Computer Says Maybe

Alix: [00:00:00] Welcome to Computer says maybe in this episode I'm sitting down with Sarah West from ai. Now to get into something that I think we talk about in generalities, but don't necessarily get into the nitty gritty specifics of an extremely important aspect of ai, and that is the kind of money and deals that structure the marketplace given there's.

Alix: An extremely small number of companies and people with the kind of cash necessary to keep this thing alive. Getting into how those different entities relate to one another and the different deals we're seeing, basically every day there's a new giant number being thrown out there, and Sarah has spent, I would say, like the last few months, getting into the very nitty gritty on all of this.

Alix: I thought I would take the opportunity of being in New York for Climate Week to sit down and basically just like talk numbers. And talk money and get a little bit clearer on what is going on. So thank you Sarah.

Sarah: What is going on? It's been such a weird moment and I feel like the deal that people are [00:01:00] most familiar with is the relationship between open air and Microsoft.

Sarah: Yeah. But like that's really just the tip of the iceberg here. Yeah.

Alix: It feels like a, a lifetime ago when, when that sort of first was the kind of, I dunno, the origin story of how all of these very, very large companies were. It felt like propping up. Something for us. But let's start like with these deals, I feel like literally every day there is some giant number being pitched by a government, by a company, by a hardware company, or one of these generative AI companies like OpenAI.

Alix: Do you wanna talk a little bit about how these deals are being pitched and shared in the media?

Sarah: Yeah. Yeah. So I mean, I think the important context here is that the leading AI companies in the United States have really gone all in on the idea of building out AI at huge scale. That means that the kinds of investment figures that are being announced are really starting to become staggering.

Sarah: So one of the most notable deals [00:02:00] this year was the Stargate announcement, which was made in the White House between. OpenAI, Oracle and SoftBank, which is an investment firm, that deal is now at $400 billion. That's going to be invested into building out the data center infrastructures needed for OpenAI to build larger and larger and larger AI models.

Alix: Yeah, I mean, I feel like even just $400 billion is like a mental, I, I find myself increasingly unable to really conceptualize the types of numbers we're talking about because that is just gigantic. And I feel like Sam Altman kind of set the tone. It was like a year ago when he was like, yeah, just like, give me $7 trillion and then we can like do what needs to happen here.

Alix: So I feel like the whole environment is just bigger and bigger price tags. And I found myself trying to relay the news about NVIDIA's. Quote unquote investment in OpenAI and thought it was a hundred million dollars because when I said a hundred billion, I was like, well, there's no way it could [00:03:00] be that big.

Alix: And it is. Is this, is it real? These deals, like are they actually happening once they get announced?

Sarah: Earlier this week there were some consultants at Bain and Company that estimated that this wave of AI infrastructure spending is going to require. $2 trillion in revenue by 2030 to be able to support it.

Sarah: That is more than the combined 2024 revenue of Amazon, apple Alphabet, Microsoft Meta, and Nvidia combined. It's also more than five times the size of the entire global subscription software market. In five years. In five years.

Alix: Yeah. And if that doesn't happen, then that means that all of these investments won't be recouped at a pace that actually makes them worth it.

Sarah: That's exactly right. So a lot of these deals are made on the presumption that if we build it, they will come. We just need to have the right infrastructure in [00:04:00] place because eventually the market demand is going to catch up. But what we're finding is that these companies are underestimating the amount of their.

Sarah: Expenditures and they're overestimating the amount of revenue that they'll be able to bring in. So OpenAI in its most recent earnings report, announced that it had underestimated a couple months ago. Its capital expenditures by. $80 billion, God. And then by contrast, its revenue was only $12 billion. So it's orders of magnitude under the scale that it will need to be to be able to recoup this kind of investment, and that's why they're turning to companies like Oracle, SoftBank, and Nvidia to help float them.

Sarah: In the meantime,

Alix: let's double click on two of these big deals to like get into the mechanics of how they're constructed. Let's talk about Oracle. First you just brought them up. They are one of the main. [00:05:00] Funders of Stargate or vendors of Stargate, or how sort of, how does Oracle connect to the Stargate announcement?

Sarah: So Oracle is a legacy cloud service provider. They're, you know, when you think about like the big cloud companies, they're always like the distant fourth. You think about Google, Amazon, Microsoft as the core of the cloud market, and Oracle is the one that's kind of been hanging on. And so they. Shocked the market earlier this year when they announced that their earnings were phenomenally bigger than expected, and it's because of this deal that they've struck with OpenAI Earlier, Oracle was projecting that its revenue would be $20 billion by 2030.

Sarah: They've now revised that projection in their most recent earnings report. Post striking this deal with OpenAI now they're expecting their revenue is going to be $144 billion by the year 2030, so it's this huge jump. The markets just rallied [00:06:00] in response and made Larry Ellison now the richest man on paper in the world.

Alix: Yeah. It's also at a time when there's been increased reporting, specifically a new statesman piece this week on how much money Ellison has put into the Tony Blair Institute as part of what some people inside the Tony Blair Institute see as. Him essentially pitching Oracle services to governments around the world.

Alix: So Tony Blair coming in and making this very glossy case about the importance of digitizing government services, the importance of modernizing economies, and then just so happens to also be ready and willing to say. And when you do that. You should probably also buy your services from our benefactor, Larry Ellison, which feels very connected.

Alix: And I wanna talk a little bit more about what this Stargate might be, but like do you have thoughts on the extent to which some of these deals are so adjacent to governments and nation states who are making pretty big bets in this area?

Sarah: Yeah, so like I said, there really does not seem to be [00:07:00] a clear guarantee of revenue coming in from traditional enterprise markets.

Sarah: In fact, the Census Bureau released a report a couple of weeks ago that said that many businesses are in fact slowing down their adoption of ai. And an MIT report said that many businesses are finding this technology to be less useful, less meaningfully contributing to their productivity than they expected.

Sarah: So while the traditional business users are slowing down, governments are going. All in. So in the uk, in the US, and really around the world, we're seeing governments striking deals with AI companies to procure their technologies. They're introducing mandates to, you know, integrate AI into government services left and right, and especially in the defense sector, we're seeing the rise of the rebrand of companies like OpenAI and Meta.

Sarah: As defense contractors as well. This [00:08:00] guarantees them that they're going to have these really lucrative contracts in place and contracts that extend on five year terms. They're lengthy contracts that sort of guarantee stability in their revenue. Mm-hmm. To cover up the fact that in the traditional business market, nobody finds this stuff very useful at all.

Alix: So like the direct to consumer type sales or the business to business sales just aren't there. So you said something to me earlier this week, which I feel like may have actually triggered us deciding to sit down and have this conversation. Yeah, yeah. Um, where I, I've been thinking a lot about in two, three years, when they're not coming up with the $2 trillion of sales that they need to actually make this.

Alix: Capital expenditure that they've overextended themselves with worth it, are they gonna be bailed out? And I made some offhand comment about we should make policymakers get on the record and say, we're not gonna bail these companies out. We're not gonna accept the moral hazard of them growing in this way with a very immature technology and sort of pushing it on us.

Alix: And then when it doesn't pan out, that government comes to the rescue. And [00:09:00] you said. That you think that they're already being bailed out. Um, do you wanna say a little bit about what you mean by that and how you're thinking about it?

Sarah: Yeah, so I mean, it's very easy because we've seen bubbles burst just left and right over the last couple of decades.

Sarah: That's the familiar story that a market gets overinflated, it disappears, and then the government bails it out. What we're seeing here is not quite the same as a bubble. We are seeing. A market that has gone all in on the most expensive way of building a technology. Failed to find a meaningful business model, but successfully persuaded key actors, including in government, to shore up their lines of business by making sure that they're gonna have these contracts in place by de-risking.

Sarah: Both the regulatory landscape through things like the proposed state moratorium on state's ability to enforce their AI laws, [00:10:00] rolling out the carpets for new sources of energy to come online. All of the things that would make a business more skittish and cautious about AI rollout are being guaranteed by the government to ensure that against all odds, that like rollout of AI is going to.

Sarah: Continue at a pace and scale that is really unprecedented.

Alix: Do you have a sense of why it's working when you, when I look at Kiir Starer and how, I can't think of another word aside from emasculated that he has been by AI companies and like American dominance in direct opposition of British people who don't want to be the data center backyard for American big tech in Europe.

Alix: Basically just offering up these extremely. Streamlined, environmentally ignorant policies around data center development for like tiny amounts of money. And also at a time when it is increasingly obvious that integrating technology into government services [00:11:00] oftentimes doesn't make them more efficient.

Alix: Oftentimes isn't something that people want or enjoy. Like do you have a sense of like, why is it working? What is the end game for these policy makers when they're thinking about shoring up these companies and making. The public use their products.

Sarah: This technology is directly harming people in the public, and we have so many examples of where that's been the case.

Sarah: There's a couple things. One, there's a clear revolving door. Between many of these governments and the heads of tech companies, so like the head of the Office of Science and Technology policy, it was formerly the head of a defense tech company called Scale ai. We have David Sachs coming in as Theis are.

Sarah: So there's like bested interests mm-hmm. That are directly shaping policy. That's one current. I think that there's also real vested interests in like. Propping up this corner of the market, the magnificent seven stocks taking up about a third of the [00:12:00] stock market. That's another contributing factor to the group think.

Sarah: 'cause if the valuations of those companies, which are like already very hot, slows down, then a lot of people are gonna lose a lot of money and they're just not willing for that to be the case. I think that there's that. And then if you look at the inner workings of what these technologies do enable, often it's enabling those who are in positions of power to like leverage asymmetries of information and asymmetries of power over others.

Sarah: So like the utility of these technologies, it's not productivity for everybody else, but it is a consolidation of power in the hands of a few people who already have it, and it cements them in place.

Alix: I also think there's a desire of center left politicians to be reasonable men on business.

Sarah: Yeah. Nobody wants to come in and be like, don't build [00:13:00] stuff.

Sarah: Don't be like, you don't wanna be having anti innovation agenda. And I think it's one of the real failures is to fill the gap with an idea of like, what is innovation and investment that works for. Everybody look like we're kind of stuck with this narrow-minded view that's come from the tech industry for like a decade and a half.

Alix: But I think that point that basically the tech companies are now a proxy for the health of the modern economy as a real, yeah, I don't know. I don't know what you're supposed to do with that. As a leader of an nation state that's relying on these companies to succeed as the primary bet for the economy, that must, like I hadn't, I hadn't quite thought about the pressure that that would put on a nation state to be like, what do we do?

Alix: We like, let. One third of our capital markets like fail.

Sarah: Yeah. And like for enforcers of like antitrust laws, for example, if you're like any single country and you're trying to enforce the antitrust laws, which let's give them the credit, we've seen unprecedented numbers of cases be [00:14:00] brought against these big tech companies.

Sarah: What kind of leverage do you have when the CEO of a company can go and flex and be like, well, we're just gonna leave your market if you try and bring a case against us. And what kind of penalties are gonna be meaningfully changing the behavior of companies when, you know, often a penalty is just a budget line item for a company that's grown to that kind of scale and size.

Sarah: Mm-hmm. It's a real fundamental small D democratic problem that we've landed ourselves in.

Alix: Yeah, totally. Okay. Well let's look at another deal, and I think this is the one where I was like. Is it a hundred million or a hundred billion? Oh, wow. It's a hundred billion. Yeah. Yeah. Um, do you wanna talk a little bit about this bizarre arrangement between NVIDIA and OpenAI that was announced this week?

Sarah: Yeah. So the Nvidia deal is a really unusual one, but it's a financing structure that is emerging in the cloud space. Nvidia has agreed that it's going to. Lease [00:15:00] its chips to OpenAI. So the OpenAI isn't reliant only on cloud providers like Oracle to build their data centers. OpenAI gets to build their own with NVIDIA's chips.

Sarah: They don't have to invest in owning those chips, which depreciate pretty quickly. They're able to lease it, and now OpenAI, as we know, doesn't have a whole lot of cash on hand, so NVIDIA is also going to loan them the money. In order to build out these data centers. Nvidia is going to be investing back into OpenAI as each gigawatt of data center capacity gets deployed and conveniently in a blog post.

Sarah: Just this week, Sam Altman said we're gonna be releasing a gigawatt of capacity in its new Abilene facility every single week. So every time they do that, they get a nice big check from From Nvidia, they get cookie. Yeah.

Alix: Do you wanna say a little bit about the Abilene facility? Actually,

Sarah: one of the big announcements around Stargate came this week when they announced that there's gonna be five different facilities [00:16:00] across the United States where.

Sarah: Stargate data centers are gonna be built and the flagship one, the one that they're starting to break ground on right now is going to be in Abilene, Texas, where there's gonna be just this huge, massive data center being built in a part of the country that's also very drought prone, which is notable given the amount of water that is needed to, you know, cool.

Sarah: These data centers. There was a big event where Sam Altman and Ted Cruz all sat together and talked about the launch of this, this new facility.

Alix: Okay, so all of this is predicated also on the fact that OpenAI finds a way to make money, right? But Nvidia then is creating conditions where the runway could be extended as OpenAI is building out its hardware capabilities in terms of its own data center infrastructure.

Sarah: Right. So there's no guarantee of return on this investment, but Nvidia has stakes in making sure that OpenAI is able to [00:17:00] continue doing what it does and kind of fueling demand in this market. And one way I heard this described is it's like invidia's, a parent that's loaning. The down payment on a home that OpenAI is buying that like they're kind of keeping it in the family, but this is, you know, more like, you know, the family in the mafioso sense than like your conventional family in that like they have this shared cartel like interest.

Sarah: It's kind of like that concept of round tripping in accounting where you have one firm that is lending to another firm, the money to procure an asset, and that money then goes right back to the original firm. You have no like net economic benefit coming out of that transaction. Mm-hmm. But you have the veneer of like a functioning market around an asset that basically has no real value.

Alix: And the purpose there is to inflate the amount of financial activity, [00:18:00] the perceived financial activity happening in a market that currently is artificially constructed. 'cause there's no customers.

Sarah: Exactly. This is the same principle that money laundering operates under.

Alix: Great. Um, and this is legal,

Sarah: uh, seemingly, I mean, I, the jury's out on, you know, the legality of this instrument.

Sarah: It's certainly happening.

Alix: I mean, I feel, I feel like this is where. There's like almost no, like the law of financial physics doesn't seem to apply to this sector. Money is being manufactured from nowhere to do these massive things. We're being told simultaneously that this is the future and also that there are no customers for these products.

Alix: I have a really hard time. Understanding how we got in a situation where companies that are making products that are very expensive and also have no one willing to buy them, are seen as the future of our economy. Do [00:19:00] you have reflections on how we got into this like narrative vortex of success and money and know products?

Sarah: We are in this world of like. Wild, magical thinking around ai, and I think part of it is that very few people are willing to contend with the idea that there is no inevitability to any of this. It's very hard once you've gotten cemented into the idea that there's this arms race that if we don't do it, someone else is gonna do it, so we better do it.

Sarah: Yeah. And then we will just tinker around the edges with trying to make this better for people rather than like contend with the core. Vision that none of this has to happen. We can just choose to do everything differently or another way. But because of that group think moment that we've gotten trapped in, it is getting harder to steer out of.

Sarah: And soon we're gonna have these data centers. Built without clear ideas on how to use them. And that's really the dangerous world. [00:20:00] 'cause then that's the environment where you start to get really toxic business models emerging and literally toxic in this case, because of the pollutants that these data centers are producing

Alix: and how all these data centers are competing on strained grids for limited pool of renewable energy.

Alix: And basically now the new meta facility in Louisiana, they've just turned on three natural gas. Plants to be able to power this facility. And I don't think that many people realize that we're moving backwards in the renewable energy revolution. I guess because we've basically decided to build out all this infrastructure that's so energy intensive that it's kind of destroying, I don't even know what the electrification of our, the renewable verification, if that's a word, um, of, of our grids.

Alix: The, the

Sarah: green transition. The,

Alix: that's a much better way of saying

Sarah: Yeah. Yeah. No, I mean, I think it. It's so true, and I think there was an executive order announced or forthcoming, maybe it's already signed. That creates penalties for turning. Off [00:21:00] existing forms of energy production that are dirty because of this new demand.

Sarah: Mm-hmm. So like one read of what's happening is that AI is giving a convenient reason. For the fossil fuel industry to be able to fabricate, demand and keep itself online and going for longer, where otherwise we would be in the full throes of transitioning over to renewable energy. That has been long, long overdue.

Alix: Way to go. Us. I mean you, you touched on this a little bit, but like the magical thinking piece, but also the like creative accounting that's happening in this space. And I'd love to talk a little bit about the specific mechanisms by which some of this investment is happening. 'cause I feel like it is complicated and there's all kinds of different sort of forms of it happening.

Alix: And just like how these relationships. There's obviously like private lending, there's some like government subsidy, which seems kind of limited. There's [00:22:00] some debt financing, but like how do these companies borrow, like how do you have the credit rating as open AI to actually be able to borrow debt?

Alix: Because the promise of this is so tenuous that I can't really imagine a more risk averse lending facility of a bank, for example, is gonna like allow you to take on debt. But yeah, how like do you have, what are your thoughts on like how these different mechanisms and forms of investment are kind of exoticizing some of these relationships?

Sarah: So we're seeing some of the shadier parts of the. Debt market lighting up now. I was actually reading in the information earlier this week, and they're like a publication that really focuses on market dynamics. It's written for VCs, and they described OpenAI as a. Risky company to lend money to, but Oracle is seen as a safer company to lend money to.[00:23:00]

Sarah: And so within that Oracle deal, they are gonna be turning to private credit markets to help them finance mm-hmm. Their data center build out. So like most of what we've seen so far has been. Cash Rich Cloud Giants like Google and Amazon that can pay for the data centers that they're building with the money that they've gleaned from their other lines of businesses, right?

Sarah: Oracle doesn't have that kind of cash on hand, so they're turning to companies like Apollo Global Management, which has among its lines of business life insurance, and so they're gonna be likely leveraging life insurance in order to be able to issue debt for. The data center build out for which there's no guaranteed revenue, like there's these layers and layers of shakiness starting to emerge within this market.

Sarah: We've also seen stuff like Core Weave, which is a very small cloud firm. They started out [00:24:00] in crypto mining and then converted themselves over into. Being a company whose core business is selling itself back to the cloud giants when they need extra capacity. A lot of their revenue comes from Microsoft actually.

Sarah: So Core we've got a huge investment from Nvidia and then turned around and leveraged its preferred access to NVIDIA's, top of the line chips to get debt from Blackstone and other financiers with the chips as the guarantee of that investment.

Alix: So wild. Yeah. Is this, I mean, this is all, these are all American companies.

Alix: Is this kind of horse trading and magical economic thinking happening in other markets?

Sarah: I know that there's been a lot of activity among VCs in Europe, for example, where there's a lot of interest in like building their own, but I don't think that we see quite this level of [00:25:00] creative financial practices emerging in other markets.

Alix: Interesting. So. You said earlier that we're kind of familiar with the traditional bubble dynamic where things get over evaluated and then investors get spooked in enough critical mass that like the whole thing comes crumbling down and there's always something left over at the end of that. Like I think when you think about the.com bubble, like we still had the internet after and actually that wave of companies became really important for the next wave of companies.

Alix: I mean, I, I don't wanna put you in a position of prognosticating 'cause it's so hard to guess what's gonna happen with this stuff, but do you have thoughts on. What might happen once the kind of dominoes start falling of some of these arrangements and like what might this capacity be used for? Like how might this market evolve when forced to,

Sarah: again, I don't think that we're gonna see like a moment where the bubbles bursting.

Sarah: I think we're in it already. We're seeing these weird dynamics where there isn't inherent or underlying. Economic value that's [00:26:00] being mm-hmm. Created outside of the fact that the build out in data centers is holding up key parts of the economy. Mm-hmm. It's, it is a huge part of the stock market. It's more valuable in the US economy than consumer spending is right now.

Sarah: Wait, wait, wait, wait. Say that again. Yeah. It's contributing more to GDP than, than the totality of consumer spending,

Alix: so. These big swing AI investments and like buying and selling and like horse trading just within the AI sector is a larger portion of GDP than all of consumer spending.

Sarah: Yeah. That's insane.

Sarah: It's insane, right? Yeah. But like it, but these insane dynamics are the new norm. Like it's not like there's gonna be some like come to Jesus. Like everybody all of a sudden has the Yeah. Wool taken off over their eyes. Like this is where we are now. [00:27:00] What we're gonna have to deal with is what? Been dealing with, which is that there are these profoundly powerful centers of power that are intermingled with the state and getting closer and closer over time.

Sarah: We're also seeing the US government taking stakes in companies that are seen as. Critically valuable. Maybe we didn't even talk about that. Yeah, we didn't even talk about that. But that's another huge factor here is like, you know, are some of these firms gonna be nationalized? This is a core part now of the economic and political reality that we're living in, and I think what we really need to understand is what it means for the distribution of power and what it means for our economic lives.

Sarah: Wholesale,

Alix: I mean, are there like. Like I'm thinking about dominant economists that have said interesting things about how all this is coming together and I'm thinking about like power and progress, for example. Yeah. That basically anytime there's a new technology over the last thousand [00:28:00] years, you, you will see the rich get richer and the poor get poor unless there's like very meaningful interventions that prevent that from happening.

Alix: But I feel like most of the economists I hear from now are like thinking on the margins about this stuff. I don't really hear anyone talking about. How new this is and the risks associated with that. Where could we go from here?

Sarah: That's the real question, right? I mean, I don't think that any kind of tinkering around the margins is going to meaningfully perturb this trajectory that we're on.

Sarah: We have to be centrally contending with these questions about what is the world we wanna live in? Do we want our institutions to be overtaken by this confluence of tech, corporate, and state actors? Where is the public? Where is the interest of the public? What does this mean for [00:29:00] democracy and the possibility of power that emanates from below?

Sarah: It is at that ground level existential question that we really have to be grappling with, and I do still believe that there are points of leverage available to us. This is not the world that we have to be mm-hmm. Entering into and we can't. Put up our hands and, and sit back because it is a future that's actively being built.

Alix: I was reading, um, ed Tron had a piece about what we're gonna see. If the bubble and he uses the frame of a bubble. But yeah, I hear you on that. Maybe being too simplistic or mean that we know what's coming because we've seen those before. But this is something very different. Um, and he starts like laying out some of the conditions that he thinks will have to come about for there to be a dramatic change in direction of the trajectory of, did you read, did you see this piece?

Alix: Mm-hmm. And it was things like you see one of the [00:30:00] main AI companies with like public. Brand and name recognition go under. So like a, an anthropic, um, or an open ai like obviously financially struggling to be able to raise money. You would see Nvidia stock price drop because their sales drop. You would see that there would be these like conditions.

Alix: And I wonder if you have any reflections on like what normal people should be like looking for when thinking about this runaway train. Maybe like going in a completely d different direction than it is headed now.

Sarah: Yeah, the things that I would be looking for are like. Real and radical departures in the behavior of these companies, and I think this Nvidia deal is one signal of something, right?

Sarah: Yeah. We've now entered a transition from a world where. There's huge amounts of capital being expended, but it's coming from somewhere. It's coming from the old business model. Mm-hmm. For the [00:31:00] tech industry financing the future.

Alix: Mm-hmm.

Sarah: We're now moving into this more speculative environment where there's debt financing, where there's round tripping or creative accounting practices emerging.

Sarah: Corners of the industry are propping up other corners of the industry. We're in this new space and at a scope and scale where like governments alone can't backstop it. That's an important dynamic too here. So. That's one thing to kind of, mm-hmm. Make note of that, that's changing the incentive structures and the behavior of these companies in ways that we don't quite yet know.

Sarah: Also, looking and paying attention to. What other parts of the economy that have not really yet been touched by AI are now being pulled into this vortex? So the fact of life insurance [00:32:00] policies being used to issue debt, like that's the signal that I'm paying attention to those kinds of wrinkles in the financial structure and where we're individually personally exposed to them.

Sarah: I think that says something. There have been a few places where we've seen companies announce significant layoffs, attributing them to ai, and let's be real, that doesn't have to actually. Necessarily be tied to, like AI has meaningfully replaced anybody's job. Companies have just decided it's in their economic interest to lay off a bunch of people, and they have AI as the justifying reason.

Sarah: So I think Accenture just announced 11,000 layoffs. Because of changes in, in AI and the, you know, value of consulting services.

Alix: I would also say it's such an easier sell to, instead of being like, we're not making enough money to continue to [00:33:00] employ these people to say we're an innovative organization that's going AI forward, and so we're laying off all these people.

Alix: Exactly. Like that's like a Exactly. It's a very clever narrative. Spin I think I sent for some of these companies

Sarah: and like a great way to mask corporate bloat. Totally. Like if a company has grown bigger than it can be justified. Yeah. Like AI gives a great reason to like trim that back down. That's another signal to be paying attention to.

Sarah: Mm-hmm. Also, it's often like the tech CEOs that are the ones that are being like AI's gonna replace your jobs. Yeah. It's almost always like a tech company that's putting forward that narrative.

Alix: Yeah. Which is. Their pitch, basically. Yeah. Like it's saying, we may not have a B2C plan, like we may not have a consumer technology that people are willing to pay for at a scale that allows us to sustain this business model.

Sarah: Yeah.

Alix: But we have good news and that's, you don't have to pay for people anymore to do the jobs in your organization. And it feels like a very, although I will say I find it really strange that we haven't [00:34:00] seen. A negative response. So like when Amai came out and said that maybe the US economy would lose 25% of its jobs in the next like 10 years or something, nobody was like, that's kind of fucked up, that as a company, your goal is to create a United States of America with 25% unemployment.

Alix: Right? No, nobody's like. Do you wanna live in that country? It's still kind of framed in this theoretical trajectory setting, narrative frame, rather than a political project that's deeply disturbing.

Sarah: Yeah, I mean like, just speak for myself here. Like, I wanna live in a world where. People have jobs that they find meaningful in their lives.

Sarah: My problem isn't that I'm too busy. I mean I am, but like AI's not gonna solve that addressing the way that capitalism is working in our economy. Like that's the thing that's gonna change things around in a meaningful way. So like, [00:35:00] why would we wanna buy in on like that particular vision of an AI future?

Sarah: I don't think that it's necessarily that. Persuasive, unless you're making money off of it, that's the key.

Alix: But it feels like a narrative project that's designed to keep them in control. Of what happens next where it's like I'm doing something that I can convince you is going to be so significantly disruptive that your only choice is to continue to listen to me.

Alix: Otherwise, you're gonna miss out on navigating that near term future that is so dramatic and could potentially like transform your whole life, your whole sector, your whole state. Someone said to me like a year or two ago, and I can't remember who it's, 'cause I would love to credit them that Sam Altman is fundamentally uncharismatic and that his communications team has found a way to have him lead via press release.

Alix: They do press releases and product releases as a way to stay in control of the narrative. And I [00:36:00] found that a really like disturbing thing to think about, that there's like these 30 guys that essentially have just played us like a fiddle and just keep us, you know, like this Nvidia announcement, the fact that like.

Alix: You could casually report a hundred billion dollar deal. That maybe doesn't even happen. 'cause maybe they can't do the one gigawatt.

Sarah: Yeah, yeah, yeah.

Alix: Thing or whatever. They have us all in this like wrapped attention, just like listening to what they're gonna say next in this way that I don't know how we like break that.

Alix: Spell, do you think? Do you think it's possible or do you think that it's just gonna have to be death by a thousand cuts? Where like everything they say, everybody starts realizing it's not actually coming to fruition. People start trusting them less. People start saying, where is all these? You've, you've set out these very clear timelines.

Alix: We've reached that time. We haven't seen what you've said is gonna happen. How do we, how do we resize their role in our public imagination about what's gonna happen?

Sarah: Yeah. Us and we are doing a lot of work there. Totally. Right? Yeah. Yeah. Like, do you remember two [00:37:00] years ago when Sam Altman was about to like go before Congress for the first time?

Sarah: Yeah. And he did this really smart thing. Which was have a dinner the night before with everybody who was ready to rake him over the coals. Yeah. Um, 'cause that was the tone at the time. It was like nobody liked tech CEOs. Everybody was ready to investigate. They did this warmup dinner the night before and he got very friendly lines of questioning the next day in his testimony, the we that has been sort of.

Sarah: Manipulated is a lot of people in positions of power that should know better. Totally. And that need to be standing up and doing their responsibility and like serving their constituents who are being negatively impacted by all of this. 'cause if you look at like where the public is Yeah. Like people are not thrilled about the direction that we're going on to say the least.

Alix: Universal diagnosis. Yeah,

Sarah: yeah, yeah. So I mean, I think one way of, [00:38:00] of slicing all of this is that this is revealing to us that we do not have a democratic system that is meaningfully representing the interests of the public. We have a space that can be like lobbied the hell out of very effectively. If you throw enough money at a problem, then you know you're going to be persuasive.

Sarah: And I think that we are so outgunned. Financially in this, but the people power part of it remains really potent. And if we can like reclaim that, then I think that we can really meaningfully change the path that we're on.

Alix: Totally. I think that's actually a really good place to end. This conversation, at least for me, has given me a framework of understanding as I see these new press releases as I see these deals, having a little bit better fluency and figuring out like what is actually happening and like not accepting that at face value or.

Alix: Trying to even absorb these numbers. 'cause a lot of these numbers are totally meaningless. Oh yeah. It's basically this attempt to kind of capture. You know, that sense that like these are the [00:39:00] serious people that are gonna tell us where things are headed. And I feel like my big takeaway is that like these are not serious people.

Alix: Yeah,

Sarah: yeah, yeah. Channel succession. Totally. Yeah. These are not serious people, just to validate like we're not losing it. There's something wrong here. Now we have to set about to the work of like making it right. But yeah, this is not past the smell test. Yeah.

Alix: Cool. Okay, well thank you. I feel like I can tell how much time you spent sort of looking at these deals and I feel like some of this language of round tripping and like thinking about this creative accounting I think is just really helpful and like kind of clearing the fog around some of this stuff.

Alix: So thank you for taking us through all of that. It was super interesting.

Sarah: Yeah, no, thank you. I'm, I'm glad to like finally get the chance to like talk all of this out with somebody and not just like. Sit in front of my computer being like, oh my God,

Alix: is someone else looking at this? Yeah. This is crazy. Yeah.

Alix: What is this?