Greg Isenberg says execution is cheap and ideas are everything. He's right. But he stops one step too early.
Where do ideas come from?
Not from your screen. Not from reading blog posts or watching demos or prompting Claude to brainstorm for you. Ideas, the real ones, the kind that become products people actually pay for, come from watching someone struggle with something in real life and thinking: this shouldn't be this hard.
AI just changed what execution means. A year ago, execution meant writing code. It meant hiring engineers, managing sprints, shipping slowly. Today, Cursor ships the feature. Claude writes the integration. A solo founder with the right tools moves faster than a team of five did two years ago.
Isenberg's point lands: if anyone can build, then building is no longer your edge.
But here's what I'd add: the founders who win aren't the ones who found better ideas sitting at their desks. They're the ones who went outside. Talked to the person with the problem. Watched them fumble through a broken workflow. Heard the thing they'd never type into a search bar.
That's where the idea lives.
I'm building in a vertical where my users don't tweet, don't read TechCrunch, and have never heard of vibe coding. The only way I learn what actually matters to them is to get off the computer and go talk to them.
This used to feel like a distraction from building. Now it is the building.
When execution costs approach zero, the scarce input becomes real-world signal. Not more brainstorming. Not better prompts. Actual conversations with actual people who have the actual problem.
Ideas are everything. But ideas are downstream of getting outside.