I started thinking like an investor in all my companies

Something changed in how I think about companies, and it took me a while to name it.

Over the last year I automated almost every workflow in my companies with AI. Once that happened, the old founder playbook stopped making sense to me.

The old playbook: pick one problem, focus on it for years, ignore everything else. Focus was the religion. And the best returns still went to investors, because investors were the ones collecting knowledge across many companies and betting on trends instead of single ideas.

Today I don’t need investment. I can build almost anything I can imagine, and agents do the operating.

So I changed my approach. When I start something now, I don’t write a product spec. I write an investment thesis. Then I place multiple bets inside it. The bets share one name, so if the thesis works, I’ve been building a brand the whole time. Flags is one thesis with several apps inside it.

I also invest in stages, like rounds. Every bet gets a small first round. If it shows signal, it gets a second round. I don’t double down until the numbers ask for it.

This is not advice. It’s just what happens when building becomes cheap: the founder job starts to look like the investor job.