Worthy Terrible ideas

I’m convinced: All ideas are terrible at birth.

Every idea starts messy, naive, incomplete. You reason with it. You poke holes. You find even one reason to move it forward. You launch fast. Gather data. Sculpt it slowly.

Conviction is earned, not declared.

I secretly want smart people to destroy my ideas. Because when an idea survives smart minds — and still leaves a sparkle in the room — you know it’s worth chasing.

Sometimes tech isn’t ready. Sometimes the world isn’t ready. That’s okay. Good ideas are often just early ideas.

Problem-Solution fit. Product-Market fit. Business-Model fit. These are just evolutionary stages of a once-terrible idea.

When you know all ideas are terrible ideas, you let go faster. You iterate better. You build with resilience, not ego.

That’s how #pifo happened. That’s how Xperian happened. That’s how 100 other experiments taught me to stay patient and playful.

Today with AI, I can generate 100,000 terrible ideas a week. And maybe — find a few diamonds hiding in the dust.

Right now, I’m working on a ridiculous idea: “Learning Fast and Slow with AI.” It’s bad. It’s messy. But there’s a sparkle. We’ll see where it goes.

What terrible idea are you working on?