Monologue: AI Isn't Getting A Bailout

Summary

Better Offline – Monologue: AI Isn’t Getting A Bailout

Host: Ed Zitron
Date: December 4, 2025
Podcast: Better Offline (Cool Zone Media + iHeartPodcasts)

---

Episode Overview

In this solo monologue, tech industry veteran Ed Zitron addresses pervasive anxieties about the so-called “AI bubble”—specifically, the myth that AI companies or infrastructure are “too big to fail” and will inevitably be bailed out by the government if the market goes sour. Zitron draws sharp distinctions between the financial rescue packages of the past and the realities of today’s AI economy, breaks down the positions of major players like Nvidia, OpenAI, and Oracle, and warns listeners against letting fear and hype shape perceptions about the stability and importance of generative AI.

---

Key Discussion Points & Insights

1. AI Is Not “Too Big to Fail”—and No Bailout Is Coming ([00:52])

• Zitron responds to questions and fears that there could be a “TARP-like” government bailout for AI if/when the bubble bursts.
  • Quote:

    “A lot of people have emailed, messaged and plucked me that the natural end of the AI bubble is some kind of government bailout... I want to be clear about something. This is different. And a bailout is very, very, very unlikely.” (00:52)

• He summarizes the purpose of the 2008 TARP bailout (systemically necessary banks/companies; direct risk to the wider economy) and contends that AI firms don’t qualify.

2. Nvidia: Huge, But Not Bailable ([01:56])

• Nvidia’s dependence on AI-fueled GPU sales has led to astronomical growth expectations—far beyond what any bailout could fix.
  • Quote:

    “It isn't enough for Nvidia to continue selling lots of GPUs, it must continue to increase revenues by at least 60% year over year, every single quarter. Which means... it needs to make $91 billion in revenue in a year's time. And that's just in one quarter, by the way. This isn’t something a bailout can fix.” (02:32)

• The idea of the government endlessly buying up Nvidia’s GPUs is dismissed as fantasy.

3. OpenAI: Not Systemically Necessary ([03:58])

• The myth of OpenAI being “too big to fail” is torched:
  • OpenAI's business isn’t big enough to threaten the economy, and its revenue/burn rate is too small relative to tech giants.
  • If OpenAI vanished, it would cause symbolic harm and chill investment, but not economic catastrophe.
  • Quote:

    “Too big to fail means that something may kill our economy or markets if it dies. And as loud and annoying as clammy Sam Altman might be, OpenAI’s death wouldn’t kill either of those.” (04:13)

• Breakdown of OpenAI’s scale:
  • Promised $1.4 trillion in compute—but revenues are comparatively tiny (projected max $13B; likely under $4.5B through September).
  • Heavy losses on inference ($8.67B in that period).
• Microsoft’s stake in OpenAI means it could absorb the impact easily.
  • Quote:

    “Microsoft owns OpenAI, IP and Research, which would mean that Microsoft would pretty much absorb those 800 million weekly active users directly into Copilot, and that’s if it even chose to retain them…” (05:48)

4. Oracle: A Possible Bailout Candidate—but Unlikely ([06:46])

• Oracle is heavily leveraged, with billions in debt tied to data center projects for OpenAI.
  • If projects fail, Oracle could be in genuine trouble—particularly since it’s politically connected.
  • Quote:

    “There is one company I could see if any bailout happens. And I’m really hesitant to say I think there’s any chance. Oracle.” (07:16)

• But even here, Zitron stresses a bailout wouldn't support the wider economy and would just delay defaults.

5. Doomerism = Free Marketing for AI ([08:10])

• Zitron strongly criticizes panic-mongering about AI’s systemic risk, calling it inadvertent hype for the industry.
  • Quote:

    “You are helping them. You are helping them market their services by making them seem systemically relevant when they are not. By making them seem like they’re innovative when they’re not.” (08:45)

• Warns against mistaking academic studies about “AI replacing skills” for concrete labor market shifts.

6. Lack of Actual Economic Impact from AI ([06:18], [08:57])

• Actual AI revenue is comparatively small: ~$60B in 2025, peanuts next to Amazon and Google’s quarterly earnings.
• Real money is flowing mainly into speculative data center construction, which may eventually warrant a targeted rescue—but only if a major financial institution is directly endangered.

---

Notable Quotes & Moments

• On Nvidia’s insatiable growth:

  “Jensen Huang needs all your money, and he needs it now. Every quarter, forever.” (03:04)

• On “too big to fail” logic applied to OpenAI:

  “OpenAI has promised $1.4 trillion in compute deals... Yet the reality is that this company isn’t even going to hit its $13 billion in projected revenue this year and will burn billions of dollars in inference alone, and doesn’t really have a path to profitability.” (04:05)

• Sarcasm on government buying GPUs:

  “Are you saying that Trump is just going to go up? ‘We're going to buy GPUs for everyone. We're going to buy them every year and every year by so many of the most we've ever.’ No, it’s not going to happen.” (02:51)

---

Important Segment Timestamps

• [00:52] – Introduction and debunking the “AI bailout” parallel with TARP
• [01:56] – Deep dive: Nvidia’s business model and the absurdity of a tech bailout
• [03:58] – OpenAI’s actual economic footprint
• [06:18] – Tech industry and genuine economic exposure
• [07:16] – Oracle as the only plausible bailout candidate, but still very unlikely
• [08:10] – The dangers of “doomerism” and AI hype
• [08:57] – Critique of shoddy media coverage and faulty “AI replaces jobs” studies
• [09:50] – Brief mention of upcoming multi-part Nvidia investigation and listener greetings

---

Closing Thoughts & Looking Ahead

• Ed Zitron previews a multi-part series delving further into Nvidia, explicitly noting it's “nothing like Enron,” and hints at upcoming interviews.
• He thanks listeners for their emails and support, offering a candid look at his recent health and stress.

---

Summary Takeaways

• AI companies are not systemically necessary in the same way as banks or automakers were in the 2008 crisis; if Nvidia, OpenAI, or Anthropic crash, there’s no pressing national need to intervene.
• Most AI revenues are small change versus the broader tech sector; the perceived risk is wildly overstated.
• Doomer rhetoric about AI bailouts is more marketing than analysis—perpetuating myths that benefit tech’s biggest players.
• If a bailout occurs at all, it’d be for infrastructure players (like Oracle) with politically strategic ties and major financial obligations—but even here, Zitron is highly doubtful.

---

Ed Zitron’s Tone

Ed’s delivery is caustic, direct, and often laced with sarcasm and exasperation, especially toward tech media hype and “doomerist” narratives—balanced with brief expressions of appreciation for his listeners and sharp, unvarnished industry critique.

---

For listeners worried about a tech crash or government bailouts, Zitron's message is clear:
Don’t buy the hype—AI, as currently configured, just isn’t big or important enough to trigger a TARP-sized rescue, despite what the industry (or its critics) might want you to believe.

Transcript

[A]

This is an iHeart podcast.

[B]

Guaranteed Human running a business is hard enough. Don't make it harder with a dozen apps that don't talk to each other. One for sales, another for inventory, a separate one for accounting. That's software overload. Odoo is the all in one platform that replaces them all. CRM, accounting, inventory, E Commerce, hr. Fully integrated, easy to use and built to grow with your business. Thousands have already made the switch. Why not you try Odoo for free at odoo? That's odoo.com.

[C]

Call zone media.

[D]

Hello and welcome to this week's Better Offline Monologue. I'm your host Ed Zitron.

Better Offline.

A lot of people have emailed, messaged and plucked me that the natural end of the AI bubble is some kind of government bailout, something resembling the TARP program that the bailed out financial institutions overleveraged during the great Financial crisis. And I want to be clear about something. This is different. And a bailout is very, very very unlikely. As a reminder, TOPS stood for Troubled Asset Relief Program and existed unsurprisingly to buy the distressed assets that banks had invested in during the great financial crisis. Primarily the ridiculous bets made on the housing market made up of people who had got mortgages they could not pay for. Now these bailouts existed to stabilize banks that were going out of business along with consumers money being set on fire if they did so, as well as auto manufacturers like GM and Chrysler that the contagion had spread to. And while I'm not going to go into massive amounts of detail, I haven't got all day, I can make it pretty simple. These bailouts existed to stabilize systemically necessary companies that had they died, would have tanked the US economy and markets at the same time and also lost people's money. Unlike the great Financial Crisis, very few of the major players in AI are either systemically necessary or in danger of dying as a result of the AI bubble bursting. Furthermore, there really isn't anything to bailout. So let's start with Nvidia, the largest company on the stock market. While GPU sales make upwards of 88% of its revenue, Nvidia will not die. If the AI bubble bursts, though, it will see a massive contraction as a result of its biggest business line collapsing. It's also important to know that Nvidia's massive stock price is a direct result of its incredible growth. It isn't enough for Nvidia to continue selling lots of GPUs, it must continue to increase revenues by at least 60% year over year, every single quarter. Which means that for Nvidia to continue being the most beloved boy in stock in the in history of the stock market, it needs to make $91 billion in revenue in a year's time. And that's just in one quarter, by the way. This isn't something a bailout can fix. Even in the most crony capitalist fantasy you can think of, a bailout would have to continually sink tens of billions of dollars directly into Nvidia, buying GPUs in increasing amounts every three months in perpetuity to continue its valuation. And while some of you may say Trump will do the you'd living in a boring doomerist fantasy based on nothing, and I'm sick of hearing about it. Good Lord, listen to me. TARP was a $700 billion program that only ended up investing around $443.5 billion or so. Or put another way, less than the $500 billion of orders that Nvidia CEO Jensen Huang claims are on Nvidia's books through the end of 2026. Are we meant to do this forever? Is that the plan? Are you saying that Trump is just going to go up? We're going to buy GPS for everyone. We're going to buy them every year and every year by so many of the most we've ever. No, it's not going to happen. Though I guess he has kind of done that with the Department of Energy. You've probably heard of this deal, by the way. It's 100,000 GPUs that isn't going to do shit. Jensen Huang needs all your money, and he needs it now. Every quarter, forever. Okay, okay. It's not going to happen.

Now. Okay. Another annoying talking point you've probably heard is that OpenAI is somehow too big to fail, when in fact it's too small to pull apart into enough parts to eat. OpenAI has promised $1.4 trillion in compute deals, convincing Oracle and others to build massive data centers in its honor. Yet the reality is that this company isn't even going to hit its $13 billion in projected revenue this year and will burn billions of dollars in inference alone, and doesn't really have a path to profitability. Too big to fail means that something may kill our economy or markets if it dies. And as loud and annoying as clammy Sam Altman might be, OpenAI's death wouldn't kill either of those. Let's break it down. OpenAI dying would be symbolically lethal to the large language model era and create an immediate and permanent chill in AI investments. Don't get me wrong, its death would also likely be an existential threat to Crusoe, the company building its data center in Abilene, Texas for Oracle. I spoke to someone on the ground there and they say if the money runs out, everyone just stops and goes home. They'll be fine. I don't know the extent of Crusoe's revenues, however, but I do know that Oracle is on the hook for a billion dollars a year for Abilene for 15 godd. And Oracle has also taken on massive amounts of debt to build Abilene and other data centers in Shackelford, Texas and I think somewhere in Wisconsin, New Mexico one now and started this with an $18 billion bond sale and an in the works $38 billion debt package to build data centers for OpenAI. While Oracle does run the risk of default and credit default swaps, which is people betting again. Well, betting that they might default on their obligations, that's going up. The risk is there, but they're still a real company that makes real things. Microsoft owns OpenAI, IP and Research, which would mean that Microsoft would pretty much absorb those 800 million weekly active users directly into Copilot, and that's if it even chose to retain them and just shut down ChatGPT, directed it to a very flimsy version of Copilot, and didn't let people do half the shit they do with it. Bailing out OpenAI would not do anything, in part because OpenAI burns billions of dollars and doesn't really do much in the large economy itself. Its revenues, which I estimate to be at most $4.5 billion through the end of September. By the way, if you're listening to this and you think that my revenue numbers for OpenAI and my inference numbers were delayed somehow, it's a cruel accounting. Stop fucking with me. I know what I'm doing. Anyway, those revenues are inconsequential, and the $8.67 billion it spent on inference in that period is while a large chunk of Azure revenue from Microsoft only really accounting for 3.88% or so of the $223 billion Microsoft has made in that period. Losing OpenAI would hurt but hardly threaten the health of anyone in Redmond. The same goes for Anthropic or any other generative AI company. These companies aren't really spending that much, nor are they fueling much of any actual real economic activity. There's only around $60 billion of actual AI revenue in 2025. In a tech industry where Amazon made $180 billion and Google 102 billion in their last quarters, the disappearance of LLMs writ large would have a minuscule effect on actual real dollars entering the economy. Other than investments in data centers, which is where I think there's if any, some kind of bailout. As I mentioned recently, there's been about $50 billion or more even every quarter for the last three quarters going into building data centers. And I think that many of these projects end up defaulting. There's a chance that a chance that these are somehow bailed out if and only if they provide an actual existential risk to a major asset firm, a Blackstone or an Aries or what have you. I don't know if Ares is in them. I'm just naming them at this point. Nevertheless, there is one that I the one company I could see if any bailout happens. And I'm really hesitant to say I think there's any chance. Oracle. Oracle, I think is the one chance. Oracle is currently massively leveraged and going into debt to the point it's eating into its free cash flow. And it's a close ally of the Trump administration. I could see them getting a lifeline to escape these ridiculous loans. And while you may say there's some sort of grand plan where the administration will agree to pay these companies to keep them alive, I don't see it happening. Is doing so won't support the economy. The markets really do anything other than prolong the inevitable. It's not going to happen. And I don't know. I'm gonna. I want to phrase this in a nice way because I don't want it to seem like an attack every time you say something like this. You're doing marketing for the AI industry. Doomerism is marketing. At this point with AI, I'm so scared of what they'll. They'll create. They are going to get bail out. Open AI is too big to fail. All these are extrapolations of your fear. I get it. They're also marketing. You are helping them. You are helping them market their services by making them seem systemically relevant when they are not. By making them seem like they're innovative when they're not.

Honestly, I need you to just live in reality and look at what's actually happening, which is sweet fuck all. In many cases. Every one of these wanky studies you hear come out so on. It's like 10% of jobs can already be automated by AI. CNBC should be fucking ashamed of themselves for running a story about it. It's a PhD. Congrats to the better offline Reddit user who saw this. It's a PhD paper and also it's referring to random skills rather than jobs. And skills and jobs are two different fucking things. I'm so goddamn tired of this era. Good lo. But nevertheless, the next two weeks are going to be real interesting. I'm working on a three or four part about my worries around Nvidia and I'm going to tell you all about Nvidia and how not similar they are to Enron. Nvidia is nothing like Enron or Lucent or WorldCom or Windstar or all these other companies I'm going to go into as well. And tomorrow I've got a fun interview with Nathan Grayson of Aftermath your Love. I might next week take a knee and just run the better offline in studio thing I just did with with Lisa from cnn. Or I might do the three or four part. It really comes down to how I'm feeling. It's been a rough few weeks. Health wise. I'm doing fine. I was just being sick and I've just been quite stressed. I must again say thank you for all the lovely emails and messages you've been sending. You're all so lovely. I love you all. Couldn't be lucky with the listeners I've got. Tell your friends about the show, tell them about the newsletter, tell them about me, tell them about my strange proclivities. But also keep listening. I really do appreciate all of you. Enjoy tomorrow's episode and whatever I put out next week and there will be a monologue regardless. Unless I do the three or four par, in which case that's going to be your monologue for the week. And yes, by the way, I realize me just talking is always a monologue. I get it. You're so smart. You are though. Really. Thank you for listening.

[C]

Amazon five Star Theater Presents Real Customer Reviews Performed by Ed Helms Tonight's Review Tactical Jacket I was living a simple life.

[E]

Didn't get out much.

[C]

Then I bought this jacket and everything changed. Women came flocking to me from lands domestic and foreign on the 245 day sailboat voyage home. I was attacked. Buy a shark. I knew it was the jacket he was after giving up the jacket in exchange for my life. 5 stars Amazon Customer 69 Shop the perfect Gift this Holiday on Amazon.

[A]

Janice.

[F]

Torres here and I'm Austin Hankwitz. We host the podcast Mind the Business Small Business Success Stories produced by Ruby Studio in partnership with Intuit QuickBooks.

[A]

We're back for season four to talk to some incredible small business owners.

[F]

The big thing about working at tech is that it's ever evolving, ever changing. Everyone's a rookie, that's how fast the industry is changing. So what I'm really excited about is to be part of that change. So listen on the iHeartRadio app, Apple Podcasts, or wherever you get your podcasts.

[E]

This podcast is brought to you by FedEx. The new power Move hey, you know those people in your office who are always pulling old school corporate power moves? Like the guy who weaponizes eye contact. He's confident, he's engaged, he's often creepy. It's an old school power move. But this alpha dog laser gaze won't keep your supply chain moving across borders. The real power move? Having a smart platform that keeps up with the changing trade landscape. That's why smart businesses partner with FedEx and use the power of digital intelligence to navigate around supply chain issues before they happen. Set your sights on something that will actually improve your business. FedEx the new power Move hey, Ryan.

[G]

Reynolds here for Mint Mobile. You know one of the perks about having four kids that you know about is actually getting a direct line to the big man up north. And this year he wants you to know the best gift that you can give someone is the gift of Mint Mobile's unlimited wireless for $15 a month. Now you don't even need to wrap it. Give it a try at mintmobile. Com.

[A]

Switch upfront payment of $45 for three month plan equivalent to $15 per month required new customer offer for first three months only. Speed slow after 35 gigabytes if network's busy. Taxes and fees extra. See mintmobile.

[D]

Com.

[A]

This is an iHeart podcast. Guaranteed Human.