The AI Bubble Nobody Wants to Talk About: Leverage, Tariffs, Inequality & the Coming Correction

The AI Gold Rush, the Great Depression, and Your 401(k): Are We Sleepwalking Into a Crash?

---

If you zoom out from the daily noise of stock tickers and viral AI demos, something unsettling comes into focus.

On one side, we have an economy riding high on artificial intelligence, data centers, and a handful of superstar companies. The stock market is at eye-watering levels. Nvidia is treated like a religious idea, not just a ticker symbol. Corporate boards are signing off on capex budgets that would make old-school CFOs faint.

On the other side, we have tariffs reshaping global trade, an affordability crisis for ordinary people, and a financial system quietly stuffed with leverage that’s harder to see than ever before.

The conversation in that script between Andrew Ross Sorkin and his interviewer isn’t just about history. It’s a warning label for right now. And for PyUncut readers—who care about long-term wealth, economic reality, and not getting wiped out at the top of a bubble—it’s worth unpacking carefully.

Let’s break it down.

---

1. The Ghost of 1929 Is Back—Just With Better Microphones

The Great Depression didn’t happen because of a single bad day on Wall Street. It started with a crash in 1929, yes, but it turned into a catastrophe because of a chain of bad choices:

• 25% unemployment by 1932
• Around 9,000 banks failed by 1933
• A generation of investors so scarred they never touched stocks again

The script draws a parallel between that era and today’s environment. The 1920s weren’t just “the Roaring Twenties” in terms of parties; they were a technology super-cycle:

• Automobiles
• Radio
• Telecommunications

RCA was treated the way the market treats Nvidia today—the company at the center of a new world.

Sound familiar?

We’re living through our own euphoric phase:

• AI models instead of radio
• Data centers instead of factories
• Cloud infrastructure instead of telephone lines

The point isn’t “tech = bad.” The point is: every bubble feels justified in the moment because the underlying technology is real.

Radio didn’t vanish after 1929. The internet didn’t vanish after the dot-com bust in 2000. AI won’t vanish after the next crash either.

The tech survives. The speculative capital often doesn’t.

---

2. The Real Threat Isn’t AI—it’s Leverage

If there’s one word that keeps coming up, it’s this: leverage.

“Every financial crisis, if I learned anything from covering 1929, covering 2008, it is leverage.”

We’re not just in a tech boom. We’re in a credit-fueled tech boom.

Big tech giants like Meta, Amazon, and Google are pouring tens of billions into AI and data centers. That’s one thing—they actually have cash flows. The more worrying part is the ecosystem growing around them:

• Real estate companies building campuses and data-center shells
• Energy providers scaling up power infrastructure to feed these data centers
• Construction companies, infrastructure contractors, and specialized service firms

Many of these players are borrowing aggressively to fund the build-out of “the future.”

Now add deals like this:

OpenAI commits $100 billion to Nvidia.
OpenAI doesn’t have $100 billion.
Nvidia finances OpenAI so OpenAI can pay Nvidia.

On paper, that’s “strategic partnering.” In practice, it’s leverage wrapped inside optimism.

Why does this matter?

Because if the economics don’t show up in time—if AI revenue and profits lag the capex cycle—the system doesn’t just “cool off.” It snaps:

• Stocks don’t just fall 20–30%
• Leveraged entities can’t pay back their loans
• That forced selling cascades into a broader confidence shock

You don’t blow up because a stock goes down. You blow up because the stock goes down while you owe more than you can pay.

That was true in 1929. It was true in 2008. It will be true next time, too.

---

3. The Shadow Banking Machine: The Risks You Can’t See

In the old world, leverage was reasonably visible:

• Companies borrowed from banks
• Banks were regulated
• The Federal Reserve could see where a lot of the risk was

Today, much of the action has migrated to what Sorkin calls the shadow banking system, especially private credit funds.

Instead of “Bank lends to Company,” it’s:

1. A fund manager raises billions from pension funds, insurance companies, and wealthy clients.
2. That private credit fund then lends directly to companies—often at higher rates and with bespoke terms.
3. Banks, ironically, still sit behind the scenes—helping lever up those private funds.

The catch?

“We don’t know about these funds. We don’t know what’s in them, we don’t know how they’re being valued.”

This matters because risk pricing depends on transparency. When nobody really knows who holds what, how it’s marked, or how interconnected it is, a credit event doesn’t just hurt one institution—it freezes trust across the system.

That’s how liquidity crises happen.

For a regular investor with a 401(k), all of this sounds abstract. But it translates into something very concrete:

• Sudden risk-off selling
• Forced deleveraging
• Broad market drawdowns that don’t care if you personally owned “good” companies

You can be holding a perfectly fine business and still see a 40% drawdown simply because someone else’s leverage broke first.

---

4. The AI Bubble: Real Technology, Distorted Economics

One of the most important nuance points in the script is this:

• AI is real
• AI will likely transform the economy over decades
• But the economics and timelines are not matching the hype cycle right now

Sorkin notes that many companies pouring billions into AI cannot actually pencil out the math in a sober, short-to-medium-term way. Projections are aggressive. Models assume:

• Rapid adoption
• High margins
• Sticky competitive moats

And yet, even a consulting firm like McKinsey finds that ~80% of companies using AI see no material bottom-line impact yet.

So what explains the frenzy?

Fear of missing out at the corporate level:

One path is: “We invest too little and fall behind forever.”
The other is: “We over-invest, but at least we stay in the game.”

Most CEOs today are choosing the second path.

This is the classic dynamic of capital cycles:

• Early, cheap capital floods in
• Everyone builds capacity at once
• Real economic value shows up more slowly and unevenly
• At some point, too much capacity + too much debt collides with slower-than-expected cash flows

That’s when correction turns into crisis—if the leverage is large enough.

---

5. Tariffs, Deglobalization, and the New “Protected” Economy

Layered on top of the AI frenzy is another structural shift: a retreat from globalization.

Tariffs—which many economists once treated as economic self-harm—are now increasingly seen as permanent:

• Trump uses tariffs to appeal to farmers and workers
• Biden keeps and adjusts many of the tariffs
• Corporate leaders who once hated tariffs now assume: “They’re here to stay.”

The logic is messy:

• Without tariffs, US car makers might be destroyed by cheaper, better Chinese EVs like BYD.
• With tariffs, consumers pay more and global supply chains get distorted—but certain domestic industries survive.

So, we end up with:

• Companies shipping goods to third countries and back just to reduce tariff costs
• Policy decisions that prioritize industrial resilience over pure economic efficiency

This can support certain sectors in the short term—autos, defense, semiconductors—but it also risks higher prices, inefficiencies, and trade tensions.

For investors, this means:

• You can’t just think in terms of “free markets.”
• You have to factor in policy-driven winners and losers.

And those policy decisions are happening in a world where AI itself is reshaping the services economy, potentially threatening the very white-collar jobs that used to feel safe.

---

6. Inequality, Capitalism, and the Fraying Social Contract

If there’s a moral undercurrent to all of this, it’s inequality.

The script makes an uncomfortable observation: many of the wealthiest people know there is deep resentment—but they often compartmentalize it.

“They think about it and then try not to think about it.”

This isn’t just a vibes issue. It’s existential for capitalism:

• Younger generations increasingly view “capitalism” as a broken game
• The American dream has shifted from “work hard, move up” to “pray for a lottery ticket”
• Social feeds sell get-rich-quick fantasies instead of sustainable upward mobility

In the 1920s, the stock market felt like the ultimate lottery ticket. Today, it’s:

• Meme stocks
• Options YOLOs
• Crypto promos
• AI startup fairy tales

The underlying message: normal effort isn’t enough anymore. You need a miracle.

That perception is dangerous. When people stop believing the system is fair or navigable with honest work, they:

• Withdraw
• Radicalize
• Or gamble even harder

None of these outcomes are stable for a system already over-levered and politically fragile.

---

7. Tax Codes, Philanthropy, and Why the Game Feels Rigged

The conversation also hits a nerve that rarely gets airtime in mainstream discussions: the architecture of the tax code.

Before anyone even touches the idea of a wealth tax, there are glaring structural choices:

• An estate tax system that allows massive fortunes to pass between generations with minimal friction
• Capital gains taxes kept lower in the name of “incentivizing investment”—even though ultra-rich investors will keep investing regardless
• Real-estate tax rules that allow depreciation deductions even as property values rise
• Philanthropy structures where the wealthy get large tax advantages while they decide the long-term fate of capital that never gets taxed at all

The point isn’t “philanthropy is bad.” Many foundations fund important work. But the aggregate effect is this:

• Trillions of dollars sit in lightly regulated or permanently tax-advantaged vehicles
• The public, via foregone tax revenue, is effectively subsidizing private priorities

It’s not hard to see why the average worker staring at rent, student loans, and stagnant wages feels like the game is tilted.

---

8. So What Does an Individual Investor Do in All This?

Here’s where it gets very real for PyUncut readers.

Most people don’t run private credit funds. They don’t get invited to Davos. They have:

• A 401(k)
• Maybe an IRA
• A brokerage account
• Some cash in the bank

And they’re being told two things at once:

1. “We’re in a bubble fueled by leverage and fragile policy.”
2. “It has historically paid to be a long-term optimist.”

Those two statements are not mutually exclusive.

Sorkin’s own stance is nuanced:

• As a journalist, he owns only broad mutual funds/indexes, not individual stocks
• For older investors (like his parents in their late 70s/early 80s), he’d want 10–20% in liquid, safe assets
• For people in their 40s/50s with long horizons and no near-term need for cash, it may make sense to “let it ride”—but with caution and some liquidity buffer

The key ideas for individuals:

1. Time Horizon Is Everything

If you need money in the next 1–3 years (home purchase, retirement income, tuition):

• You cannot treat that money as “risk capital.”
• A 30–40% drawdown hurts most when you must sell to pay real-world bills.

If your horizon is 20–30 years:

• Market crashes become violent chapters in a long story, not the ending.
• Historically, over long spans, the market has rewarded disciplined optimists more than doomsday skeptics.

2. Liquidity Is a Form of Psychological Armor

Having 10–20% in cash or near-cash isn’t about “timing the market.” It’s about:

• Being able to sleep when volatility spikes
• Avoiding panic selling at the bottom
• Having dry powder to buy when others are forced to sell

3. Don’t Confuse Narrative With Risk Management

The Nvidia and AI story might be structurally correct over decades. But that doesn’t mean:

• Nvidia’s current valuation is bulletproof
• AI spending can’t overshoot
• Leverage won’t amplify any corrections

You don’t need to “call the top.” You need to build a portfolio that doesn’t blow up if someone else’s leverage implodes.

---

9. The Market as a Political Guardrail

One of the more fascinating insights in the script is this: the market itself may be the strongest guardrail on presidential behavior.

We’ve already seen glimpses:

• Announce extreme tariffs → bond or equity markets react violently → White House “downshifts”
• Push too hard on policies that frighten capital → spreads widen → political pressure mounts

That doesn’t mean markets always “get it right.” But it does suggest:

• For all the talk of populism vs. elites, capital markets quietly discipline policy
• Politicians ignore them for too long at their own risk

It’s an odd twist: the same markets that can inflate bubbles and fuel inequality might also be the emergency brake when policy veers too far off the rails.

---

10. Where This Leaves Us: Cautious Optimism, Not Blind Faith

So, are we in an AI bubble?

Yes—in the sense that:

• Valuations in parts of the market are stretched
• Leverage is building in opaque corners
• Corporate investment is often fueled more by fear of missing out than clear cash-flow math

But also yes—we are in:

• A legitimate technological revolution
• A period where AI, like radio or the internet before it, will likely reshape productivity, business models, and daily life over decades

Both can be true:

• The technology endures
• The current pricing does not

For PyUncut readers, that suggests a simple but demanding stance:

• Stay in the game—especially through low-cost index funds and diversified exposure.
• Respect leverage—even if you’re not directly using it, you live inside a system that is.
• Hold liquidity—so you’re a forced buyer in crashes, not a forced seller.
• Think in decades—because all the historical evidence suggests that over multi-decade horizons, being a disciplined optimist beats being a clever pessimist.

And finally, do not forget the human element.

The Great Depression scarred a generation so badly that they never invested again. Our goal isn’t just to avoid losing money in the next downturn. It’s to avoid losing faith—in markets, in the possibility of upward mobility, and in our own ability to navigate complexity without resorting to lottery-ticket thinking.

AI will likely fulfill much of its promise. The question is whether your financial life will still be intact when the dust settles.

That isn’t decided by Nvidia’s next earnings call.

It’s decided by your asset allocation, your time horizon, your cash cushion—and whether, in the middle of all the noise, you can stay rational in a world that keeps trying to sell you a dream at 50x earnings.