The agentic gold rush is here. I'm selling shovels.

Everyone's down at the river panning for the AI app that prints money. I'd rather sell the equipment. Here's the honest math on why.

Ryan Doubrava

— 3 min read
Share

The rush is on

There's gold in the river. You've seen the clips: the $80k-a-month agent, the app that prints money while its maker sleeps, the thread that ends with "and I built the whole thing in a weekend." Gold fever. It's everywhere right now, and here's the part nobody wants to say out loud: it isn't entirely wrong. There really is gold in that riverbed. People really are pulling it out.

So everyone grabbed a pan and waded in. Myself included. I published the receipts from my own nights and weekends last month. The question I keep getting is some version of: okay, so which claim are you staking?

Here's my answer. I'm not staking one.

Every gold rush has the same winner

Go back to 1849. The man who started the California gold rush was Sam Brannan. He ran through San Francisco waving a vial of dust, hollering about the American River. Then he went quiet about the gold and got loud about something else: he'd already cornered every pan, shovel, and pickaxe he could lay hands on. Sold them at a markup that'd make your eyes water. Levi Strauss showed up not long after and sold the denim.

The miners? Most of them went home broke. The ones who got rich and stayed rich were the ones selling the equipment.

It never gets old, watching the same pattern run. The reliable money in a gold rush has never been the gold. It's the gear.

So what's a shovel, exactly

Here's where the analogy earns its keep, because "sell shovels" is easy to say and easy to get wrong.

The shovel is not the app that prints money. That's a gold pan, and like every gold pan, it works right up until the easy gold in front of it runs out. The shovel is the thing underneath it: the production-grade rig that turns "I got it to work once" into "it works every time, on somebody else's problem, without me standing over its shoulder." Specs in, finished product out, with the boring rigor (the testing, the traceability, the guardrails) that makes the output something you'd actually put your name on.

Short version: vibe coding is the EZ-Bake Oven. Fun, accessible, genuinely good for everyone, and it makes a mini cake. Agentic engineering is the industrial oven. (That distinction has earned its own post, the next one in this series.) The shovel I want to sell is the oven. Not the cake.

The honest read on the math

Now the part I actually care about, because it's a math argument, not a mood.

The one selling equipment gets paid whether or not any given miner hits gold. The one panning only gets paid if they hit it, and most pans come up with sand. So one of those is a bet on a rare event, and the other is a bet on the rush itself. The rush, I'm confident in. Any single claim paying off is a coin flip wearing the costume of a plan.

I'm not above the fever myself. I keep a pan in my own bag: a couple of plays I run for giggles, a horse in the race, skin in the game so I'm not just narrating from the bank. But that's the side bet. It was never the bet.

Which is why the pitch changed

The last couple of things I wrote were, more or less, "here's what my agents shipped." The natural next sentence is "want your team's agents to do that too?" And for a long while, the honest end of that sentence was hire me.

It's slowly turning into hire them. The equipment, and how to run it. Not me, down at your stretch of river, panning on your behalf.

Which means saying the quiet part out loud: I can build the shovels. I've always been able to build the thing. Building is the comfortable part. Saying it plainly, standing here telling you this is what I'm doing and why, that's the uncomfortable part. So that's the part I'm practicing. In public. One post at a time.

The rush is real. I just know which seat at the table I want.

If you know someone betting their nights and weekends on striking it rich with AI, send them this. Not to talk them out of it. Just so they've seen the other seat at the table.

More from this series:

Read more