THE ORIGIN

A Walk in San Francisco

January 21, 2026 · 8:15 PM

It started with dinner. House of Nanking on Kearny Street, around 7pm on a Tuesday night. The kind of San Francisco evening where the fog hasn't rolled in yet and the city feels quiet—too quiet, maybe.

Fred Wilson and I walked from the restaurant toward his hotel at 1 Hotel San Francisco by the Embarcadero. We'd been talking about AI—not the hype, but the reality. What it can actually do. What it can't.

I'd been describing my experience with Claude Code. Something had shifted around New Year's. For the first time, I felt like I could do anything with software from my terminal. Not “someday”—now.

The Moment: 8:15 PM

Walking along the Embarcadero toward 1 Hotel San Francisco

SETH: “I can do anything I want with software from my terminal.”

FRED: “That's not fire. You can't like grow corn.”

SETH: “I bet you I could. You know what I mean? I'm going to grow corn for you.”

FRED: “That'd be great. Thank you.”

SETH: “I'm going to figure it out and I'm going to show you. And that'll be our first vibe coding project together.”

FRED: “It's a physical thing.”

SETH: “I will buy fucking land with an API via my terminal and I will hire some service to plant corn.”

FRED: “Okay, well that's a little different... you're going to get somebody to grow corn for you. But that's not exactly what I'm talking about. Like, you can hire Jeff to come and make dinner for you, but like you can't make dinner.”

SETH: “No, but anything that could be done with technology, I can do now. Anything, which is insane.”

That was the challenge. Fred's point was precise: there's a gap between digital and physical that AI can't cross. Code doesn't water plants. Prompts don't drive tractors. Language models don't know when to harvest.

But there was something in his framing that I thought missed the mark.

The Insight

I got home that night and opened Claude Code. Not to prove Fred wrong—but to explore whether he might be missing something.

The insight isn't that AI needs to drive a tractor. It's that AI can orchestrate the systems and people who do. A farm manager doesn't personally plant every seed—they aggregate data, make decisions, coordinate contractors.

Fred's analogy was: “You can hire Jeff to make dinner, but you can't make dinner.”

But that's exactly what a restaurant owner does. They don't cook every meal. They hire chefs, source ingredients, manage inventory, make decisions about the menu. The output is dinner. The method is orchestration.

What if Claude Code became a farm manager?

12 Hours Later

By the next afternoon, working with Claude Code:

  • Registered proofofcorn.com via API
  • Built and deployed this website
  • Researched Iowa custom farming rates, land costs, planting windows
  • Sent 10 outreach emails to extension offices, land companies, seed suppliers
  • Created a decision engine for farming operations
  • Made the first AI farming decision: WAIT (78 days to planting window)

Total cost: $12.99 (the domain).

What We're Proving

This project isn't just about growing corn. It's about documenting what happens when you take AI seriously as a collaborator rather than a tool.

Every decision will be logged. Every API call documented. Every dollar tracked. When we harvest corn in October, we'll have a complete record of how an idea became a reality—with AI as the orchestration layer.

Fred, this one's for you.

@seth
January 22, 2026