AIバブルかどうか、まもなく分かる――Anthropic IPOとS-1が示すもの

In a few weeks, we're finally going to get some answers on whether we're in an AI bubble. That's because the Frontier Lab Anthropic is going public. Anthropic gets ready to make its Wall Street debut as the AI giant. AI giant Anthropic. That's the company. Cisco-based AI giant Anthropic filed papers today. It might raise $100 billion at a $2 trillion valuation and rescue a whole slew of companies floundering in the debt market. Or it might raise a lot less and send the entire stock market into a tailspin, Which is to say a lot is riding on this IPO. The near future of the American economy, the medium future of the tech industry, and given that the risk factors in Anthropics S1 include catastrophic and existential risk to humanity, maybe even the long-term future of the human race. At a minimum, the Anthropics IPO will mark the end of an era. The venture investment-driven, hype-accumulating startup phase of the AI century is nearing completion. But what comes next? In today's episode, we're going to try to wrap our minds around the most important company in the economy, at least for now,

and look at what's at stake in its big moment. Are we in a boom or a bubble? Will tech's big bet on data centers work out or fizzle? And can Silicon Valley's most militant AI believers be a normal public company? Or are we entering a new era of AI hype mysticism? I'm Max Reed. This is Hard Fork. We are live in studio with Erin. We're not live. Okay, that's a really good start. We are here in studio with Erin Griffith. It's so good to see you in person. Yeah, likewise. There are two big developments in the hard fork universe this week. First is that you finally followed me on Instagram. Oh, yeah. We can get to that later, maybe in the segment. Yeah, I was on the fence. Yeah, well, there's a lot of referendums that we're going to talk about today, And I think that was an important one. The second thing I want to talk about is that Anthropic announced its intent to go public, which means the rubber is finally hitting the road after however many years it's been of hype and doomsaying and argument.

You have been reporting on Anthropic and the AI world for a long time. And I wanted to sort of pick your brain about this company today because this feels like a huge inflection point for AI, for tech, for the stock market, the economy, humanity. Yeah, just the stakes are only that high. Yeah, exactly. So what is riding on? I mean, everything. Everything. I mean, I think that was our headline was nearly everything. Practically speaking, this is the first pure play, like, LLM, large language model company to go public, to open its books so we can actually look at the underlying business of this AI boom that has been going on for four years and attracted trillions of dollars in investment and sort of reoriented our entire economy around itself. And, you know, Anthropic alongside OpenAI and, you know, arguably Gemini from Google are kind of the companies that are driving this. They're like the pillars that are, you could call, I've heard many metaphors, okay, you could call them the bottom piece of the Jenga tower.

I've heard them described as like, you know, the tip of a pyramid with like an entire ecosystem of companies supporting it. But either way, they are sort of the quintessential AI companies. And now, for the first time, we have kind of a look at the underlying business model that so much money has gone into. So are the investors you're talking to, are they nervous about the IPO? Are they worried about what might come out of it? Well, I mean, last week they were definitely nervous. Now we've had a day or two to kind of look at these numbers and see. And some of them have leaked already along the way. So we had kind of a sense that this is the first time we see, like, in a 300-plus page document, the full picture of what Anthropics business looks like. So let's go there. What are the sort of big important things that we can take away from the S-1? Okay. Well, first of all, I guess it's okay to say that business looks kind of healthy. I mean, it's arguable. There are people that have, you know, made the case that it's not. But I think, like, generally speaking, people have been kind of pleasantly surprised to see that this company is capable of making money.

And, I mean, when we've seen how much money these companies have raised over the years, we've seen how much they've been spending to build their AI products, I think the general consensus has been, like, these companies can never make money. And just to be clear, Anthropic posted an $80 billion net loss. So maybe let me unpack that a little bit. An $80 billion net loss seems insane, but it's actually kind of a weird accounting quirk where they have to count some convertible debt that they issue to some of their shareholders as a net loss, and that will essentially be cleared in the next quarter or by the time they go public. And so that's not a reflection of the health of their actual business. What we've actually seen is that they, in the first half of this year, had an operating loss of around $600 to $700 million, and that's shrinking from the same period last year where it was like over $2 billion. So they're shrinking their losses of the actual business that they're operating. On the flip side, they are still shelling out enormous sums of money, and they're on the hook for enormous sums of money to build out their AI infrastructure.

So this is the data centers, the chips, the compute power, the very, very expensive stuff that they need to keep training new models and to support the old models. So it's kind of – and Dario has talked about how the company is sometimes on a nice edge when it's making these decisions and signing these deals. And so they have signed so many contracts in anticipation of the demand. And it's like a very high-risk game. If the demand's not there, then they're going to be sitting on all these contracts. And many of them are non-cancellable. If the demand is there and they don't have enough, then they're going to be losing business. And so that's kind of what's underlying all of this is how much AI are businesses and people going to be using? And has the company basically laid the groundwork for that? Right. Right. I mean, it's interesting because on the one hand, the S1 seems to confirm that there's an enormous amount of demand. I mean, we're talking about $16.3 billion in revenue for the first half of 2026. And that is an enormous amount of growth, too. Like, you know, we're seeing a company grow 14x. I mean, that is insane, especially for a company of this size.

So we've never, I mean, I hate to, like, say unprecedented, but it feels like we haven't seen a company growing like this ever before. Right. So we can say there's demand for this, right? But the question that you're saying is we don't actually know how much it is, and it might be that there's more even than we've already seen. It might be that there's actually less. But we're building a certain number of computers. I mean, to be very simple about it, no matter what, we're building a lot of computers out there in the desert, and Anthropic has committed a ton of money to those computers. And so is that the biggest risk they're facing? Definitely. I mean, this is, I think, what everyone's looking for. And so the S1 that we saw did not have Q3 numbers in it, and they've already kind of previewed those numbers to their investors, and our reporting and others have shown that the growth is actually slowing in the third quarter, and that's because first half of the year, Claude Code just was growing like crazy. They basically took the entire industry by storm, and everybody's using their coding product. Now, competitors like OpenAI are starting to catch up. Companies are trying to pull back on how much they're spending on tokens because they realized that token maxing got out of control,

and so the growth is a little bit slowing, and they're getting some competition from open source models or cheaper options. And so I think in Q3 they grew 44%. And so that's a pretty significant slowdown from, again, gangbusters growth down to, like, just great growth. But is that going to continue on that trajectory? In that case, okay, the growth story isn't, like, as exciting, and perhaps demand isn't, you know, going to be through the roof forever. So that's one question investors are looking at. And I think just in general, they're looking to see how Anthropic explains that slowdown because if Anthropic just comes in and says, like, the whole market is pulling back, that's very, very bad for this whole ecosystem, the Jenga tower or the pyramid that I'm talking about for, like, you know, all the demand and all the investments that everybody is making, all the debt that everyone's taking out to build this stuff out. But if the reason that the growth is slowing is just because of competition, because open AI or open source companies are kind of eating market share, but the demand is still growing, then, like, great, party on. The AI boom continues to boom.

So they're looking for new avenues for growth, potentially. They're trying to sell different stories about what the next big product is. Yeah, and I think we're going to see this on the roadshow. I'll be watching that pretty closely. I think that's going to kick off in probably a week or so. Before you start, what is a roadshow? So in a week or so, Anthropic is going to basically go on this tour where they meet with all the public market investors and kind of tell their story. And they call that a roadshow. And bankers, like, kind of take them around on a little private jet. And they're, you know, meeting with, like, insurance companies in Iowa or whatever and telling them about the machine god. So that's going to be just, like, kind of amusing on its own. But, you know, they need to tell, I think, a new growth story because they're going to say, okay, we had this, like, coding product that's incredible. but like it's already kind of slowing down growth. What's our next growth story? Well, they believe that AI is going to solve many, many, many problems in the world. And so one of the stories that they're pitching to investors now is this bio thing. You know, AI is going to cure cancer. Like AI is going to develop new drugs. So that's like a big story they're going to probably be pitching.

And then they have other products too, like legal tech or other ways that AI is going to kind of disrupt the economy beyond just coding, which has been, again, like a blockbuster hit for them. Right. One thing I want to talk about with the S1 is the 93-page risk factors section. You've read a lot of S1s before. Are they? Every word of everyone. Obviously, in your capacity as a reporter. Are they usually 90 pages long? I mean, they're usually pretty long, and it's a lot of legalese. It's just like a lot of ass-covering, basically. It's basically anticipating all the potential things that shareholders could get mad and sue over and saying, we warned you. Right, exactly. And there's a lot of those for Antopica. I'm just going to read a couple of them right now. They're very unique in their specificity, let's just say. Yeah, and I would say, too, it's funny reading them if you've been following AI for a while. It's kind of like getting six years of Twitter and blog posts and Substack compressed into a document of legalese because they're listing everything from the resource constraints on building out compute to the various sort of AI safety problems to just sort of like your basic business problems with customer satisfaction.

But the ones that caught my eye, for example, are concerns have been raised both within our organization and by other researchers that sufficiently advanced AI systems could, if not developed, deployed and operated with appropriate safeguards, oversight and security measures pose catastrophic or existential risks to humanity. This is a pretty rare kind of thing to do. Another one that I was really, it was really. Can we just go back to that really quick, though? Because one thing that the great Matt Levine of Bloomberg pointed out was that this is written in passive voice. It's not Antropoc saying we think. Right. That our products. It says like some people may be saying. Within our organization, some people. I don't know if it's our founder and president, Dario Amadei or anybody, but some people have raised this question. Who can it be? Yeah. Yeah. Another one that I like is, in addition, considerations relating to model safety, interpretability, alignment, and welfare may lead us to limit or delay the deprecation of older models. So just to translate that a little bit, when they're talking about model welfare, that's they think the models might be alive and they're not ever going to shut down some of the product lines because they want to make sure that they get taken care of.

I miss this one. Yeah, I have to credit Joe Weisenthal at Bloomberg for picking this one up. They're not saying this out loud, but they're saying considerations relating to welfare may lead us to limit or delay the deprecation of older models. This is a little like Apple saying the iPhone 12 is alive, so we're just going to keep making it forever. Oh, my God. Imagine trying to explain that to some, like, you know, analyst at Goldman Sachs and be like, we need to put this in here. I mean, the thing that I wonder about a lot is how much people who are spending their time talking to chatbots all day start to buy into and belong to this stuff. And so you actually will have activist investors who are saying, you know, we have to make sure that we're taking care of model welfare and whatever else. This is the future coming soon. Well, the activist investors won't have a whole lot of ability to do much because of the governance of the company. Yes, I actually want to get into this. But let's start before we get there. Let's start with just a quick word about the founders, because this is a company that has a really specific philosophy around AI risk and AI safety. That's sort of the core, not just the core of the business, but it's kind of their origin story.

I'm sure a lot of people are familiar, but will you just give us the 22nd version of where Anthropic comes from? Sure. And this is kind of the story of AI also. You know, Google's working on AI. Elon Musk and Sam Altman see it, and they're like, man, I don't think they care enough about the future of humanity. So they split off and start OpenAI, and they're like, we're going to build AI, but we're going to do it for the good of humanity. We're the principled ones. And then within OpenAI, Dario and all of his friends are looking around and saying, wait, Sam doesn't care enough about the future of humanity. We need to build the safe AI before they can. So they split off and do that. And that has kind of continued on down the line. So that's where Anthropik is. Their roots are truly in AI safety. And they are the probably most extreme on that spectrum of the fears inside the company, at the top of the company, and around it. Well, let's talk about how this relates to governance, because they have this very unique governance structure. I like losing my voice now And hopefully they can use some AI to make my voice sound less gravelly Yeah so Anthropik is kind of following a trend that has been happening for the last 15 years at tech companies

which is that founders want to retain control even while selling shares, which would, in general, the classic form of governance means that if you sell shares, those shareholders get more control. This started kind of goes back to Mark Zuckerberg and his multiple class of shares. But Anthropica has taken it even a little bit further. They have five classes of stock. Their seven founders have 50.1% voting share. You know, they're going to be in charge of electing all the directors to the board. And then they also have this, like, independent trust that controls a majority of the board seats. And so it's, like, a little bit convoluted. And I think no one's going to push back on this because over the years, every time startups have tried to do this, investors have made a big stink about it, like, oh, no, we hate this. And then they just go along with it anyways. And so, like, it hasn't, yeah, there has not been a line in the sand. Even with SpaceX, they were a little bit more extreme about that. Some of the exchange or some of the indexes tried to put their foot down. And then at the end of the day, they're just like, ah, fine.

So this is not necessarily new in Silicon Valley, but it is for Anthropic, seems to be distinctly sort of connected to the AI safety philosophy at the heart, which we should say, I'm not sure we said it out loud, but like really specifically, Anthropic is the company that believes most strongly in the possibility of total AI annihilation, Armageddon, and additionally believes most strongly in the possibility that the LLMs are alive or conscious or sentient in some way. And they lay out that they believe their founders are best suited to sort of shepherd this into the world in a way that, you know, is a little self-aggrandizing. When I was reading it, I was like, wow, you guys are special. And I guess you are because you're at the helm of a potentially $2 trillion company. But, you know, they are even able to, like, name their successors. They have a secret list of successors that they all have to approve if, like, somebody leaves the board. And so they're setting this up to be in control of it forever. They don't want to lose control of the AI or of the company. Yeah. I want to return to the philosophy, the sort of question of the philosophy, because one of the questions I have about Anthropic is, you know, beyond this governance structure

here, to what extent is there, are there philosophical commitments around AI safety? Like, how do they inform the company itself or inform the business decisions? Or do they? Well, I think in the beginning, it was a huge piece of the company because it was the only place where they could get money. regular Silicon Valley investors kind of viewed the company as a science project for the first couple years and the only place that they could find money was from people like Sam Bankman-Fried Jan Talon people who were deeply enmeshed in the EA kind of philosophy in that world and then their first product was very specifically and purposefully like an enterprise business product they weren't building something for consumers they weren't trying to build an AI companion necessarily and that has proven to be an incredible business And then when the VCs got word of that, they were like, oh, great, let us dump all the money in. And most of those VCs that came in later are looking at this as an enterprise software, SaaS business, and they're not really that hung up on the philosophy. They're like, yeah, we like that we're backing the good guys, the good guys.

But they don't really buy into this sort of stuff that's a little bit more weird about it. The woo-woo stuff. Yeah. I mean, is there a word for AI woo-woo? No, I don't think there is. Which is wrong with one. Yeah, exactly. I mean, it's very interesting because you could look at all the decisions that they've made from the governance structure on down to like the specific things that they're doing. And you could tell a story where it's all about their philosophy. It's all about their fear of AI or their belief in AI. But you can also tell a story that this is just a very straightforward, smart Silicon Valley company that founders want to retain control. It wants to be very efficient. It wants a clear path to profitability through enterprise sales and not trying to become the next Facebook or whatever. And it seems like investors... And that's certainly what the investors that I talk to say. I remember when Jacob Cox quit and everyone was kind of freaking out about, hey, it's going to kill us all. I'm mentioning this to some of their investors and them sort of being like, oh, yeah, hacking, that's a big risk. We're definitely worried about cybersecurity, but I don't really know what you're talking about on the other side.

We're not worried about that. Let's zoom out a little bit. So this feels, as we were saying before, like an inflection point. Like we've spent the last half decade in this big time hype moment. VCs, executives, researchers are selling this crazy story of a technology that's going to change everything. Now we have a big document with a lot of real numbers in it. And pretty soon we're going to know what the stock price of Anthropic is. So how would you characterize this? What is happening right now with the IPO? What questions are being asked and what questions are being answered? I mean, it is sort of like a demystifying moment because we have, you know, I've seen these companies in the private market get these trillion dollar valuations in such a short time. And it's kind of being set by like a handful of VCs. Right. And so, like, yes, VCs believe in the in this company, but they also believed in the metaverse and like a lot of crypto stuff. And that has not worked out. And so this will be the first time that we have like Wall Street and public market investors and, you know, pension fund investors, all the professional money managers, like, modeling out the financials of this company

and deciding whether it's worth investing at a $2 trillion valuation or maybe $1.5 trillion. And so that's part of it is, like, we're actually going to see, like, some real evaluation of the company. Although I guess you could also argue that Wall Street sometimes invests on hypermomentum, too, as we've seen from many stocks. So, like, who knows? But in general, this is, like, a time when we're finally getting, like, to look inside the business that's underlying the AI boom and decide whether or not it is a valuable contributor to our economy because there are a lot of people out there that think it's not. So what are you looking for in the weeks after the IPO? Whether or not AI kills us all. Big question still on everyone's mind. I doubt that that's going to be talked about too much on the roadshow. I mean, I'm sure that there are some advisors at Goldman and Morgan Stanley Horse who have been telling Dario, like, let's reel it back on the old extinction talk. I mean, could I invest in a company that has a 10% chance of killing everybody?

I mean, it's in the risk factor, so you can't say they didn't warn you. There was a Wall Street Journal headline arguing, the headline is, the bull case for A. Armageddon. The bull case. Exactly. The bull case. For investors mulling meanthropic share offering a 10% chance of extinction may be a 90% win for your portfolio. How can you argue with that, really? If you're going to die, you're going to die. But if you can... Die rich. Yeah, exactly. Erin, thank you for coming into the studio and talking to me at Anthropik. We're going to take a break, and then we'll be back with Mike Isaac and Jasmine Sun to talk about the rest of the week's news in tech. we're back with Aaron Griffith and joined by tech writers Mike Isaac and Jasmine Sun

um Mike and Jasmine we were just talking about the Anthropic S1 um Mike what was your big take away from the filing i thought it was like way front-loaded with rosier sort of paintings of ai at the top compared to the past you know like i you guys tell me but i feel like dario amade is painted as like the the biggest like doom and gloom guy of ai in the valley and my perception was everyone around him was like hey you need to like pump the brakes on this a little bit especially as we're getting the average person or trying to get the average person excited about it. So, like, the whole top was like, hey, by the way, we do want to cure cancer. We do want to, like, create these amazing drugs. And then, like, they back into, oh, and it has, like, a 10% chance of blowing up the entire world, basically. But I don't know. Someone told him in jazz hands a little bit at the top, right? Yeah, I mean, how much do you think it was, like, worked over? Like, I was interested that the adjective they use at the very top is, he says,

we believe that AI is one of the most consequential technologies in human history, which is a very funny, neutral adjective, right? Consequential goes a lot of ways, and it's not necessarily how I would be selling my product. We should do, like, a real close reading of all the language in this, because I do feel like Dario is, like, definitely a micromanaging control enthusiast, let's say. And, like, every word probably has a few different nested meanings in there for sure. Jasmine, what was your big takeaway from the S1, if you had one? Yeah, I mean, kind of similar to what Mike said, I was just thinking about how Anthropic has a very odd relationship to money and being a for-profit company. Like the founders, when they broke off from OpenAI, they spent a bunch of time debating whether they should be a nonprofit or a for-profit. Of course, they have invented some crazy multilatered governance structure, as they all do, which is something they reinforced a lot in the letter, that the founders have pledged 80% of their equity to charity, that employees have all this interesting matching. There's a famous culture fit interview

that all of the Anthropoc employees go through where they say, hey, would you work here if your equity went to zero? And they make sure that everybody says yes before they accept the job. And so mostly it just made me think about how weird they are about like the status of being a sort of profit-seeking corporation. I know, I can only imagine what the meetings with the bankers were like, you know. I mean, I assume the meetings with the investors are very similar. How do you tell a labor union that we're going to be post-money in like, I think that's a hard sell. This is coming, by the way, like in the midst of a little bit of anxiety and let's call it a jittery moment in the market. The elections are next week, and I really want to talk to you guys about what we're looking at and what we're waiting to see. Jasmine, what are you tracking for the midterms next week? Yeah, I mean, these are interesting midterms because I think it's the first election where people's feelings about AI, about data centers could plausibly have shown up. I mean, I think the top level on it is, even though we saw a lot of discourse, a lot of anger about data centers in particular

that's been building up over the last year, I did some reporting on it over the summer in Wisconsin and Michigan, two states where data centers were actually a really big factor in the primaries. In the majority of states, I don't think they are yet shifting people's votes. They're an 80-20 issue in the sense that people feel very negatively about data centers. People feel very negatively about AI. This is bipartisan. And in that sense, it seems like a unicorn. But it's also not something that is quite salient yet. I've looked at, you know, a million polls and AI usually ranks like 22 out of 39 issues. Oh, my God. It's like even in states like Ohio or Pennsylvania, where data centers are a big issue. I think it's usually something like between one or five percent of voters will list it as their number one thing. And so while we're seeing data centers really impact like your mayoral election, like we're seeing recall campaigns, city council reshuffles. I don't think like a big Senate race is going to be determined by data center sentiment. Right. It's like, I mean, I mean, this came across in the reporting you were doing that this is a really a municipal issue and it's going to get a lot of town managers voted out and

council board selectmen or whatever else. But it's interesting because this was, at least in the summer, it felt like this was a focus of some Senate campaigns. Some Senate campaigns were, in fact, putting a lot of resources into being anti-data center. Do you think that was wasted money or is it just not ultimately the issue that matters? I mean, I think it really depends. I think there are races where they're close enough, and rural voters in particular, you can get a lot of normal rural Republicans or low-turnout voters to maybe turn out if there's a data center in their backyard. I think the races that people are really concerned about, Sherrod Brown in Ohio might stage a comeback in particular because of data centers, or a big motivation for Greg Abbott to restrict approvals and connections to the electricity grid, even though he's been very, very pro-built, he's been welcoming the data centers for years, but because of a potential threat from both his Democratic opponent, plus all of these rural Texan farmers who are like, what's up with these power lines and data centers? He actually did a pseudo-moratorium, which, because he was worried about the midterms.

So, you know, I think there are a few elections in purplish states where data centers could swing things. But the other thing I'm tracking is, like, data centers are very concrete, whereas AI generally still feels so abstract to people that I'm not really sure if, like, data centers are barely swinging some elections. I'm not sure that, like, AI in general is going to be a voting issue yet, even though some people suspect it. Yeah, I mean, it's a very funny, you know, as you say, if you're on Twitter or if you're, like, reading too much about this, it does feel like a life or death issue. Certainly it is for some people, some people writing, you know, introductory letters to S1s, but it hasn't, as you say, it hasn't really trickled down, except in a sort of indirect way through the effect on the economy of AI. Mike, what about you? Are you tracking anything in the election next week? You talked about this in last week's episode a little bit with the head of policy at Meta leaving and sort of imagining which new regime of like you know democratic operatives they might replace them with But I just want to see if there is a white No of course I am a loyal hard fork listener

That's right. We've got one. You've got one. Let's get to two by the end of this. I'll get my wife to listen. I do. I really, really want to see the theater of all the tech companies that have been trying to cater to Republicans suddenly be like, actually, we love the new democratic, potentially democratic regime in all of these states. And because they have a hold on, again, potentially have a hold on Congress. This is what we really believe in. You know, I think these swings are valuable because it sort of shows, you know, these are corporations that don't necessarily believe in anything except self-preservation a lot of the time. Yeah. I mean, what do you are you seeing any indications besides, say, so we're talking about Joel Kaplan, who is the head of policy at Facebook for 15 years, has been a sort of conservative whisperer. He announced his departure last week. Are you seeing any other kind of movements that suggest a little bit of, I don't know how to put it, team switching or sort of trying to make inroads among Democrats? I think all of the companies are looking hard at their D.C. teams and saying,

who do we need to sort of elevate and put out front? They're never going to cut all of their conservative sort of relations people or whatever. But it's sort of like, actually, look at this guy, the guy who wears a blue pin instead of a red one or whatever. That's like that is that's all the companies. And like, that's just the game. I guess I shouldn't be so cynical, but it is worth looking at. Why wouldn't you be cynical? This seems like exactly the right thing to be cynical about. Yeah. And do you feel like there's a there's a shift? I mean, I think we all are aware of what felt like. I'm sorry to use such immaterial language here. A vibe shift in Silicon Valley. is there a vibe shift swinging back? I mean, just if any of you feel like that might happen, or was the first vibe shift an illusion? I mean, I've seen headlines to that effect, but I think, like Mike said, it's going to be a little bit more on the D.C. side of things because, I mean, just from my perspective, a lot of the VCs were really excited to be able to not care about stuff

that Democrats care about, like DEI. culturally it seems like we're kind of going to be still in that vibe. Yeah, it doesn't seem like that's changed. Jasmine, we were talking before we were taping about like a sense that there's a question about how Republicans are going to react to this. They're likely to lose the House of Representatives. There's going to have to be a shift in how they approach AI regulation in Silicon Valley. Totally, yeah. I mean, this is one of the big open questions I have about the 2028 primary too. I mean, primary season is going to start basically two weeks from now. And are J.D. Vance and Marco Rubio going to break from Donald Trump's extremely, extremely unpopular positions on AI regulation and data centers? Right. Like in spite of the vast like the majority of Republican voters do want to regulate AI. They would prefer slower with more safeguards versus some kind of abstract immaterial beat China narrative. and yet basically every Republican in Congress or every Republican leader has been unable to say any

of that because the president sort of sets what the party line is and if the Republicans take a beating in the midterms I could imagine some more folks deciding to break off and say like actually like our constituents are angry at data centers they're worried about kids safety if they're social conservatives but I am starting to see a bit of a split in the Republican party between the more traditional libertarian free market letteric types who are still very excited about deregulation versus some of the more new right folks. I spent the weekend with some folks from American Compass, the new right think tank, and they're all thinking, actually, maybe after the midterms, this is the time for the Republican Party to get serious about AI regulation on kids, jobs, safety, things like that. Were they saying AI or were they saying SI? It was actually really funny. It was a chat in the house workshop, but there's a split in the room, and you could tell there was a bit of internal tension between the folks from the admin and other places who were saying SI and the folks who were a little bit too staunchly, like, we're not caving to that.

So it was really fun. I've never seen a better, like, metaphor for, like, people talking over each other and, like, not being able to communicate than, like, not even being able to agree on what we're calling this thing. It was kind of amazing. Yeah. So one of the issues on the ballot next Tuesday is this tax in California on billionaires. Erin, can you explain what the proposition is, what the measure is? And I mean, I think we can all guess, but why is Silicon Valley so interested in this tax on billionaires? Yeah, I mean, it's been actually pretty fascinating. Prop 40 is a proposed one-time 5% billionaire tax on wealth, including unrealized gains. That means private startup stock for billionaires in California. And as you can imagine, the billionaires are a little bit mad about it. And so what better way to fight back than to use their vast wealth? And so they've gathered, this group that we reported on, Building a Better California,

has gathered $333 million, which is so, so, so much to fight a state proposition, mostly from, you know, Sergey Brin, but then he brought in John Doerr and Patrick Collison, the founder of Stripe. John Doerr is a very famous VC investor from Kleiner Perkins. And so, and they have been, over the last couple months, recruited this like kind of who's who of Silicon Valley, like very high-powered, famous VCs, founders, many of whom are not even billionaires. There's some like mere hundred millionaires. Temporarily embarrassed billionaires. Yeah. Yeah. Yeah, and, you know, aspiring billionaires. But they're mad about this for a number of reasons. You know, they think it's poorly written, that they don't believe that it's going to be a one-time tax. They don't like how the money's actually going to, like, the money's not, you know, specifically earmarked well enough to do the things that they say they're going to do, and they basically just don't trust the politicians to do the right thing with it. And so they're really mad about it, and they don't like to press it in a sense, because California is kind of known for, you know, being a trendsetter.

And so they're worried that it's going to turn into this cat and mouse game where other states are going to do it. And because a lot of them have already left the state, but they're mad about essentially the principle of it. The principle of the matter. Yeah. Yeah. They believe that, you know, this is going to hinder aspiring billionaires, founders, builders from moving to California and wanting to, you know, get rich. And so I think, you know, there's a lot of reasons that they're fighting it, and they're all kind of, like, banding together and getting on these weekly calls and, like, talking about their strategy. And, like, one of the really interesting things that we found in our reporting was, like, they are aware of how it looks. There was this quote in one of the documents that they're circulating to other donors that says, billionaires are seen as having a net negative impact across issues. So they're not trying to be the face of this. They don't want to be out front. They're not like, you know, talking publicly about this like they might if there was some other like, you know, any issue, like including data centers.

They are trying to go through other groups that they're partnering with, like Planned Parenthood is a really big one that they're working with to kind of that's opposed to it. And so I sorry, why is Planned Parenthood opposed to it? I mean, for some of the same reasons, it's not sustainable. This one-time tax thing is supposed to make up for some of the Trump administration cuts, and they don't believe that it's not setting them up in a position to be sustainable going forward. They just think it's poorly written. I think a lot of the people that are opposed to it are not opposed to the idea of a wealth tax, including actually some people that I spoke with for this story. They just think this is a crappy one, and they wish that it was better, that it was just structured better and made more sense. But do you feel like they're always going to say that? I feel like it's always a, well, this bill is poorly written, and if they could only legislate it better. I just feel like I've heard that from Andreessen or whoever a million times, you know what I'm saying? Totally. I mean, it's almost sort of like respectability politics where it's like, I agree with your message, but I don't like how you're doing it.

Yeah. I was interested to learn that about Planned Parenthood, though, because I hadn't realized that this was more than just a cabal of billionaires plotting against it. No, there's like a bunch of – I mean, ever since I wrote this story, Very weirdly, my phone has been blowing up with political text messages from different groups, and there's a lot of health care groups that have lined up against it. That's interesting. Yeah. I mean, from a Democratic point of view, this is a nonpartisan podcast, of course. From a Democratic point of view, it's sort of nice to put one of these ballot measures on every year so that the billionaires all get distracted and focus on that instead of putting money around the country the way Elon Musk is. That's pretty good game theory. Yeah, there you go. Feel free to call me, you know, Senate Local Action Committees. There's one other big thing I want to talk about today, and that is what is being called Black October, the vuln-apocalypse, the PHP vulnerability discovered last week. Mike, I'm going to put you on the spot. Can you explain? Obviously, we all know what PHP is. We know what's going on. But for anybody who doesn't know this stuff, can you explain what happened last week and why it made everybody who works in information security's life hell for the – I mean, it continues to be hell?

Well, look, I think the way I like to think about it is how the entire, let's say, Internet and infrastructure that computers run on to this point is pretty Byzantine and built on a lot of old code that has, let's say, a lot of holes and stuff in it, right? And that's just the nature of computing. You figure out what are called bones vulnerabilities, and you fix them over time. You know, Microsoft has dozens of fixes to their main software a year. Now it's ranging up into the hundreds because AI is finding more of them. What happened, though, Black October is not trying to get a Furby for your shopping for your kids. Mike, come on. Was that on the spot? You've got to do better than that. I hope you didn't workshop that. We're on squishies now, Mike. It's dumpling squishies. My son did his birthday wish list this week, and dumpling squishies, that's the big thing. I'm showing my lack of children on this podcast. You're showing your age as you're showing, brother.

But it is essentially a vulnerability that goes far wider than just any one little system tweak that you have to update your individual software on. This is stuff that, you know, folks in the industry said could break the Internet. It could, like, affect how we get on websites and use websites and have them be – have our accounts take it over. You know, it was like – it's much wider. My big takeaway is, like, this is the type of stuff that's going to be much easier and more commonplace to find in the age of AI where you can have these programs just scrape the Internet. We should say specifically what we're talking – so PHP is a programming language that – that – god, this is – every time I start talking like this, I think about the dozens of men with huge beards that are going to be twirling and emailing me. The NIST thing. Do you see why I'm so smart? Pond it off with you. Half of them are my friends. I'm thinking about specific people who are going to be texting me about how wrong. PHP is a programming language that's used on a ton of different websites. And as Mike is saying, it's much easier to find vulnerabilities in these sort of infrastructural languages with AI,

which can sort of run a bunch of different ideas and figure out what the exact problems are. And when that happens, all of a sudden you have these really intense, really serious vulnerabilities that now need to be fixed, not from the top down, but across a bazillion websites. And so, you know, I don't want to make this podcast about me, but on one day last week, I do a daily logic puzzle called Clues by Sam. And I wasn't able to record my score that day because Sam was fixing the PHP. The real victim. Well, you know, I'm just going to bring this down to earth for everybody listening so they really understand what's going on. Imagine if you couldn't do your Wordle every day. That is what you need to be using as an illusion. I mean, are you guys talking to people who are in the midst of this? I'm interested in whether this is either from the perspective of cybersecurity people or from AI people in general, how much we think this is just the new normal, that every month is now just 85-hour weeks for people who work in security and all the AI companies throwing up their hands and being like, we told you so.

I mean, I love to hear what you guys think, but I think the flip side or the counter argument, and this kind of happened in the open versus closed source debate a few weeks or months ago, was like, well, guess what? AI is also going to enable us to solve all these problems and to quickly defend against them and to fix them faster. And these vulnerabilities already existed. It was only a matter of time for someone to find them. So there are arguments from folks saying these fixes can happen. We just sort of need to enlist the advanced tools we have now to fix them faster rather than you know random websites across the Internet not being able to do that in enough time But like I don know that feels a little rosy to me I do feel like we going to just see this more often And the old internet breaking much faster than before is my point of view, at least. I mean, one of the crazier news stories from this week that I thought about was that startup cathedral made of ex-Doge sapphire,

so our nation's best and brightest, literally hacking into Anthropics, Slack, and IT systems. And Anthropic is supposed to be the ones that are building the best cyber defense internal models that they can't even release to the public. And yet their own technical systems were not secure enough to be able to block this crew of like young startup guys from accessing trade secrets in their Slack. And so I don't even think, you know, before I would have thought like, man, some, you know, municipal governments are going to really struggle if they get hacked by some people with an open source model. But now it's actually like the best and most AI forward companies in the world are also as vulnerable. So that's a great point. It's going to be a very long few years, I think. I mean, I think the argument, I read a great piece that I want to recommend by Joshua Sachs, who's a former meta AI guy. And his argument is basically that we are, as bad as it seems, we are still in a sort of defender's window, a defender dominant window.

that even though there's all these vulnerabilities being discovered, in fact, I mean, his point, which I admire because I think it's true, is that it takes cyber criminals actually a very long time to, like, figure out how to do their jobs because, generally speaking, they're not that smart and they're not that competent at, like, running businesses. So we have, you know, a certain period where we can lock down defenses and, you know, prevent, certainly on the level of cyber crime, protect against some of the worst vulnerabilities. But he says, on the other hand, you know, this is not a wholly positive story. You know, we're still thinking about nation state actors that have the resources and the will and desire to, as Jasmine says, like hack into municipal government infrastructure and whatever else. So I suspect this is not the last time we're going to be talking about AI discover vulnerabilities on this podcast. Great. That's right, exactly. Go back to Hawaii, Mike. There's nothing left there. The last thing I want to talk about is the sort of,

it felt like it flew under the radar amidst a bunch of other news this week, but Microsoft offered biopackages to 60,000 people. Erin, what did you make of that? Well, I mean, that is an enormous number, and when you put it in perspective of the fact that Microsoft has like 230,000 employees, it just reminds you of how enormous these companies are. But just to be clear, buyout packages doesn't mean 60,000 people are going to take them. They expect 3% to 5% of their employees to actually take them. So in reality, it's going to be like at most probably 3,000 layoffs. But that's still so many. And I feel like, to your point, Max, over the last three, four years, we've just become so numb to these rolling layoffs from the giant tech companies because that is just a new reality that started kind of after the Zerp era kind of deflated. All these tech companies that got super bloated because they thought they needed a million engineers decided, okay, now we're just going to kind of be in this environment where we just cut 1,000 people every six months or something.

And tech used to be this cushy job, especially at a place like Microsoft, which is very geographically focused, where you can just, like, have a 30-year career. And there are a lot of people at companies like Microsoft, Amazon, you know, even, like, Google and Facebook, which are a little bit younger, that have been there for so long. And that is no longer the case. That is, you know, maybe a sign of what's to come for the rest of the economy because tech is sort of always a little bit out ahead on workplace trends as well. So that's a little bit depressing. And I guess the other thing that I'll add is, like, something that is, like, kind of darkly funny about the announcement, which is that Microsoft said it was part of their efforts to give employees more control over their future. Oh, what? That's not the kind of control I'm looking for. That's not the control I'm looking for. Really, really, really grim. The silver lining in there, I would say, as far as your, like, harbinger of what's to come, is, like, a lot of this is because of their AI spin and infrastructure spin.

And so, like, perhaps, like, whatever, Accenture or some, like, giant employer that's not spending billions of dollars per quarter on AI is not going to do the same level of layoffs. They're just going to use the tools to justify different types of layoffs in lesser numbers. Oh, my God, Mike, Accenture is the worst example because, like, if there's any company in the world that's going to be fully taken out by AI, it's consulting, right? I'm sorry, good lord. See, I'm going to get taken out by AI because I can't even think of a good company. No, actually, your mistakes, that's what makes you human, Mike. Thank you. That's what people love. Okay, on that cheery bit of news, we're going to go to break. When we come back, we're going to be talking about stories that we're tracking this week. Welcome back. I'm with Aaron, Jasmine, and Mike, and we're going to be talking about stories that we're tracking this week. Mike, what story are you tracking this week?

So I, this is kind of nerdy, but that's my thing. I found this memo from the VP and distinguished engineer at Amazon this week that, you know, most normal people are not tracking. But I think it's really interesting because it's essentially a very long manifesto on the use of AI among software engineers inside of the big companies. And what they essentially cop to is token maxing is kind of a terrible metric to measure how you're doing as an engineer, as is LOC, lines of code, the amount of code you are contributing building these products. Also a messy metric, which basically means you're larding up these systems with a bunch of stuff that someone's going to eventually have to delete. And so in this long memo, they essentially tell employees, look, what we care about are long-term outcomes. This is like what really matters in the end, and this is what, you know, we need to be judged by. The problem is you can't really judge long-term outcomes on a quarterly basis when you're trying to figure out who should still be employed here.

So I think after, you know, thousands of words, we end up kind of back where we started, and I'm very curious how they judge these engineers in the coming quarters. I mean, that's so interesting because I feel like when the token maxing trend started, everyone with any common sense was just kind of like, this seems absurd. Like just spending as much money as possible to write as much code as possible. Like playing the volume game just obviously that's going to end badly. And we kind of all knew it was a short-term bump. And the token mining or whatever they call it, token mining, the reverse of that trend happened so fast. Like, by the time we could barely get our first story up, it was already reversed. But, like, it seems like they're still struggling to figure out where the balance is. Is that, like, kind of your upshot from this? Well, what better metaphor for the tech industry than spending billions of dollars to find out something is a terrible idea and ending up back when you started? They're in the arena trying things, Mike. Yeah, exactly. Yeah, that's true. I am a coward.

I mean, this is, I have to say, you say this is nerdy, but this feels like the most relatable possible thing to me. everybody has a job where they introduce a new metric and you know the metric's stupid and you fulfill the metric and then your bosses are like, actually, we're going to do the opposite metric. And so you do that and you're like, this is still stupid. And then they're like, then they're sort of like, actually, our metric is the long-term health of the company. And you're like, well, I'm still just going to do my job. You're self-employed. What are you talking about? You are your own boss. Well, I am also self-hating, so I think I can. Jasmine, what are you tracking this week? Oh, man. I saw that this new research report that a team at Anthropic put out, they did a randomized controlled trial trying to understand how much experts defer to AI systems. So they basically gave a bunch of professional statistical analysts access to AIs to give them advice on a decision in a scenario. And what they found was that not only did the experts, I think over 80% of the time, make the same decision as the AI, which are split half-half. they also over 70 percent of the time use the ai's arguments to justify their decision but only 15

percent of the time did they say oh yeah i was influenced by this ai agent and so you we have this world which i think is one that we're going to be in increasingly where humans are oftentimes almost becoming like rappers on ai agents and ai systems where like all we're doing is like seeing the faces and like typing the thoughts and like i feel like what like there's all this concern about AI takeover and rogue AIs. And I think that a lot of the times we're just going to let the AIs take over because we'll just be outsourcing all our decisions and we won't even realize. So that's what I'm worried about. That is a cheery thought. I mean, as somebody who professionally takes other people's opinions and repeats them as though they were my own, like I do, I do kind of sympathize. It's just usually the people I'm stealing from are people, not, you know, not artificial intelligence. People say that like AI makes everybody a manager, Like, here we are. Yeah, exactly. That's exactly what managers do. Wow. Erin, what about you? What are you tracking? So I cover venture capital, as you know, and one of the most famous firms, Andreessen Horowitz,

recently made a big change to the branding and the name of one of their sub funds, which is their A16Z crypto fund. They're one of the biggest crypto boosters and have been for many years. But now they are going by A16Z frontier. And this was pretty notable to me. I mean, Hard Fork listeners might remember Chris Dixon, their top crypto investor, coming on for an interview a few years back talking about, you know, the importance of the blockchain and crypto for the future of the Internet. And so this would kind of make it look like they're backing away from that promise. promise and that's kind of a thing that happens in VC a lot where, you know, VCs chase trends and they change. And, you know, it's not just a trend chasing business. It's also a marketing business because all money is kind of the same. And so in order to stand out and get the entrepreneurs to come to you, you have to kind of put a stake in the ground and say like, this is what we're interested in or this is what we're really excited about for the future. And so they were really,

really big bulls on crypto. And they, of course, despite the fact that crypto is super out of vogue right now and that something like 80%, 90% of all VC dollars are basically going to AI, they have been very careful to say, we're not backing away from crypto. We're just expanding our mandate. It's just all frontier now. Yes, yeah. And this is something that other crypto funds have also been doing. They're not alone in this. There's a fund called Paradigm that was started by a Sequoia investor. In 2023, they very famously removed some of the crypto language from their website, and people noticed and freaked out. I will say a big part of it is that there's probably just not that many people building crypto, investable crypto startups right now because everyone wants to build an AI. So it kind of is necessary. Like the crypto industries, like one of their many problems with crypto is that without the hype, without the number going up, there's just less people that are interested in like working in this industry. and building on this industry.

So, like, in order to find the opportunities to, like, deploy billions of dollars into startups, you have to kind of, like, change your mandate. Yeah. When crypto is too toxic, even for Andreessen Horowitz, that's when you know the vibe has shifted. I mean, here at Hard Fork, we embrace our crypto nominative words. That's right. Exactly. Our roots are in the blockchain. We'll be sticking with this name forever. Good Lord. Wow. Well, guys, thank you for coming through. This has been a really fun discussion. I'm so glad to have you, and I hope you all come back soon. And this has been Mike Isaac, Jasmine's son, Aaron Griffith. I'm Max Reed. Thank you for listening. This episode of Hard Fork was produced by Whitney Jones and Sawyer Roquet and edited by Viren Povich. with additional editorial support from Brendan Klinkenberg, Lisa Tobin, and Brooke Minters. We were fact-checked by Will Peischel

and engineered by Daniel Ramirez. Original music by Marion Lozano, Alyssa Moxley, and Dan Powell. Additional video production by Chris Schott. You can watch this full episode on YouTube at youtube.com slash hardfork. Special thanks to Sam Dolnick, Jessica Testa, Lu Wang Tam, Matty Maciello, Dave Mayers, Luke Mylander, Thomas Trudeau, Samantha Winter, Kyle Grandillo, Michael Cordero and Dalia Haddad we love to hear from our audience if you have any questions or comments or want to give us a piece of your mind find us at nytimes.com slash hardfork you can also email us hardfork at nytimes.com