The moats are drying up.
One is still wet.

At 19 I built my first software company. It made $8,000 in sales before I moved on. At the time I believed coding was one of the greatest investments a person could make. Most people couldn't do it, and syntax was the least of it. Great code was the fingerprint of a great thinker. And if you built something, a moat protected you: the time, the failures, the thinking required kept anyone with the same idea from copying you overnight.

That moat is gone. Anyone can walk across it now.

Around September 2025 I started building another software company. I had 200 million views of experience behind me and was confident I could market whatever I built. But the failures felt different this time. A few years ago failure excited me. It meant I was learning. This time it came with dread, because something kept whispering: this will be buildable in one prompt in a few months.

It felt like the future version of me would look back and see wasted time.

The value of coding is declining. The moat is dry.

The only moats left are the things a machine can't produce: emotion. Uniquely human experience. The ability to make someone feel something real.

A model can hold a shape. It cannot feel. Humans can feel, but most can't articulate what's happening beneath the surface — why certain moments stop you cold, why some stories change you and others disappear the second they end, why the same song hits differently depending on the day. The people who can feel deeply and think precisely about why — those are the last ones with a real moat.

For the past year I've been sitting at that intersection. Building frameworks around attention and emotion, and why certain content makes people feel alive while most of it disappears. I've looked at storytelling through the lens of neuroscience, mapped viewer attention the way you'd map a probability distribution, and tried to understand every emotion not as something vague and poetic but as something with variables you can reason about.

These ideas are still theoretical. I'm testing them, finding holes, learning where I'm wrong. But in the few weeks since I started applying them deliberately, those videos have gathered over 8 million views. I don't yet know which variables are doing the work.

That's what makes it worth figuring out.