Sequoia's Controversial Bet Just Proved That the 'AI Wars' Are Already Over

Reading time: 16 minutes | For: VCs, Founders, Investors

$25 billion. $350 billion valuation. And the most interesting part: Sequoia already invested in OpenAI and xAI. They're not betting on a winner. They're betting there won't be just one.

I've covered venture capital for eight years. I've never seen a move like this.

Sequoia doesn't hedge. That's been their philosophy for decades. Find the winner. Back them exclusively. Win big or learn.

Until now.

The Anthropic investment isn't just a check. It's a signal. And if you understand what it signals, you understand where AI is actually heading.

---

The Philosophy That Broke

Let me explain Sequoia's traditional approach through a poker analogy.

Most VCs play like cautious poker players. Spread bets across the table. Win some, lose some. Hope the math works out.

Sequoia plays differently. They play like a shark. Identify the hand that will win. Put everything behind it. Dominate the pot.

This philosophy produced Google. Apple. Instagram. WhatsApp. Stripe.

It also produced spectacular failures. But the wins were big enough that the strategy worked.

The philosophy assumes one thing: markets have winners. One dominant player emerges. Being right about who that player is matters more than being diversified.

For AI, Sequoia just abandoned that philosophy.

They already backed OpenAI. Now they're backing Anthropic. They were already connected to xAI through Elon's network.

That's not shark behavior. That's diversification. That's hedging.

Why?

---

The Market Structure Insight

Here's what Sequoia sees that most people don't.

AI isn't going to have one winner.

Not because the technology prevents it—the technology would allow dominance. Not because regulation prevents it—regulation might actually favor consolidation.

AI won't have one winner because the market structure doesn't support it.

Let me explain.

Dimension 1: Deployment Context

Different contexts need different AI.

Enterprise AI needs different safety guarantees than consumer AI. Government AI needs different compliance than startup AI. Healthcare AI needs different validation than entertainment AI.

One model optimized for everything is a model optimized for nothing.

The market will fragment by deployment context. Multiple leaders in each context.

Dimension 2: Capability Profiles

Models have trade-offs.

Claude is better at careful reasoning. GPT is better at certain creative tasks. Gemini is better at multimodal integration. Grok is better at… being Grok.

Users don't want "the best model." They want the best model for their task. And "best for this task" varies.

The market will fragment by capability profile. Multiple leaders in each profile.

Dimension 3: Economic Positioning

Different price points serve different markets.

Some users will pay $200/month for the best. Some need $20/month that's good enough. Some need free with ads. Some need on-premise at any price.

The market will fragment by economic positioning. Multiple leaders at each price point.

Dimension 4: Values Alignment

This one's harder to see but might matter most.

Anthropic prioritizes safety. OpenAI prioritizes capability. xAI prioritizes… anti-wokeness?

Users increasingly care about values. Which AI reflects my values? Which company do I want to support?

The market will fragment by values alignment. Multiple leaders for different value systems.

---