🎭 Big Short on War vs. Palantir Rebound:

Which Side Are You On?

Michael Burry is back — not with a hedge fund, but with a paid Substack manifesto telling the world that AI is the next dot-com bubble.

He’s short “war stocks,” skeptical of AI profits, and openly bearish on Nvidia and Palantir — two names that have become the poster children of this AI cycle.

So…

Is Burry right, or is this just classic Burry contrarian theatre? 🎬🧐

Let’s break it down below 👇

⸻

🧨 1. Burry’s Case: “AI = Dot-Com 2.0”

Burry’s argument is basically:

(1) Exponential hype ≠ exponential profits

Investors are projecting infinite future earnings based on early breakthroughs.

This happened in 1999 too — adoption is fast, monetization is slow.

🧪➡️💸 takes years.

(2) Hardware cycles eventually compress margins

Nvidia’s current margins are historically abnormal:

• 75–80% gross margin on H100 class chips

• Supply constraints artificially boost pricing

• Competition (AMD, custom ASICs, in-house accelerators) inevitably erodes profits

Chip cycles always normalize. Always. 📉🎡

(3) “AI will save everything” = peak sentiment indicator

Every sector claims AI will transform their P&L.

In 1999, everyone said the same about “the internet.”

AI has real value — but not at this velocity.

(4) Valuations assume a straight-line adoption S-curve

Reality: adoption is S-curve, but revenue is staircase-shaped.

Corporate budgets lag hype.

From Burry’s perspective, AI is real, but the stocks are fantasy. 💭💸🐂

⸻

🧠 2. So… Are Nvidia & Palantir Overvalued?

Let’s examine each.

⸻

🟣 NVIDIA (NVDA) — AI Monopolist… but at what price?

Bullish fundamentals:

• 90%+ share in AI training chips

• CUDA dominance is a near-monopoly

• Demand is still outrunning supply

• $200–250B in annual data center TAM expanding rapidly

• Blackwell ramp in 2025 is monstrous

🔥📦🚚

But…

• NVDA trades at ~35–45x forward earnings depending on scenario

• Market prices in multi-year 30–40% revenue growth

• Hyperscalers (AWS, Google, Meta, Microsoft) are pushing for:

• In-house accelerators

• Open-source alternatives

• Lower capex intensity

• Margins will compress (history of semis says so)

• China restrictions cap part of growth

Verdict:

NVDA is not crazy like dot-com stocks were — it has real earnings —

but valuation assumes perfection.

🟡 Mildly overvalued but not in a bubble.

⸻

🔵 PALANTIR (PLTR) — AI Defense Darling but Early in Monetization

Bullish side:

• Strong national security moat

• Gotham + Foundry + AIP = sticky enterprise platform

• AIP adoption is accelerating

• U.S. gov + NATO conflict cycle = steady revenue runway

• Operating margins finally improving

🛡️📊⚙️

But…

• PLTR trades at 70–100x forward earnings

• Commercial revenue still inconsistent

• AI platform adoption is high on POCs, low on large-scale deployments

• Government spending is stable, not hyper-growth

• Real AI monetization is still early

Verdict:

Yes — Palantir is overvalued relative to current earnings.

The story is huge, but the numbers aren’t huge yet.

🔴 Story stock priced like a mega-cap margin machine.

⸻

🔥 3. Is the AI Boom = Dot-Com 2.0?

Short answer:

❌ No — fundamentals today are far stronger.

BUT

⚠️ Pockets of exuberance absolutely rhyme with 1999.

🚀 Why It’s NOT the Dot-Com Bubble

• AI companies today have real revenue and cash flow

• Cloud hyperscalers have balance sheets the size of nations

• AI is already deployed in production (search, ads, logistics, drug discovery)

• Data infrastructure and compute demand are measurable

• Customers are real enterprises, not pets.com

This is a genuine technological platform shift.

💥 Why It IS Similar

• Insane valuations on long-dated promises

• Retail FOMO on anything labeled “AI”

• Capex cycles are being extrapolated infinitely

• Everyone is claiming “AI will redefine our business”

• Margin compression risk ignored

It’s like 2004 Google IPO + 1999 pattern of hype blended together.

⸻

🧭 4. My Position: Bullish AI, Bearish Prices

This cycle is fundamentally real, but the speed and the valuation multiples are the danger.

• AI is transformative ✔

• Nvidia is dominant ✔

• Palantir has a moat ✔

But:

• Multiples are stretched

• Profit cycles take time

• Competition always arrives

• Rate cuts + liquidity have pumped everything

We’re in an AI-led bull market —

but parts of it are priced like a bubble. 🎈🔥

The next 2 years will separate:

• Survivors (NVDA, MSFT, GOOG, AVGO)

• Narrative stocks (PLTR, UPST, AI, smaller caps)

• Casualties (AI ETFs filled with junk)

⸻

🎙️ TL;DR (For Your Inner Trader)

• NVDA: slightly overvalued but justified by dominance

• PLTR: overvalued relative to near-term earnings

• AI sector ≠ dot-com bubble but AI valuations = bubble pockets

• Burry is early — as usual (and may still be right later)

• AI boom is real, but the market priced in 2030 earnings today