. So maybe that was when he had slipped into Satoshi mode. On the other hand, there was no way that anyone in 2008 or 2009 or 2007 could have known really what a gigantic phenomenon Bitcoin was going to be. Why, you know, like Adam Back clearly was very comfortable talking about many of his projects under his own name. He talked about Hashcash. All of these people had been talking about their work in applied cryptography for years under their own name. So have not having been able to know how big of a deal Bitcoin was going to be. Why would this particular project suddenly oh he would like have the foresight to like oh I have to be pseudonymous for this particular project. I don't totally understand why someone who had like talked about digital money, digital money for years under his real name would have had reason to think, okay, with this project I'm going to have to talk pseudonymously or anonymously. That is not entirely obvious to me.

Like again in retrospect like sure like in 2026 we could look back yes it makes a lot of sense that this person would have like the pseudonym aspect but when it's just a bunch of people part of the community trying to solve the problem of peer-to-peer payments who have been using their names why suddenly like when these pieces come together this way would the creator feel impelled to be private in a way that they didn't feel impelled to be private with all the other Bcash Hashcash digital gold etc. That's not totally obvious to me.

Well, the way that I mean obviously the way you're talking about it is I'm becoming more suspect that it was in fact you behind. No. Well, so here so this is a Yeah. Wait, John, it sounds like you don't find this component.

No, I just I feel like after watching a number of projects sort of sputter where they did have a specific founder that could be just not even attacked but just like the motivations of that individual could be questioned and there's so much like risk tied to like a single person. It doesn't strike me as that crazy I mean clearly someone went anonymous, right?

One exercise is like is there a person that if it came out and we could determine that it was entirely factual that this one person who's the kind of person that would make the Bitcoin sort of trade up in value, right? Because there's like I can't really think of anyone. I can think a lot of different groups that it would trade down dramatically. It comes out oh some government entity, you know, that goes down dramatically. And so there's not a lot of like ROI for anyone on a personal level if they're.

One of the funny things about Adam specifically if you go to like his Twitter profile it says, you know, it lists his accomplishments and it says he is the inventor of Hashcash in parenthesis Bitcoin mining and what's funny is many people over the years have actually accused Adam of overstating what his claim there. It's like yes, there is clearly a lot of overlap between Hashcash and the notion of proof of work that was elemental to how Bitcoin mining worked, but Hashcash itself is not Bitcoin mining. It is a sort of a thing that preceded it was one of the antecedents or the preced whatever to Bitcoin mining a similar technique the whole proof of work concept. So there's been this claim that actually one of his issues is that he overstated his invention or role in the creation of Bitcoin which is very funny like if it turns out it's him that for a long time the accusation was that he overplayed his role. And then the other thing that's strange is that like since, you know, he's had like so and this is mentioned in the New York Times piece he was one of the guys who last year or whatever launched one of these Bitcoin treasury companies right like a MicroStrategy like entity that was just going to hold a lot of Bitcoin and does that sound like something Satoshi Nakamoto would do in my opinion not really but then the flip side is well that's like the best opsec in the world do a bunch of stuff that feels a little bit scammy, feels a little bit cheesy, feels a little bit like opportunity.

This guy is definitely broke. There's no way. There's no way he's got, you know, the tens of billions of dollars.

Yeah, that's what I'm saying. So, it's like on the first pass, I was like, 'No, like Satoshi is not going to like launch a little like mini me MicroStrategy company. This does not seem very likely.' But then the flip side is, man, if you were someone who like maybe some suspect of being Satoshi, what would be the thing that would be the least Satoshi like thing to do and throw people like maybe it would be like create a Bitcoin treasury company?

Or if you were Satoshi, but you lost the keys in some sort of silly mishap, then you would want to get back in the game because you're still a believer, but you don't have access.

Anyway, last question. Last question on the Bitcoin thing. Did you ever go down the Paul Le Roux route? And were you ever a Paul Le Roux believer? I was really convinced.

Paul Le Roux guy. Which one was that?

So he was a former criminal cartel boss informant to the US drug enforcement agency. He created TrueCrypt and was really early in basically spamming like the classic email spam of like, you know, mail enhancement. He would spam that out, scam everyone. He became so deeply entrenched that at one point he didn't just own a like domain name and he would switch the URL. He owned a TLD or like a registrar so that no one could kick him off and he could create unlimited new websites. That's pretty funny. So he would just create a new website every day and so the Gmail filters would get confused because well we've never seen this one before. I love that.

You know, the thing is I've always thought like that I don't know if any of you have ever read it, but there's an amazing book by an NYU professor Finn Brunton called I think it's called Digital Cash. And it's about the prehistory of Bitcoin. So, it's about this community of the cypherpunks in the 90s and 80s and actually even going back to the 1970s. Where people were talking about this fact that as the world was going to go online, it would be very difficult for me to send you a payment without the need for like a third-party database. The idea of like we need something that's cash-like, a bearer instrument on the internet. And I think this is like a really important thing to know is that there was a small group of people that basically for three decades or more before Bitcoin had been trying to solve this problem. And so like in the New York Times piece it points out that Back had been talking about all of these things that Satoshi had been talking about. And my first response is yes. Of course they did because this was this core group of people who for years understood that in the online context if payments were ever going to be a thing we would need a third party. And if there was a third party they can block a transaction they can reverse a transaction. They can penalize you for making a bad transaction etc. And so this was understood by some people from day one of the, you know, pre-internet that this would become an issue. And so I'm not surprised that they were all using roughly the same analogies and so forth.

Yeah. The book is Digital Cash, the unknown history of anarchist, utopian technologists who created cryptocurrency. And I will recommend a book to you about Paul Le Roux. I think it's called The Mastermind. And I believe it was optioned into a movie. I'm not sure. I think Michael Mann was going to do it and that sounded really good. But the book is fantastic. Anyway, that was really fun.

Has this been like a nice reprieve for you from covering the war? How have you been processing the constant gyrations in the oil market and the broader financial market? It feels like such a hard thing to cover, but obviously you've been doing a great job with Odd Lots and beyond.

Thank you. The stakes are I mean it is a hard thing to cover. It's an awful thing to cover, right? It's a war and wars are awful and the stakes are so, you know, there are so many ways that wars can get more, as catastrophic as they are, get even more and more catastrophic. So, it's like, you know, it's but it's also like an extraordinary story for the things that we like to cover, which is like understanding the real flows of the economy. And you know, prior to this war, I had not fully appreciated essentially just like how I mean, I knew about for oil, I was aware that the Strait of Hormuz was a very big deal. But like the degree to the fallout, whether we're talking about like fertilizer, whether we're talking about real estate in Dubai, of course, which you guys probably know more about.

People than myself, elevated to camping out to Dubai for the low taxes and stability, and then suddenly, you know, you're like, oh, this could be...

Honestly, credit to Americans, like, Dubai has just not been that popular.

I think it's different. It's very much, yeah, a European thing. When I was out there, I mean, it very much is like a melting pot. I've been there a couple times and very limited number of Americans. And the Americans that are there, for the most part, aside from on, like, you know, scheduled, you know, fundraising trips, are not really in the same type of, like, business community that you and I are in.

Yeah, it does seem like more of, like, a British thing specifically in terms of, like, oh, going there for the low taxes or the stability, etc. But yeah, like, all of these things getting upended, or the specific — not just oil and gas, that yes, that goes out of the Strait of Hormuz obviously, but then like this specific acuity, then pain points that are being felt in Asia specifically and how they're already having to go into sort of like a 1970s rationing mode where you see these headlines about if you have, you know, if your license plate ends with an odd number then you can drive on this date. I mean, the economic reverberation, it was clearly the biggest sort of global econ story since COVID.

Yeah. I was reading this article in The Economist this weekend, in Buttonwood, about how it feels like the market is not actually able to process what a 20% decrease in oil rates is. And it feels like, I don't know if that's just like hope that there will be a taco, that there will be a quick resolution, but the key stat that they were latching on to was that even if everything was world peace tomorrow, just bringing resources and infrastructure back online, just starting the flow and moving the ships again takes, you know, three months, six months. And so you will see reverberations through the economy for months if not years even if, you know, we get a good outcome immediately, which is what we're all hoping for.

Yeah. Some sort of benign outcome. But like, but this is the weird thing too, and you're spot on, and this is the weird thing, which is that if you talk to the commodity guys, they're like, this is unbelievable. This is like a scenario that we couldn't even really have contemplated. And then you look at the stock market and, you know, oil is up a lot. It's not like oil markets are not registering. It's up a lot, but then you look at the stock market and it's flat on the year, like as of the time I w— or it's like down 0.3%.

I mean, it's stunning because you have to remember that even like the last thing we were all talking about prior to the war was like, oh, every legacy business model in the world is going to zero because AI is getting so good. So that was the first two months of the year. And then the third month of the year is this war that massively shoots up the price of oil, kind of like takes the prospect of rate cuts off the table. And we're looking at a market that's flat on the year. Like it's really like, you know, I try to — I'm sort of an EMH guy and so I always think if I'm missing something I just trust the market. It's very strange, you know. It's very strange.

It is very strange. Have you found any particular corners of the economy that are serving as white pills these days? Because it feels like a year ago things were sort of — there was political chaos, we were talking about the tariffs and whatnot, but it felt like in broadly most industries were sort of seeing reasonable advances. Things were much less controversial and we're in a much more tumultuous time.

I mean, it's hard to disagree with that. I mean, I think that there's basically — I think most people say there's basically two parts of the economy that are growing, and one is anything related to AI and data centers, etc., obviously, and the other is like home healthcare jobs. And I don't think much has, like, changed on that front. And unfortunately, like, okay, like again, let's just go back to the world of February 26th before the strikes had begun. You know, the issue was we were already still at that point well above target for the Fed's inflation. Now, reason to think maybe it's trending down, maybe there's some softening, etc. But even with just essentially two cylinders out of — I don't know if the cylinders analogy is great because I don't know how many cylinders, but two cylinders really firing, like, okay, ongoing need to, like, provide people healthcare and then AI, even with that the economy was like under a certain level of, like, strain with price stress. Yeah. And so then you add in, okay, now, like, all forms of commodities are going to be more expensive. They're spending, of course, wars are very costly from an expenditure perspective. And so where does that leave the rest of the economy that's not AI or healthcare? I have yet to see — I don't know, I'm certainly not feeling white pill to the extent maybe the closest is like, look, I would say there's still evidence a little bit out there that, for example, that we're not seeing the evisceration of tech and software jobs the way some people might have anticipated, and there continues to be like evidence like, oh, actually, like that sort of like the economist optimism that maybe AI will create demand for software engineers because there were more things that could be softwareized. I actually think maybe that's looking okay, that thesis. And so it's white pilling in the sense maybe we don't have like the total evisceration of the sort of the labor market, like that part. We actually got a good jobs report last week on the Good Friday report. So like, there are things...

And we trust all the jobs reports now. We could definitely lean on...

The revisions have been a little tumultuous for me.

There have been a lot of revisions. It's very tough to measure the economy, but I salute our brave statisticians at the Bureau of Labor Services.

And we did talk to some other folks who were looking at like private market data, payroll data, and there were green shoots there as well. So I'd be — I'm optimistic that that holds. Yeah, I've heard the most optimistic take about just more companies building more new things. We've built a bunch of internal tools here. The actual, like, NPS score for most AI apps is actually very high even though AI as a concept polls very poorly. So there's this difference there.

What does NPS stand for?

Net promoter score, or like approval rating. So like people may give ChatGPT directly like an 80% approval rating and it has like a lot of good reviews in the app store, for example. And if you just ask them about that or what they're building with Claude Code, they'll tell you an amazing story of some custom software that they built for their business that they would have had to spend a million dollars on a consulting group to do it and it maybe would have been really tricky to get it to math out. But then they'll just tell you like, oh yeah, we needed this dashboard and we just made it in a day or a couple weeks. So people love their specific things they built, but then it's like terrible and they hate...

Yeah. No, I mean, look, I think this actually seems to be a recurring phenomenon over the last several years to some extent. Said this about the economy generally, like, I'm doing okay, my financial situation is actually fine, but the economy is like really terrible. So it would be kind of interesting like if maybe there was a pattern that we're seeing across domains where, you know, everything else — I was, yeah, I like AI, I like how it's doing, I like this thing I built, etc. It's all terrible. Maybe that's just sort of like a general disposition that people have towards a lot of things.

Yeah, it's the Kyle Scandlin vibe session concept. And people maybe they're happy in their relationships, but they were like, oh, but dating for everyone else is a total wasteland. Everyone these days — maybe that's actually kind of a thing.

Last question. Any prediction market predictions on your side?

What do you think? We have an episode out today with Thomas Peterffy, the founder of IBKR, which is also one of the most impressive trading entrepreneurs. They're really getting into prediction markets. So I think my prediction is that it's not just going to be like a two-entity race for that much longer. I think as it gets bigger, there's probably too much money to be left on the table for it to just be these two companies that most people never heard of up until like a year ago.

Yeah. No, makes a ton of sense. Well, thank you so much for taking the time to come with us both.

Take care.

Thank you. We'll talk to you soon.

Cheers.

Our next guest is Andrew Dai from Allian. He's the co-founder and CEO, with big raises coming in to the CBP and Ultra. Let's bring in Andrew from the waiting room. Andrew, how are you doing?

Hi. Good. Thanks.

Thanks so much for taking time to join us. I would love to hear a little bit about your background. Maybe we should start even before Google Brain and DeepMind. What was your academic track like going into tech?

Yeah, so I grew up in the UK. I lived in Manchester and Sunderland and London, all across, following my dad's work. And ended up in London, did my high school there. I went to the University of Cambridge for computer science and then I went to the University of Edinburgh for my PhD in AI.

And what did you work on when you were at Google?

At Google I worked on a whole bunch of things. I worked on Google Now. I worked on Smart Reply and Smart Compose. But really I had a paper about 12 years ago where we proposed language model pre-training and supervised fine-tuning, and that's the paper that all the GPT papers cite, and here we are today.

Turned out to be a big deal.

Yeah.

Yeah. Did the...

Wait, yeah, sorry. I don't know if you're going to dig in deeper there, but at the time when you were writing that, how much conviction did you have around the paradigm? Did you expect...

And scaling laws in particular?

Yeah, I knew it was going to be something big. And I could tell because when I was giving my poster at EACL that year, one of the inventors of LSTMs, which was the biggest model at that time, they said it just works. Like the method just works. And there I could tell, oh yeah, there's something happening here. But I never imagined it would get to this scale and I never imagined I'd be doing it for more than a decade.

Yeah. So, did you go to NeurIPS this last year?

Uh, yeah. Yeah, I did.

How has it changed? How has it changed?

It's all now. A lot of language and VC. A lot of VCs sneaking around too.

Lot of VCs. That's right.

Well then, yeah, take us through the decision to launch this company. What you're thinking needs to be different about the current strategies employed by AI labs.

Yeah. So, our company is built around visual reasoning as a first-class citizen, multimodal reasoning. So looking at all the labs, you see there's a lot of focus on text, on language. That's been very effective, right? We have new paradigms in cybersecurity right now, but the visual capabilities are kind of getting left behind. And there you have problems where the models on visual problems are at the level of like a preschooler, of like a three-year-old.

Can you give me an example of that? What does that look like in practice?

Yeah. So in practice, you can tell these models to generate a pool table and they will make a perfectly good-looking pool table, right? But if you ask them to count the number of balls on the table or count the number of bottles in a bar, then they will just hallucinate. They'll be off sometimes by a large amount.

Yeah. So I've seen sometimes the reasoning models will ingest an image and then wind up writing a bunch of Python code to basically count pixels and do things like that, very much not the way I would imagine a normal human would process counting something. At one point I was asking to estimate the height of a desk and it was, you know, writing all this math to sort of manually, you know, try and understand the size of things, which was a reasonable approach. It wound up just being a normal-sized table, which was sort of underwhelming. But how are you thinking about the actual development and what you want to do differently? Are you focused more on a new architecture, new data sources, more scale? What's the shape of your strategy here?

Yeah. So, we are basically a full-stack team. We have experience with pre-training, data, multimodality. So, we are essentially building specialized models that includes new architectures for visual reasoning, very specialized sets of data with specialized data processing and new algorithms. So we expect to be running the full gamut of changes and this is really needed to really make a breakthrough in visual reasoning.

What does good training data look like today? Is it synthetic? Is it images, videos, 3D, you know, virtual worlds? Like what's the shape of what's valuable?

As you can expect, it's a bit of everything. But yeah, definitely natural data, so data that's not generated from a model, is more valuable and more useful. Data from a model, that synthetic data, has the risk of putting the model into like a weird place where it just outputs like m-dashes or just like tries to repeat the same thing all the time. But yeah, data around the natural world, around the 3D world, that's invaluable.

What do you think the cybersecurity analogy is for visual reasoning? Like cybersecurity is so equipped for text-based models, coding agents. The entire cybersecurity threat can sort of be understood as a big string of text more or less. Where do you think visual reasoning goes in terms of applications?

Yeah. So for applications, one of the most promising ones that we're seeing is engineering. So right now all these engineers, mechanical engineers, hardware engineers and also architects, they're drawing all these diagrams in CAD software which has been developed over the last few decades, but there hasn't really been AI breakthroughs there. People are still doing basically the same thing they've been doing for the past few decades. And so what we will do is we will produce models that really understand these drawings. Like say you have a real estate floor plan and you want to say make this bedroom bigger or add an extension to my house, right now that would take weeks and lots of manual time and then you have to follow building codes, make sure like everything is correct. And that's because there's not really any reasoning in our current models, and so we think there's a huge potential.

One of the hardest times we've ever laughed on this show was reacting to a floor plan that was generated by AI. And it looked so high fidelity. Every line was perfect. It didn't have any of the fuzziness that you'd expect from earlier models, but when you dug in, it made no sense. There were like 12 toilets and two baths and it didn't really understand the problem. So, very exciting. Well, you raised some money. Tell us about it. How much did you raise? We want to hit the gong for you.

Yes, we raised $55 million.

Oh, congratulations. From who?

And that's from Sequoia Ventures, Menlo Ventures, Automator, and Nvidia, and 49 participating. And we have some great angels there, including Jeff Dean.

Jeff Dean. Awesome.

He's been on an angel investing tear. It's really exciting to see. I mean, obviously he's a legend, but he's clearly very optimistic about all these different approaches going forward. So, very excited. Well, thank you so much for taking the time to come. Great to meet you, Andrew. We'll talk to you soon. Have a good rest of your day.

Cheers.

Up next we have Lumini and Kesava Kirupa Dinakaran in the waiting room. We'll bring him in to the TVPM Ultra Dome. They have raised a $38 million Series B to scale AI automation for health systems. How are you doing? Good to see you.

What's going on? Doing great. Great to see you both.

Great to see you. I believe it's your first time on the show even though we've met in the past. Please introduce yourself and the company.

Yeah. I'm Kesava, one of the co-founders and the CEO of Lumini. We help large health systems drive automation in their operations and back office.

So is this a diffusion story or is this a foundational AI research problem? Like what is the secret to actually driving improved healthcare outcomes in these larger organizations?

Yeah, I think the models have kind of got to a place probably in the last maybe six to twelve months where some of the opportunities to actually drive automation end to end in a very reliable manner is very possible. So it's 100% a diffusion story at the moment because the reality is like, you know, American healthcare still runs on faxes in many ways and so you're kind of at this opportunity now to dramatically change how it's done in the first place.