Simplify Training with AI-Generated Video Guides

Are you tired of repeating the same instructions to your team? Guidde revolutionizes how you document and share processes with AI-powered how-to videos.

Here’s how:

1️⃣ Instant Creation: Turn complex tasks into stunning step-by-step video guides in seconds.
2️⃣ Fully Automated: Capture workflows with a browser extension that generates visuals, voiceovers, and call-to-actions.
3️⃣ Seamless Sharing: Share or embed guides anywhere effortlessly.

The best part? The browser extension is 100% free.

Check out Guidde

Hey, Josh here. A lot just got announced this weekend. Check these stories out.

Peter Thiel just dumped every single share of Nvidia he owned. All 537,000 of them. Gone.

Meanwhile, Google just announced they're dropping $40 billion—with a B—on data centers in Texas.

So what is going on? Are we watching the smartest investors in tech run for the exits, or are they just making room for the real players who understand where this is actually heading?

Listen, I've been watching this AI boom with the same mix of fascination and dread you probably feel scrolling through your feed. Every day brings another breathless announcement about artificial intelligence changing everything. And look, some of it is real. But Thiel's exit—combined with the $1.8 trillion that just evaporated from AI stocks in early November—suggests we might want to pump the brakes on the hype train.

Let's break down what's actually happening.

Here's what Thiel did: Between Q2 and Q3 of 2025, he slashed his total equity holdings by two-thirds. We're talking from $212 million down to $74.4 million. Portfolio turnover exceeded 80%. This isn't trimming around the edges—this is a fundamental restructuring.

And he didn't go to cash. That's the interesting part. He moved into Tesla, Microsoft, and Apple. The megacaps. The boring, established giants.

Why does this matter? Because Thiel himself has said that Nvidia is "making all the money while everybody else is collectively losing money." He knows the chip maker is dominating. Nvidia commands 92-94% market share in discrete GPUs. They're printing money with 50% net margins and 70% gross margins. Jensen Huang just announced $500 billion in AI chip orders.

But Thiel's out anyway.

At the All-In Summit back in September 2024, he said something that keeps echoing: the current environment feels "uncomfortably close to 1999." As in, dot-com bubble 1999. The kind where everyone gets rich on paper until suddenly no one is rich at all.

Let's talk about OpenAI for a second, because this is where things get really interesting.

OpenAI—you know, the company behind ChatGPT, arguably the most important AI company on the planet—lost $11.5 billion in Q3 2025 alone. That's nearly equal to their entire first-half losses of $13.5 billion. They generated $4.3 billion in revenue during that time, which means... well, you do the math.

The thing is, it gets worse. According to their own projections, OpenAI expects to accumulate roughly $74 billion in operating losses in 2028. Cumulative cash burn through 2029? Over $115 billion. In 2025, they're planning to spend $22 billion against $13 billion in sales. That's $1.69 spent for every dollar earned.

Here's the kicker: Nvidia is making a fortune selling shovels while the gold miners go bankrupt.

This is the paradox that keeps serious investors up at night. The infrastructure layer is phenomenally profitable. The application layer is hemorrhaging cash. And everyone keeps saying, "Just wait, the monetization is coming." But when? And at what scale?

Remember Michael Burry? The guy who called the 2008 housing crisis, made famous by "The Big Short"? Yeah, he's back.

In Q3 2025, Burry disclosed $1.1 billion in put options against Nvidia and Palantir. Let me translate: he's betting these companies are going to crash. Hard.

The Palantir bet is particularly spicy. The company just posted strong quarterly earnings. Everything looks good on paper. But it's trading at a forward P/E ratio of 280x. For context, Nvidia trades at about 20x forward earnings. Burry's basically saying, "I don't care how good the business is—that price is insane."

And then, after two years of silence on X, Burry posted this on October 30th: "Sometimes, we see bubbles. Sometimes, there is something to do about it. Sometimes, the only winning move is not to play."

Subtle, right?

Here's something that should concern you: The "Magnificent Seven"—Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, and Tesla—now make up approximately 35% of the entire S&P 500's market capitalization. Nvidia alone accounts for over 8% of the index.

During the dot-com bubble peak, tech concentration hit around 25%. We're past that now.

The S&P 500 Equal-Weighted Index, which treats every stock the same regardless of size, actually declined 1.75% in early November while the regular market-cap-weighted index stayed elevated. Translation: Most stocks are going down. Only the massive tech companies are holding everything up.

What happens when they stop going up?

Between early and mid-November 2025, AI-linked equities destroyed approximately $1.8 trillion in value worldwide. The top eight AI stocks in the U.S. lost $1.2 trillion in a single week. That was the sharpest tech setback since Trump's tariff announcement in April.

Tesla dropped over 5%. Nvidia fell more than 3%. Alphabet shed over 2%. The Nasdaq Composite fell 1.8% in one day and headed for a weekly loss of 5.5%.

We're watching wealth disappear in real-time.

So against this backdrop—Thiel running, Burry betting against everyone, $1.8 trillion evaporating—Google drops the news that they're investing $40 billion in Texas data centers through 2027.

Three new facilities. One in Armstrong County near Amarillo, two in Haskell County near Abilene. Google's largest U.S. state investment ever.

Why Texas? Land is cheap. Energy is abundant. Regulations are friendly. The state already has transmission infrastructure for renewable energy. Plus, OpenAI and Anthropic are also building there. Texas is becoming the Silicon Valley of AI infrastructure.

Google's trying to be smart about it. They're partnering with Intersect and TPG Rise Climate to build solar and battery storage alongside the data centers. They've secured 6,200 megawatts of new energy generation capacity. They're training 1,700 electrical apprentices by 2030—more than doubling Texas's pipeline of new electricians.

But here's the problem: Texas might not be able to handle it.

The Electric Reliability Council of Texas (ERCOT) projects peak electricity demand could more than double by 2031. Data centers alone could add 86 gigawatts of new demand. Some projections go as high as 218 gigawatts by 2031. The current record, set in August 2023, was 85.5 gigawatts.

A University of Houston analysis warns that without massive infrastructure investment, Texas could face an electricity deficit of up to 40 gigawatts by 2035. Joshua Rhodes, an energy expert at UT Austin, put it bluntly: "I just don't think we can physically build that much infrastructure that fast."

Lead times for gas turbines have stretched to seven years. The renewable energy supply chain is getting squeezed by tariffs. And the Trump administration just reduced tax credits for solar and storage, threatened clean energy subsidies, and restricted federal renewable funding.

Google's betting $40 billion that they can make this work anyway.

Economists are converging on three scenarios for how this plays out:

Scenario 1: The Soft Landing
Markets gradually recognize that AI's real economics differ from the hype. Valuations adjust downward in a measured way. Investment decelerates but doesn't crash. Companies that actually generate profit consolidate gains. The speculative players fade into obscurity. We all breathe a sigh of relief.

Scenario 2: The Bubble Burst
Investor confidence erodes when it becomes clear that many AI companies can't achieve profitability at current capital deployment levels. Capital Economics predicts this could happen in 2026, with valuations correcting by more than 30%. Rising interest rates and inflation compress multiples. Things get ugly, but not catastrophic.

Scenario 3: The Massive Contraction
Think dot-com collapse. Retail investors eventually recognize the overvaluation across the sector and trigger a cascading sell-off. The Magnificent Seven get taken down. A decade of tech sector outperformance unwinds. Markets spend years recovering. Everyone who bought at the top becomes a cautionary tale.

The challenge with predicting bubbles is that timing is everything—and impossible. The dot-com crash took years to fully unwind. If you called the bubble in 1998, you were right... but you missed two more years of gains. If you stayed in until 2001, you lost everything.

Here's what I think is happening: Both Thiel and Google are probably right.

Thiel's looking at current valuations in the AI ecosystem and seeing peak risk. He's watched Nvidia multiply 13x since 2023 while the S&P 500 only gained 1.8x. He's seeing OpenAI burn through cash faster than anyone thought possible. He's watching the Magnificent Seven hit concentration levels that exceed the dot-com bubble. And he's thinking, "I'd rather be in stable, cash-generating businesses than betting on whether AI companies figure out monetization."

Google, meanwhile, is looking at the same data and saying, "Long-term demand for AI compute capacity is real, and we need to secure infrastructure now or get left behind." They're not betting on valuations. They're betting on physics—that AI requires massive amounts of electricity and hardware, and whoever controls that infrastructure controls the future.

Both can be true simultaneously.

The real question isn't whether AI is overhyped—parts of it definitely are. The question is whether the infrastructure spending can be monetized at the scale necessary to justify these investments. If AI applications eventually figure out profitability and adoption reaches the levels everyone's projecting, Google's $40 billion looks prescient and Thiel's exit looks premature.

If the bubble deflates and capex spending contracts sharply, Thiel will look like a genius and Google will have some very expensive, very empty data centers in Texas.

Real AI demand exists. The technology works. Productivity gains are measurable. But the gap between "this is useful" and "this justifies a $5 trillion market cap" is enormous.

Remember when Nvidia became the first company to hit $5 trillion in market value on October 29th? That happened less than three weeks before $1.8 trillion evaporated from AI stocks.

The infrastructure layer—Nvidia, the data centers, the electricity grid—will probably be fine. Someone's going to need chips and computing power regardless of whether ChatGPT-47 figures out sustainable monetization.

But the application layer? The thousands of AI startups burning through venture capital? The enterprise software companies bolting "AI-powered" onto every product description? That's where the carnage happens if this goes south.

What we're watching is the classic innovator's dilemma playing out in real-time. The picks-and-shovels sellers (Nvidia) make a fortune. The prospectors (OpenAI, enterprise AI startups) mostly go broke. A few strike gold. Most don't.

Thiel's betting the prospectors are overvalued. Google's betting they'll be one of the ones who strikes gold.

Place your bets accordingly.

The Bottom Line: When legendary investors start making dramatic portfolio moves, it's worth paying attention—not because they're always right, but because their decisions reveal how the smartest money is thinking about risk. Thiel sees 1999. Google sees 1995. The question is whether we're at the beginning of something transformative or the peak of something about to collapse.

Either way, it's going to be fascinating to watch.