Four Takeaways from Claude Founder House at SF Tech Week

After spending time at Anthropic’s Claude Founder House, particularly the discussions on Building for the Next Model with Jared Kaplan and Building AI-Native Companies with Sameer Dholakia, four convictions stood out to me.

1. Build for the next model, not today’s limitations

Model capabilities are advancing faster than traditional product cycles. Don’t build your moat around what AI cannot do today.

Build products that become more valuable as models get smarter.

2. Build AI-native organizations from Day One

An AI-native company is not a traditional company with AI tools added.

Rethink team size, talent density, decision-making, and how work is divided between humans and AI.

The organizational advantage could be as consequential as the technology itself.

3. Think global from Day One

The traditional playbook was sequential: win one geography, build a local team, then expand market by market.

AI is changing the economics of localization, distribution, and cross-border operations.

Think globally from the beginning. Test multiple markets early. Concentrate resources where traction is strongest.

4. Speed is essential. Focus is the multiplier

When model capabilities and competitive dynamics change within months, slow execution is expensive.

But speed without focus is just accelerated distraction.

The winners will move exceptionally fast while remaining disciplined about what not to pursue.

My biggest takeaway:

Build for tomorrow’s intelligence. Organize differently. Think globally. Move fast. Stay relentlessly focused.

The next generation of AI companies won’t just build different products. They’ll be built differently.