Paul Graham wrote in 2002: "The recipe for great work is very exacting taste, plus the ability to gratify it."
For most of human history, the people with the taste could not gratify it. That sentence has become a lot more interesting.
https://t.co/NRFr3tFCTa
In 2001, I told Dick Clark our target audience was twenty-five to forty-four.
He paused. Then he said: "I hope you don't mind, but my app will target a different group."
He was seventy-two. He understood the moment better than I did. ๐งต
In 2026, that gap has collapsed.
Claude Code. Lovable. Cursor. Not productivity tools for developers. Real creative tools for people who think in products but could not build them. Creative instinct was always the rare thing. Tech the obstacle. The obstacle is mostly gone.
The strongest startups recruit a portfolio of Market Discovery Partners, not a handful of design partners.
A single customer provides anecdotes. A portfolio provides patterns. Patterns reveal markets. Markets create companies. Full take: https://t.co/oE97rZc7BO
There is a difference between a Design Partner and a Market Discovery Partner.
Design Partners give you features.
Market Discovery Partners show you the market.
Most startups are recruiting the wrong one.
The question nobody answers directly: how long should a Market Discovery Partner remain one?
Shorter than founders think.
When they stop teaching you about the market and start asking for a custom report, they have succeeded. Graduate them to paying customers. Celebrate it.
The Frankenstein product is not a mystery.
It happens when founders confuse customer feedback with product strategy. Customers are experts in their problems. Not your market. Not your business model. Signal, not strategy.
The pattern: the foundational platform matters, but enormous value accrues to the layers around it.
The question is not who wins AEO. It is which layer you want to own.
We know where we are investing. Building one of them? [email protected]
Full essay ๐ https://t.co/Zn3LhNRyOo
Google won search. But nobody ever truly won SEO.
That one sentence is worth more than $30 billion, and it explains where the real value in AI is about to go.
Here is what we are watching.
3. Streaming and AI.
Discovery is moving from the search box to the answer.
Nobody wins streaming discovery, the same way nobody won SEO. The value goes to whoever aggregates, recommends, and earns trust, not whoever has the biggest catalog.
Today Rachel and I are announcing we closed Construct I with $140m to focus on seed and series A investments in companies transforming the most foundational industries of our economy. More here. https://t.co/bQMvNwEa9N