AI

Loops, Not Divides

February 25, 2026 2 min read
Loops, Not Divides

We inherited a clean division from twentieth-century management: thinkers in one room, builders in another, a PowerPoint deck (vF_real_actual_definite_FINAL3) passed between them like a baton. It was never a particularly good arrangement. Now, with the cost of building collapsing toward zero, it just doesn't make sense.

Every impressive person I've met shares a similar trait: they refuse to choose between thinking and building. Ask them if they are strategy or execution people, and they'll give you a quizzical look. It's a false choice. They are not one or the other, they're both.


When you look at the slate of things being built (and talked about) on social media today, you'll find the same smattering of things: a personal CRM, a travel itinerary builder, a recipe guide.

We finally received the magic wand to make anything, and we built more derivatives and declared that SaaS was dead.

Instead of exploring what is possible, we're creating (cheaper, better, faster, more personalized) copies of known objects. What a crisis of imagination.


I think of emerging technologies as a creative medium. The only way to fully understand what is possible or not is by actually building and playing with it.

Is it brittle or pliable? Is it hard or is it soft? Is it static or evolving?

So, what do you do?

Some people will say, build more and have more shots on goal. Others will say, pull back. Focus on what you should build and why. Spend more time thinking, exploring, researching, getting into the texture.

Again, these are false choices.

The best people will do both. Think deeply about what to build and why, and then quickly learn from what happens when you actually make it tangible. Then loop back.

It's not about picking one side or the other. It's about faster loops.