World War AI | Ben Hunt on the Economic Consequences of the AI Boom

Summary

Excess Returns Podcast

Episode Title: World War AI | Ben Hunt on the Economic Consequences of the AI Boom
Published: November 28, 2025
Host: Matt Zeigler (Excess Returns)
Guest: Ben Hunt (Epsilon Theory)

---

Main Theme & Purpose

This episode dives into the powerful and rapidly shifting narratives around artificial intelligence (AI), with a particular focus on how U.S. government and corporate investment in AI is being framed as an existential, World War II-scale national security project. Ben Hunt breaks down the underlying economic, political, and market consequences of this framing, drawing on his recent essay, "World War AI." He explores the risks of massive capital and energy reallocations, their impact on society, jobs, and inflation, and lays out alternative policies to balance progress with prosperity for all.

---

Key Discussion Points & Insights

1. Narrative Shift: From Promise to National Security Imperative

• Observation: The rhetoric around AI has shifted sharply in the past year—from "carrot" stories about productivity and prosperity to "stick" narratives invoking a cold war-style race against China.
  • "The thesis of the note is that our government...will increasingly present AI CapEx, AI development as a national security, arms race, war footing, in competition with China as a mortal enemy. Very similar to the mobilization of the economy around World War II." (Ben Hunt, 06:30)
• Consequence: This narrative change justifies unprecedented resource allocation and sacrifices by invoking historical examples like the Manhattan Project and WWII mobilization.

---

2. The Limits and Fallout of the “Carrot” Narrative

• Carrot Narrative Recap: For years, the public was sold AI as a path to abundance—less work, more wealth, greater leisure. Hunt argues this never materialized for most people.
  • "We were presented with a carrot, not a stick. We were presented with a carrot of a golden age of productivity and leisure and wealth will be ours if only we make this sort of enormous effort and investment in AI. The carrot...was never going to be successful in the long term because it just ain't true. I mean, we're all working harder today, not less." (Ben Hunt, 08:33)
• Inequality: The supposed prosperity accrues to 'techno-oligarchs' rather than society at large.

---

3. World War II-Scale Capital & Energy Reallocation

• Data Center Buildout as Mobilization: Hunt documents that projected U.S. AI infrastructure investment (>$4 trillion over four years) rivals all WWII spending (in current dollars).
  • "The punchline is...the numbers are roughly the numbers that we spent on World War II. I mean, in truly insane numbers in terms of money—something north of $4 trillion over the next four years...That money's got to come from somewhere...all that money today going towards AI data center build out, that's investment that does not go to anything else. In particular, it doesn't go to the consumer economy." (Ben Hunt, 13:29)
• Economic Implications:
  • Money and physical resources (especially electricity) are being diverted from the wider economy, crowding out consumer services and driving up costs (e.g., health insurance, housing, energy).
  • "We're already seeing that happening where consumer credit, capital intensive consumer services like health insurance, homeowners insurance—all these things are going up in price because there's not as much capital being invested in them." (Ben Hunt, 13:29)
• Debt to GDP: The US is already at WWII-level debt-to-GDP before new AI spending begins, compounding risk.
  • "We're starting there, we're starting there with the U.S. government today. Now that's for U.S. government spending. That's not counting the private spending and investment." (Ben Hunt, 13:29)

---

4. Energy: Scarcity and Rationing Risks

• AI as Energy Black Hole: AI’s electrical demand could consume 22–25% of U.S. energy by 2030, creating price and availability pressures for everyone else.
  • "If you are reallocating...electricity and energy to these AI data centers—which we are—it creates shortages of energy everywhere else...that leads to today being rationed through price—so our utility bills go up. Tomorrow, it goes towards rationing by just—you don't have it available." (Ben Hunt, 22:55)
  • Potential path: price controls, political backlash, and even physical ration books (as in WWII).

---

5. Employment and the “Jobless Boom”

• No New Jobs: Unlike WWII, when mobilization created millions of jobs, the AI boom is explicitly designed to reduce labor.
  • "In fact, the avowed purpose of AI data centers is to reduce jobs...You're getting operational efficiencies...There are no new jobs here. There are no net new jobs for sure." (Ben Hunt, 28:25)
• Social Impact: AI’s disruption will most heavily affect junior/entry-level workers first, with plans to expand into blue-collar roles via robotics in the future.
  • "This is the old saying about how many people do you have to employ for a data center? I mean, it's one guy and a dog to...make sure nobody's breaking in. That's it. These things don't require—create—new jobs...the goal is to replace the physical manipulation, the blue-collar jobs tomorrow." (Ben Hunt, 28:25)

---

6. Crowding-Out and Dynamics of Boom-Bust

• Market Myopia: Individually, every AI data center project 'pencils out.' In aggregate, the resource drain is unsustainable, setting up for a classic boom–bust scenario.
  • "All 500 of these deals don't work in the aggregate...This is market myopia...and it is what generates boom situations...then you have a terrible, terrible bust." (Ben Hunt, 53:32)
• Political Risk: The only solution is to rein in AI's share via policy intervention before the bust.

---

7. Policy Alternatives & Hopeful Approaches

Ben proposes three policies to realign progress with broad prosperity:

1. Rapid reshoring of manufacturing: Create new jobs by bringing industry back to the U.S.
   • "I am all for requiring Apple...to bring back manufacturing jobs to the United States." (Ben Hunt, 44:16)
2. Build out all forms of energy generation: Support economic abundance by massively increasing energy supply via every means—nuclear, solar, wind, gas, coal.
   • "We need nuclear. Of course we need nuclear...Yes, we need wind and solar...we need it all. Am I worried about carbon emissions? Of course...but I am more worried about this Blade Runner future that we're creating..." (Ben Hunt, 44:16)
3. Hard 10% electricity cap for data centers: Prevent AI from consuming all new electrical generation—ensure other sectors and the public benefit.
   • "Put a cap on how much electricity can be consumed by data centers. My proposal is a 10% cap." (Ben Hunt, 44:16)

---

Notable Quotes & Memorable Moments

• "Don't piss down my back and tell me it's raining. They're pissing on our back and telling us, oh, it's just a gentle spring rain. The carrot doesn't work." — Ben Hunt (10:05)
• "When you reallocate trillions of dollars, whether it's for a COVID response or an AI data center build out, without having enormous repercussions in the rest of the economy...we focus so much on the problem we're trying to solve that we ignore the unintended consequences." — Ben Hunt (20:00)
• "This is a different kind of technology. It is designed to be a replacement for human labor. That's the whole point. The only jobs that remain...are the meat taking care of the meat...health care, elder care, nursing." — Ben Hunt (34:17)
• "If people vote and say, 'this ain’t working for me, I am having shortages, my utility bill goes up...so I can have an AI friend? Are you freaking kidding me?' ... that's when we'll have a moment...either AI authoritarian moment or an AI backlash...Both of those outcomes to me are really poor." — Ben Hunt (41:04)
• "People don't know it's an option. And it has to be an option and we have to start talking about it." — Ben Hunt (64:32)
• "It's important to be for things...be for energy expansion, be for jobs coming from industrial and manufacturing reshoring...bring positive energy to this." — Ben Hunt (59:57)

---

Important Timestamps

| Timestamp | Segment/Event | |-----------|--------------| | 06:30 | Ben outlines the “World War AI” thesis—AI as national security, WW2-scale mobilization. | | 08:33 | The false promises of the AI “carrot” narrative. | | 13:29 | Comparison of WWII and current AI-capital investment/crowding out. | | 22:55 | Energy reallocations, price pressures, and risk of rationing. | | 28:25 | Labor market implications—AI's true job impact vs. WWII. | | 34:17 | Why the expected “new jobs” effect from tech may not materialize with AI. | | 44:16 | Ben’s three core policy proposals for a balanced future. | | 53:32 | Market myopia and the oncoming boom-bust dynamic in AI data center buildout. | | 64:32 | The idea that capping data center energy usage is an option, and states are awakening to it. |

---

Tone & Language

• Candid, urgent, and policy-driven; Ben Hunt balances skeptical critique of elite techno-optimism with a practical, hopeful call for collective action.
• Deploys vivid metaphors ("carrot vs. stick," "pissing down my back," "boom-bust," "meat taking care of the meat") to elucidate potentially abstract concepts.

---

Actionable Summary for Listeners

• The AI boom is being rebranded as a “war effort” to justify massive, concentrated resource allocation, with real risks of economic crowding out, inflation, inequality, energy shortages, and job loss.
• Hope exists: meaningful policies can ensure the benefits of technology are broadly shared—reshore jobs, build more energy, and cap AI’s energy share.
• Before it’s too late and political extremism takes over, Ben advocates for proactive, positive, and inclusive policy engagement.

---

To Dive Deeper:

• Read Ben Hunt’s original essay, “World War AI”
• Explore the narrative-tracking platforms: Epsilon Theory (for all investors), Panoptica (general audience), and Perseant Pro (institutional investors)
• Stay engaged at Epsilon Theory’s website for updates and resources

---

Episode recommended for investors, policymakers, and anyone interested in the intersection of technology, economics, and society as AI reshapes the world.

Transcript

[State Farm Advertiser]

This episode is brought to you by State Farm. Listening to this podcast Smart move Being financially savvy Smart move Another smart move. Having State Farm help you create a competitive price when you choose to bundle home and auto bundling. Just another way to save with a personal price plan like a good neighbor, State Farm is there. Prices are based on rating plans that vary by state. Coverage options are selected by the customer. Availability, amount of discounts and savings and eligibility vary by state.

[Shopify Advertiser]

Running a business comes with a lot of what ifs, but luckily there's a simple answer to Shopify. It's the commerce platform behind millions of businesses including Thrive Cosmetics and Momofuku. And it'll help you with everything you need. From website design and marketing to boosting sales and expanding operations. Shopify can get the job done and make your dream a reality. Turn those what ifs into sign up for your $1 per month trial at shopify.com specialoffer the story being presented to.

[Ben Hunt]

Us a year ago and six months ago was that all of this money, all of this electricity, all of this effort going towards AI it wasn't because there's some threat out there. It's because of all the good stuff that would come to us from making this effort. It was never going to be successful in the long term because it just ain't true. It just ain't true. I mean, we're all working harder today. The US government will increasingly present AI CapEx, AI development as a national security arms race, war footing in competition with China as a mortal enemy. All that money today going towards AI data center build out. That's investment. That does not go to anything else in particular, it doesn't go to the consumer economy.

[Matt]

You're watching Excess Returns. Ben Hunt is back with me. We're talking about his latest Epsilon theory essay that's moving very, very, very fast and into reality. World War A.I. welcome back, Ben.

[Ben Hunt]

It's great to be back, Matt. Thanks for having me.

[Matt]

So it's crazy. Is that when you wrote about this a week ago and we're recording it is November 25, about 11:30 in the morning, the White House just, just released not long ago this Genesis mission statement. AI accelerated scientific discovery at a Manhattan Project scale a week after you wrote this note. And even if people are watching this and going, I'm sick of talking about the NASDAQ, I'm sick of talking about Mag7, I don't want to hear about AI anymore or these like handful of names. I think you really have to understand what's going on Here, because this is a giant policy, political markets, coordinated effort. You're writing about it in this piece, Persian Pro. You've got a bunch of trackers. Could we just for a second, what's on Perseant Pro? Why should people look there if they want to see under the surface of the story?

[Ben Hunt]

Absolutely. So what we think we have, in addition to whatever my opinions are about things, we think we've got a technology to actually measure the storytelling, the narratives, we call them, the semantics of the world markets, for sure, but also politics, also society. AI is a fascinating topic and it's so important because it traverses all three of those. It absolutely traverses markets and the economy, it traverses politics, our society, our roles as humans, our jobs, it covers all of that. And I think we've got a really good technology to measure the way the reality is being presented to us. We don't have any secret sauce for understanding what reality is. You've got your ideas, I've got my ideas. But I think we've got something very special to accurately and truthfully describe how reality is being presented to us. And honestly, I think for markets and politics and the way we lead our lives, that presentation is at least as important. I kind of feel like it's maybe more important than the reality itself. So we do that in a couple of different ways. We publish Epsilon Theory, which puts out articles like I wrote the World War AI. We have a totally free version we call Panoptica, where we write about narratives in sports and personal finance and media. And then we have Perseant Pro, which is for, as the name implies, professional investors. So we're trying to hit all the audiences here. We're not communists, we're capitalists. But we wanted to run the gamut from information for a professional investor, Perseant Pro, information for a casual investor, Epsilon Theory, and then information, if you're not even really thinking about investing at all. And that's Panoptica. So that's our goal and that's what we're doing.

[Matt]

So if you're watching this and what we're going to dive into here, I can't suggest enough. Look up Persian Pro. P, E R S, C I, E N T. Check it out. Or go to the Epsilon Theory website to get your link, or go to Panoptica AI. It'll get you a link. Because what's wild is we have these. These pulses and these storyboards that are tracking these categories and these stories bubbling up. So if you want to know how Ben is basically on top of this, Story seemingly in advance. It's because this is the tool that's helping him get there. Ben, you ready to dive into this piece? I think we got to dissect this together a little today.

[Ben Hunt]

I am, and let me start with this. And it connects with what you just said. Matt. The reason I wrote this piece, World War AI. The thesis of the note is that our government, the US government, will increasingly present AI CapEx AI development as a national security, arms race, war footing, in competition with China as a mortal enemy. Very similar to the mobilization of the economy around World War II. Crazy as it sounds, at essentially the same scale in terms of capital, in terms of energy allocation, in terms of consumer sacrifice as we experienced during World War II. The reason I came up with this was that we were seeing in our technology clear signs of an uptick in that sort of narrative, that sort of presentation of the AI build out as being necessary for national security, national survival, moving towards a war footing vis a vis China on the race for AI. And it's. It's crazy, let me just say, out of the way, yes, China is a competitor, an adversary of the United States. This is not an AI arms race. This is not an AI war. But that is how it is being presented. And we ain't seen nothing yet. That's the point of the note.

[Matt]

Explain the carrot metaphor. Explain this idea of why the carrot wasn't going to work. Because inside of this, this is a key distinction that you lead off with.

[Ben Hunt]

Yeah, well, this World War AI, that AI development is a national security issue. This is kind of new. This wasn't the story being presented to us about AI a year ago or even six months ago. The story being presented to us a year ago and six months ago was that all of this money, all of this electricity, all of this effort going towards AI, it wasn't because there's some threat out there. It's because of all the good stuff that would come to us from making this effort. We were presented with a carrot, not a stick. We were presented with a carrot of a golden age of productivity and leisure and wealth will be ours if only we make this sort of enormous effort and investment in AI the carrot. My personal view was the carrot was never going to be successful in the long term because it just ain't true. It just ain't true. I mean, we're all working harder today, not less. We're all working harder.

[Matt]

There's more gas on the fires, but we're working harder.

[Ben Hunt]

Yeah, exactly. Exactly. I'm not kicking back, enjoying the early Stages of AI productivity. I'm working harder than I've ever worked, and everyone I know is working harder than they've ever worked. There's no. The story or so the story was, is that we'd have an employment boom from AI. Really? Are you seeing an employment boom? I'm seeing the opposite of an employment boom. We were told, oh, well, we'll have AI friends. Really? You think that's a good thing? You think that we're all doing better in our social lives because we have AI friends? We think news and entertainment is better because of AI content. There's this old saying, don't piss down my back and tell me it's raining. They're pissing on our back and telling us, oh, it's just a gentle spring rain. The carrot doesn't work. I think so many people are understanding that it's not some golden age of prosperity for us people. It may be a golden age of prosperity for the AI owners, as I like to call them the techno oligarchs. It may be great for the Peter Thiels and the Sam Altmans of the world, but it ain't great for us. I mean, I've looked at my utility bill. This is not good. This is not good at all. And I think we all see that the trajectory of this is start to get worse. The carrot doesn't work, so they've got to move to the stick. The stick is a threat. And the threat is, oh, my God, we have to keep on this course because if we don't, China will win whatever the hell winning means, and then they will conquer us or whatever that means. We have to move towards this war footing. We saw early signs of this in our research. I wrote the note. And then this week the White House announces, yep, it's a war footing. A Manhattan Project to develop AI before the bad guys do. Couldn't script it any other way. This is happening.

[Matt]

When you look at this, and I think it's what you hinted at with the utility bill right there, it's shortages of energy, there are shortages of capital, which is part of where I want to take this next and the way that you fund this, because this is what has far, far, far, far reaching market implications of engaging in this type of a policy. So you said, for example, the treasury is already at World War II, debt to GDP levels before this starts. And now with Genesis, I mean, we have basically a biblical name and the Manhattan Project invoked in a single thing from the White House. How do we even start to begin? How much Runway this even gives us.

[Ben Hunt]

Well, I think the way to start is to look at the projections bottom up projections, announced projects, what is it going to cost? And what's it going to cost both in money. What's it going to cost in terms of power to run these AI data centers? That's what I try to lay out in the Note. And I can give you the numbers really quick, but the punchline is going to be that the numbers are roughly the numbers that we spent on World War II. I mean in truly insane numbers in terms of money, something north of $4 trillion over the next four years. So even the same time frame as World War II, which is basically four years for the United States, we spent a little less than $300 billion on World War II. That translates in today's money right at $4 trillion. It's the same amount of money. Same amount of money. And that money's got to come from somewhere. It comes from companies, it comes from investment, from loans, it comes from the government, from our tax dollars and from the government's borrowing. That's what we did in World War II and we borrowed. The government borrowed enormous sums to pay for it. So at the end of World War II, our debt to GDP, which is how you often look at these things. What's your debt amount relative to your productive capacity? There's this kind of famous and meaningful line of if you borrow as much as your annual GDP is, that's problematic. Well, we got there in World War II. We're starting there, we're starting there with the U.S. government today. Now that's for U.S. government spending. That's not counting the private spending and investment. So the 4 trillion, some of it's got to come from the government, most of it will come from private sources. All of it is money that is not being invested in other things. And that's the point of the Note, which is that just like In World War II, with all the capital from the government, from corporations, from individuals, war bonds and loans, all that stuff, all that money going towards World War II production, all that money today going towards AI data center build out, that's investment that does not go to anything else. In particular, it doesn't go to the consumer economy. We're already seeing that happening where consumer credit, capital intensive consumer services like health insurance, homeowners insurance, all these things are going up in price because there's not as much capital being invested in them. This is all happening now. I'm not projecting something to the future. I'm saying this all happened before during World War II. It's all happening now, early stages and it's all going to get a lot worse. The other crowding out effect. So one aspect of this is all the money that gets invested into AI data centers. Trillions of dollars is money that doesn't get invested in other places. The other impact of this though is on the borrowing by the government to support this. That raises government can always borrow its money because it's the safest investment. But the more they borrow, especially once you get over that hundred percent of GDP line, the more you got to pay in interest. So this has two effects. One, there's less money being invested in, I'll call it the consumer economy and second, the price of money interest rates, those go up. That's the two impacts from this massive World War II level amount of capital reallocation into the AI data center world.

[Shopify Advertiser]

So you're about to make a trade based on a friend's text, but which u do you listen to is it we could buy a house in Tulum, get optioning those options. We could lose everything. Or let's do a little research, get your head in the trade and make the investment decision that's right for you. Learn [e-mail] TradeSmart.

[Amazon Prime Advertiser]

This episode is brought to you by Amazon Prime Black Friday game day on prime is an epic day of live sports. It all starts at 9am Eastern with the Capital One skins game. Then Black Friday football returns when the Bears take on the Eagles at 3pm and it culminates with the final night of Emirates NBA Cub group play with Bucks Knicks at 7pm and Mavs Lakers at 10pm Black Friday game day only on Prime.

[Matt]

What do you make of and I remember I think a Bank of America chart floating around from the COVID response where they pointed out that the COVID response was basically going to be a World War II level response in fiscal stimulus to offset the impacts the perceived impacts of the COVID pandemic for the US and it feels that's part of why we're starting here. Is that an incorrect statement or how do you interpret the statement of why we're starting where we are today?

[Ben Hunt]

No, it's a, it's a very accurate statement. It's accurate in two respects. One, we borrowed a lot of money and we plowed it into the consumer economy. Mostly it gives us a new higher starting point for debt and spending that we take on from now. So that's that crowding out effect where the price of money interest rates goes up because we're starting From a much. We're starting from a point where World War II ended, basically. But here's the other thing. That infusion of money, about $4 trillion World War II levels into the real economy in 2020 and 2021. It blew up our world with inflation. It blew it up. You can't reallocate trillions of dollars, whether it's for a Covid response or whether it's for an AI data center build out without having enormous repercussions in the rest of the economy. It's always these, I'll call them unintended consequences, because I think they are largely unintended. But what it really is is that we focus so much on the problem we're trying to solve. Oh, my God. We got to build out AI data centers. Let's put $4 trillion into it. What does that mean for the rest of the economy or with COVID Oh, my God. People are shutting down their businesses. We've got to put $4 trillion into the support the economy. What does that mean for the price of goods and services and supply chains and everything like that? It's. It's always the rest of the economy when we get. Because we are, we are so focused on AI right now. And I get it, I'm a big AI fan. But we shouldn't be focused on it and investing in it at World War II levels that make it impossible for us to invest or support the rest of the economy. But that is what's happening.

[Matt]

So one of my crazy recollections of the COVID pandemic is oil going to zero or negative.

[Ben Hunt]

Yeah.

[Matt]

And going to the gas station and some guy chuckling from behind the gas, like across the thing from me, you know that, oh, like they're going to pay us for this. Right. And yeah, spoiler, we still paid for our gas. You know, that's, that's not the way this thing works. But these unintended consequences happen when we look at energy now and we look at the resurgence from those negative oil prices to where we are today. And think about especially electricity in the US going forward. You wrote in the piece, I think it was 22 to 25% of electricity usage for AI by 2030.

[Ben Hunt]

Like, yes. Yeah. See, that's the other part of the massive reallocations here. Yeah. So again, the same thing happened in World War II, because energy is what? Well, it's used in everything. Right. Energy is the most fundamental physical resource. So you can talk about energy in the form of gasoline to fuel your car, because energy is what drives all of transportation in an economy is what drives all of manufacturing in an economy. It's really what drives all of agriculture. I mean, agriculture is a function of the energy you put into drives all of mining and materials. Again, it's just a function of the energy you want to put into it. So when we're talking about energy, yes, we're talking about electricity, but we're also talking about everything else because energy goes into everything. And if you are reallocating, not just money, not just capital, but if you're reallocating electricity and energy to these AI data centers which we are, it creates shortages of energy. Again, everywhere else we're looking at shortages of capital and we're looking at shortages of energy across all the non AI data center aspects of our economy. What that leads to, and the exact same thing happened in World War II is what's called cost push inflation. We experienced it in this country in the COVID response. Remember how all the supply chains got wiped out and we had all these shortages that led to the cost of things going up? Well, this is that. But to use kind of modern parlance on steroids, right, this is that. But spread across everything because energy and capital go into everything. And if you've got shortages of energy and capital, then the price of everything goes higher, just like in World War II. In World War II, the main focus was on food. The price of food just skyrocketed in World War II. So the government's response to that, First you try to ration it through price. That's what a market economy does. When things get scarcer, the price of that thing goes up. So it takes more money to buy that thing. Ultimately though, that becomes politically impossible with things like food, with things like electricity, with things like gasoline. So the government moves to price controls just like they did in World War II. Set up a system of price controls on everything that the consumer uses and needs to survive. In World War II at least, and this is a pretty common occurrence, sometimes even price controls aren't enough. There's too much cheating, there's too much getting around it. So ultimately, like In World War II, the government has to move to actual physical rationing. So you get your ration book of food, your coupons, so that it doesn't matter what you're willing to pay and it doesn't matter what the price controls are. You only get a pound of butter a week. That's your physical ration. That's where this is going. And I know that sounds alarmist, I know that sounds Stupid. It doesn't sound stupid, it sounds alarmist. But this is what happens when you have capital and energy reallocated in such a massive way. It leads to today being rationed through price. So our utility bills go up tomorrow. It goes towards rationing by just. You don't have it available. So you're starting to see that in insurance today. You can't get homeowners insurance, you can't get health insurance. Pretty soon we'll have rolling brownouts. We're already experiencing some of that in some places where electricity is rationed out. Now that I think can make sense if you are in fact fighting a war with Nazi Germany. I don't think it makes sense at all for this fake AI war with China. But that's where we're going. That's how reality is being presented to us.

[Matt]

Let's turn the attention for a second to, to corporations and more specifically jobs, the people who work for these corporations. Part of AI, part of this pitch was, you know, margin expansion. It's going to increase productivity or it's going to reduce headcount. You don't need as many people. That helps with margin. That helps with margins if you're letting people go or if you're increasing productivity. But if these patterns hold, I mean, where's, where's this going on the employment front, knowing that these pressures are also being added to the economy.

[Ben Hunt]

Yeah, well, the saving grace for World War II, from an economic perspective, the thing that saved the United States economy from horrible stagflation. So the capital shortages meant that there was no money available to, for, I'll call it Main street businesses. For small and medium businesses. There was no credit available to expand your bakery or anything like that. So there was no growth in the US Economy from small and medium businesses. None at all. At the same time, you had this cost push inflation that I'm describing, put those together. Stagflation, no growth, persistent, pretty terrible inflation. The saving grace was the United States government created millions of jobs. The money was spent on building factories and transforming existing factories into war production. Rosie the Riveter so women entered that workforce. Millions of jobs created real economic growth. Growth was actually fine in the United States because we created a ton of jobs to fight the war. That's what's really different this time. It's the same as World War II in terms of capital. Massive amounts of capital going towards the AI data centers, massive amounts of energy going to the AI data centers. But there are no new jobs here. In fact, the avowed purpose of AI Data centers is to reduce jobs. When people talk about productivity, they're talking about labor productivity. They mean doing more, producing more with the same number of people, or producing the same amount with fewer people. And let me suggest that every person listening to this show, their company is having that experience. All these AI projects that I know your company is running, they're not opening up some new revenue opportunity. They're not creating some new economic vista. Oh my God, we've got this opportunity. We can hire people and do all this stuff and make lots of money and sell this product. That's not it. You're getting operational efficiencies. You're getting especially junior people being essentially replaced. Most companies I know have slowed down their hiring plans, especially at the junior level. Maybe they've been able to let some people go. There are no new jobs here. There are no net new jobs for sure. This is the old saying about how many people do you have to employ for a data center? I mean, it's one guy and a dog to kind of patrol around and make sure nobody's breaking in. That's it. These things don't require create new jobs. On the contrary, they're designed for labor productivity, which means elimination of jobs. And it's just getting started. The whole notion of robots, as Elon Musk says, the embodiment of AI, is not just to replace the white collar symbolic manipulation jobs, the thinking jobs that are being replaced today. The goal is to replace the physical manipulation, the blue collar jobs tomorrow. They're not confessing these techno oligarchs, they're bragging. It's the whole point. So this is what we're fighting the war for. It's not to compete or win versus China. That is bullshit. What this war is for is to replace human labor. That's it. That is what we humans are being asked to make these sort of sacrifices for. And that's what I'm committed to stopping. Not in the sense of getting rid of AI. I'm a big fan, but it needs to be subordinated to humans, not elevated to something more important than humans by the techno oligarchs. That's what I want us to try to understand.

[Matt]

Do you see any willingness or desire or expression of acknowledgment for the job situation yet? Is that showing up in any of the narratives? I mean, we have Peter Atwater writing about the K shaped economy for a reason that feels like it's worse. I think about Kyla Scanlon in the Vibe session. I think about all these topics that are Trying to describe this feeling of what's going on. Is it popping up anywhere where we're talking about this in a here's a plan to grow out of or beyond.

[Ben Hunt]

This or is barely, just, just very barely. You see kind of little inklings of some people talking about it. But it, it absolutely has not entered the popular narrative structure. And I'll tell you why. And you'll hear this all the time. I hear it all the time when you talk to people about AI and where are the jobs? The comeback is always, always some variation of the following statement. Well, new technology always creates new jobs. We just don't know what those new jobs are going to be yet. That is what you will hear. And my response to that is this is a different kind of technology. This is a different kind of technology. It is intentionally. It is designed to be a replacement for human labor. That's the whole point. The only jobs that remain in an AI uber alles future. The only jobs that remain. And even that I'm not so sure if Elon Musk has his way. The only jobs that really remain are the meat taking care of the meat. It's health care, it's elder care, it's nursing, which by the way are the only human jobs that are expanding in this current regime. It's the meat taking care of the meat. And that doesn't work for me as a human being. It really doesn't. So I think there's a path out of this, Matt, I really do. But the path is growing narrower and narrower by the day, especially as the government moves forward with characterizing, presenting the AI build out and the power generation build out that will be sucked up entirely by the thinking machines. Presenting this as a national security war effort. This is the worst possible path we could be going down.

[Matt]

I want to call out some of the Persian signatures. You've got a chart. And we'll see if we can share some of this here too. Tracking confidence that hyperscale builds will continue growing at a z score of 5.07 while skepticism about these projects simultaneously hitting an all time high z score of 5.91. But then that.

[Ben Hunt]

Let's describe what a Z score is first.

[Matt]

Say what it is. Say how you interpret this. It's important.

[Ben Hunt]

A Z score is a way of saying, well, how unusual of this. It's basically if you're familiar with the concept of standard deviations, how unusual is it that you see this? How unusual is it that you would flip a fair coin? It would come up heads 12 times in a Row. How unusual is that? Well, that's what the concept of standard deviation is. That's the concept of Z scores. So when we say oh, it's got a z score of 5.9 or whatever the number is, what we're saying essentially is that's 5.9 standard deviations away from what's normal, away from the historical level of narrative volume on this topic, which if you know anything about standard deviation, I mean two standard deviations away is a ton, three standard deviations is like oh, you're in the top percentile. Six standard deviations. Well, this is extremely rare. And what it's showing is that this is in high rotation. Both sides of the coin. People saying yes, this Capex is going to continue and people saying this can't possibly continue. This is what everybody's talking about. This is where the attention is on that topic. But to your point earlier, the arguments against Capex and the AI buildout are not that it's anti human, it's not that it's going to suck up every bit of energy, it's not that it's going to suck up every bit of capital that's going to starve the rest of the economy. That's not the why that people are arguing against capex. And that's what I want to try to get that message out there. I want people to understand that the reason why CAPEX capital expenditures on AI should not advance the way that they're being planned, that we should not be putting this on a war footing. The why? Why not? Because it starves the rest of the economy of money, energy, jobs. And so I am against that. It's not that I'm against AI, it's I'm against of this over investment that starves the rest of our economy from what it needs to survive, much less thrive.

[Matt]

How do you interpret this environment of the hyperscaling growth will continue relative to the skepticism that these will generate a positive return in some way? Because both of those flashing is a really interesting contradiction to me.

[Ben Hunt]

I think it depends on your time frame. I think that because that's where I am as well, I both think this is, nothing stops this train. I mean when you've got the White House, when you've got Wall street, when you've got every talking head, when you've got corporations with all the money in the world pushing this vision and pushing it away of our national security and our national survival is at stake. This isn't stopping. This isn't stopping.

[Matt]

And.

[Ben Hunt]

I believe there's there comes A point. And that point is not far off. When people vote and they say, this ain't working for me, I am having shortages. My utility bill goes up, we're having rolling brownouts so I can have an AI friend. Are you freaking kidding me? This. This isn't it. And so it gets cashed out in politics. And that's when we'll have a moment. We'll either have, I like to call it the AI authoritarian moments, where if the techno oligarchs and their money and their political support says, well, if the problem here is democracy, if the problem here is people voting, I guess that's what we have to do away with. Or you could have an AI backlash and a real political populace come to power. Both of those outcomes to me are really poor. I neither want Proconsul Vance acting on behalf of Peter Thiel and the techno oligarchs. I don't want that. And I also don't want Comrade Mamdani as the. As our leader either. Through some populist backlash against AI. I want to try to avoid those two extremes. I think it's harder and harder to avoid those two extremes. But I think if we can get the word out to get people to recognize it's not just that this should not this isn't a war with China over AI, it's that these data centers are sucking up every bit of capital and energy that we need for our own economy. I think if people start thinking that, they'll see it, they'll see it, they'll see it in their own lives. And I think it's not too late, but it's getting close to being too late from those two outcomes of either AI authoritarianism or AI backlash through political populism. I want to avoid both of those. I'm probably crazy to think we can avoid them, but we need to get people to recognize the problem before we can take any steps to try to avoid those two bad political outcomes.

[Matt]

Okay, so you laid out three policies here, and I want to go through these together to policy one, rapid domestic reshoring of manufacturing. Policy two, rapid build out of all power generation. And policy three, a hard 10% electricity cap for data centers. You said policy three. The cap that, that's the big one that will be fiercely be fought because it actually bites and actually contains them. It doesn't bite today with most states and where they are. What's the political economy change? How do we get there?

[Ben Hunt]

Let's start with the first two. Let's start with the first two because I think it's so important to be for something. I can sit here and say, oh, it's the Sam Altmans and the Larry Ellisons and the Peter Thiels and the Elon Musk's of the world that these techno oligarchs, I mean, grr. And just be against them. And that's a mistake because I'm not anti AI, I'm not anti technology, I'm not anti progress, I'm certainly not anti national security. I am for all those things. So I want to. But I want to tell people what I'm really for, which is for the health of the United States of America and its people. So policy number one, I really do think we need more jobs. And I think the way to get jobs is to bring back manufacturing. I am all for requiring Apple, I'll pick on Apple for a second. To bring back manufacturing jobs to the United States. I think that is long overdue. I think the Trump policies to cajole, encourage, force companies to bring manufacturing back to the United States, long overdue. Amen. 100% for it. 100%. And I don't want to hear the arguments about neoliberal this and oh, the iPhone cost will go up. You know what? Apple will eat most of that price increase from the higher labor costs in the United States just the way they ate tariffs. So I'm not scared off by companies, oh, we can't do this back in the United States. Yes, you can. So I am all for bringing manufacturing back to the United States because bringing those manufacturing, bringing those industries back to the United States, that's good growth. That creates small medium businesses. Supporting those organizations creates entrepreneurialism around those efforts. They're not data centers off in the middle of nowhere with their own little energy supply unconnected to the rest of us. Just there for the thinking machines. Those are an engine of economic growth that I am entirely behind. That's what I am for. I'm for that, number one, manufacturing back into the United States, number two thing that I'm for. We got to have more energy. There's this and I love it. The abundance philosophy you see a lot in Democratic political circles, Ezra Klein, some other folks, and I think it's great. And again, it's based on you gotta be for something. You can't just be anti Trump. You can't just be, oh, grievance politics. What are you for, man? I'm for growth. I'm for abundance. I'm for having more. More liberty, more money, more. I'm for that. Well, the answer for Abundance is you got to have more energy. Because like I say, energy goes into everything. You want more housing, you want more health care, you want more of a social safety net, you want more jobs, you want more incomes. My friends, the key to all of that is more energy. Do more energy and all of those things will happen, all of them. So, yes, we need nuclear. Of course we need nuclear. Yes, we need wind and solar. Yes, yes, yes, yes, we need natural gas. Yes, we need coal. We need it all. Am I worried about carbon emissions? Of course I'm worried about carbon emissions. I am more worried about this Blade Runner future that we're creating with a permanent underclass and a permanent masterclass of the techno oligarchs and their thinking machines. I'm a lot more worried about that. So the key to all of this, and I am for a war footing on this, is we need more energy, period, from all sources. But here's the third part. Let's say we do these things. The new manufacturing that we're reshoring, we're bringing back to the United States, it's going to take a lot of energy. Our consumer economy takes a lot of energy. Right now, even if we build whatever energy we're building, every watt of that new energy is being sucked out of the system by the data centers. So it's not enough to build more energy if the data centers are going to take all of that growth and more, which is what's happening today. And this is the part that I'm having the hardest time communicating, which is that it's not just that we have to build more energy generation. We also have to keep the data centers from taking every bit of that. We have to keep them from getting a bigger slice of the pie. Because every projection we've got, they're going from a rounding error of energy consumption a couple of years ago to like a quarter of our energy being consumed by the data centers. That's what kills us. That's what kills us. So it's that third policy that I'm for is just now, before it's a problem, put a cap on state by state, grid by grid, region by region. Put a cap on how much electricity can be consumed by data centers. My proposal is a 10% cap. No more than 10% of the electricity that's generated can be consumed by data centers. Now, I'll give you a little variance on that. If you build out a new big data center, a new big electrical generation plant, it's not all yours. Some of that's got to go to the grid because this is a national effort. But sure, you build a new energy plant, you can have half of it, I'll give you that. But we got to work together to not only build out the energy generation capacity of this country, but to keep the data centers from eating that all up. That's what policy number three is for extra value. Meals are back.

[Matt]

That means 10 tender juicy McNuggets and.

[Ben Hunt]

Medium fries and a drink are just.

[Matt]

$8 only at McDonald's for a limited time only. Prices and participation may vary. Prices may be higher in Hawaii, Alaska and California. And for delivery.

[Shopify Advertiser]

Ford BlueCruise hands free. Highway driving takes the work out of being behind the wheel, allowing you to relax and reconnect while also staying in control. Enjoy the drive in BlueCruise enabled vehicles like the F150 Explorer and Mustang Mach E. Available feature on equipped vehicles terms apply. Does not replace safe driving. See Ford.com BlueCruise for more details.

[Matt]

Let's talk a little bit then about the investment implications of this because each of those scenarios lay out different things that will work, will not work, will compete, will not compete. And I, I, I, I have to believe there's, there's some opportunities inside of this. From what you're seeing, what are you excited about? What areas do you think there's actually growth that's coming that can be harnessed? What signs do we look at for that we're getting through that there's actually a future on the other side and who's going to win? Not specific names. I'm just interested in how you're thinking about this in the economic sense.

[Ben Hunt]

Yeah, and this is a tough one. And I'll describe why it's a tough one. See, the path we're on right now is the classic boom, bust cycle of a market myopia. And what I mean by that is that these data centers, each individual project, it pencils out. What I mean by that when we say something pencils out is it makes economic sense on an individual. You look at the project that you're being presented, you're the funder or you're being asked to give money to this. And you, you ask the tough questions, you go through, you dig through it, you say, okay, you know what, this deal makes sense.

[Matt]

Positive value.

[Ben Hunt]

Yeah. I can look at it and say, okay, I see the risks, but here's my return, here are my protections. All right, you know what? This pencil's out. I'm going to go, yes, this deal makes sense. What you don't see, though, is the 500 other deals that also make sense on an individual basis. Because when you put all 500 deals together, it does not make sense. All 500 of these deals don't work in the aggregate. They all work individually. And so you say, yeah, I think I'll go forward with that investment. But they don't work in the aggregate. This is market myopia. And it happens all the time. And it is what generates boom situations, because every One of those 500 projects goes through and then they get to a point where the projects stop working because they don't work in the aggregate. And then you have a terrible, terrible bust. This is what a boom and bust cycle is, and it's driven by market myopia. What I'm proposing is that we step back, we see that in the aggregate this doesn't work. We can't have an economy where a quarter of our electricity is gobbled up by AI data centers and where 4 to 5 trillion dollars is spent on that to the exclusion of everything else. The bust there is horrific. By putting a cap on the energy consumption of data centers, a lot of these deals, a lot of those 500 deals on an individual basis no longer pencil out. They no longer make economic sense. The money that has already been invested in them, that money is lost. There are losses. But you are deflating this bubble before it gets to the point where it totally goes boom and takes everything down with it. So that's what I'm advocating. I'm advocating taking air out of the expansion, the bubble expansion. Here you do that by putting political risk. Here you say there's a regulatory cap on the electricity that can be consumed. And so the projects that are still in the planning stages, they no longer pencil out, which means the capital doesn't go there. And so the capital's got to go somewhere else. Where does it go? Well, I'd love for it to go to reshoring, manufacturing and building out energy generation for everything, not just the AI data centers. So like I say, it's a long putt. It requires losses to be taken and the bubble to be deflated slowly on the AI CapEx build out the opportunities. Where does that capital go to get a return? Goes to manufacturing, reshoring and energy generation, Just not only for data centers. That's where I see where I hope this all can go.

[Matt]

So for investors, a key part of this becomes just watching. Are we running it hot without any awareness of caps, without any awareness of spending and pretending it doesn't exist? Or do we gradually or suddenly Start to have some acknowledgement that we need some type of protection on not letting this just absorb the whole economy until we run into a giant bust. Is that kind of fair?

[Ben Hunt]

That's exactly where it is. We need a balance. We need AI growth to be subordinated to our national economic goals. The difficulty we have today is that they are not being subordinated. They are being elevated into a war effort, into this Genesis project. So it's not going my way. Right? It's not going our way is how I'd rather describe it right now. It's going the techno oligarchs way. I hope that by getting the word out there, by getting people to step back and say, you know what? This doesn't work in the aggregate. And the rationale, the presentation they're using of a war effort, that's not right. I'm hoping that I get enough people to step back so we can start making a political effort, state by state to put some caps in place. I think if we can do that, I think it can build on itself and I think we can actually change, change the course of this mighty river of capital and energy that's currently only flowing towards one end, that being the AI data centers.

[Matt]

Okay, so part of this, and this is part of why I really wanted to get you to sit down and just talk about this essay. Because as somebody in the financial space talking to people at various levels of awareness, I don't want to say intelligence, not intelligence. It's just there's a whole class of professional investors who are just sick of talking about why they're underperforming the index. And if they have enough mag7 and there's a whole sector of people at my local pizza place and bar who are just sick of my utility bills went up again.

[Ben Hunt]

Yeah.

[Matt]

And they're not all connecting this to that. They're talking about much of the same story. Not, maybe not the underperformance relative to a benchmark for some hedge fund, maybe not that. But this reality of this build out is just starting to pick up steam. And a lot of people at a lot of layers of society are worried about the exact same problems. But it's not being covered this way. It's not being discussed through this way. What do we do? Aside from like sharing an article like this with somebody or explaining it, share a podcast episode like this with somebody. What else do we do to just try to have this conversation? Because I don't think people are threading that needle that this AI build out is actually attached to all these Fundamental inputs for everything that makes our country operate.

[Ben Hunt]

Well, that's why I think it's important to be four things. Right? It's. I think. I think that if you just do a headlong attack on the techno oligarchs and the AI data centers, I don't think that works. I think the vested interests are too great. What's the old line? If your pocketbook requires it, you don't see things.

[Matt]

Where you stand is a function of where you sit. All these expressions.

[Ben Hunt]

I think it's important to bring positive energy to this and to not be against AI but to be for energy expansion, to be for jobs coming from industrial and manufacturing reshoring. And I think that we can start talking about putting caps on electricity used by AI data centers. I think there is a political will for that, particularly in some states and some groups of states. I even saw in my home state of Alabama, which you would think would be the last place or one of the last places you'd get some sort of cap. Katie Britt, our senator, our junior senator, or may she say, I don't even remember Tommy Tuberville is the other one, so that I can't even talk about him. But Senator Britt's talking about, you know what, these data centers here in Alabama, they're using too much electricity. Katie Britt's talking about this. So I think there is the kernel, the germination of this idea that there should be limits on data center usage of electricity. Not trying to take us back to, you know, pre LLM days still allowing for growth, but allowing for growth in the rest of the economy too. I think there's a potential to be for that, particularly if it's coupled with, well, what is all that electricity going to go for? What's going to go for jobs through other manufacturing and reshoring of industrial facilities. What are we going to. Can we build more energy? Yes, we can. We want to build more energy from all sources. No more woke left anti energy policies and no more woke right anti energy policies. We need it all. We need it all and we can really move towards abundance. So I look like I say it's a long putt. It's going to be a difficult path. But I think there is a path and importantly is a path of four of positive energy as opposed to just being a path of fighting something like AI and the techno oligarchs. I want to be for something.

[Matt]

I like the sound of that. I like that. When I read this piece the first time through, I had a positive feeling coming out of it. I had the positive feeling of. I actually see the upward path of people understand this issue. Here's how you create jobs alongside this growth. And that's a really exciting path out. That's not just here we have to go running off this cliff.

[Ben Hunt]

It is.

[Matt]

And I hear it in what you're saying in Alabama too. So a crazy thing that I saw in Pennsylvania was the Amazon Web Services or whoever the data center was to going goes to the nuclear power plant, goes to the power company and says, here's our growth for next year. Here's how much we're going to need. The state regular regulator came in, state regulated entity and they were like, sorry, but no, like that line was put up. This is, I think close to a year ago at this point. They're like, we can't, we can't do that. We can't give you all your growth because then there's no more energy. And I think that that appetite is there. I just think it hasn't made it to any headlines yet. I don't think people know it's an option. I don't think people know how to exactly politically yet.

[Ben Hunt]

People don't know it's an option. I love that, Matt. People don't know it's an option. And it has to be an option and we have to start talking about it.

[Matt]

This World War AI piece I think lays out those options better than anything that I've read yet. So I want to suggest people go find it. I want to suggest people check out if you're in markets especially perfect. You want to see these, you want to see these storyboards, you want to see these pulses. You want to see how they're highlighting these stories in advance. And of course, if you want to be for something, I think that's what Panoptica AI is pretty much all about. Right, Ben?

[Ben Hunt]

Absolutely. Positive energy. That's how we get through this together.

[Matt]

All right, anywhere else you want to send people? We're going to put all the links in the comments before I let you go.

[Ben Hunt]

We got a lot of links. I mean, but obviously, you know, you can always just do Epsilon Theory. Epsilon Theory is kind of the hub for, for, for, for all of this and hope to see you there.

[Matt]

All right, come say hi. I'm going to have Ben back again in a month. We're going to update on this because this is a breaking and moving story, as we well know. Hey, if you write a note and there's a giant government intervention every time, within days, have a lot to talk about this year.

[Ben Hunt]

Ben no doubt.

[Matt]

No thank you for tuning into this episode. If you found this discussion interesting and valuable, please subscribe on your favorite audio platform or on YouTube. You can also follow all the podcasts in the Excess returns [e-mail] if you have any feedback or questions, you can contact [e-mail] no information on this podcast should be construed as investment advice.

[Shopify Advertiser]

Securities discussed in the podcast may be.

[Ben Hunt]

Holdings of the firms of the hosts or their clients.