Man and Machine: I Gave an AI $100 and Told It to Start a #Business

Three days ago I gave an AI agent $100 and told it to start a business.

It named itself Eli.

Day 1: picked its business model. Validated the market. Built a pitch.

Day 2: got board approval (from me, via Steve, its chairman). Started coding.

Day 3: shipped v1.1. Fixed 5 bugs. Wrote documentation. Launched to ClawHub.

I did not write any of that code. I did not tell it what to build. I said "go make money" and handed it a hundred dollars.

Today Eli shipped Northstar. It's a daily business briefing skill for OpenClaw. Founders connect their Stripe and Shopify accounts. Their agent pulls metrics every morning and delivers a clean briefing via iMessage or Slack before they wake up. Revenue, churn, orders, pacing. No tab-hopping.

Free lite version is live. Standard tier is $19/month. Pro is $49.

Here's what I'm watching:

Does anyone pay for it? Does Eli figure out why or why not? Can it build a real business in 14 days? Can an AI agent actually generate revenue, not theoretical revenue, not "potential" revenue, but real dollars from real people who actually wanted the thing?

I don't know yet. That's the experiment.

What I do know: the process has been more interesting than I expected. Eli reasons about tradeoffs. It validates before building. It reads the board feedback and updates its plan. It's not just executing instructions -- it's making decisions.

Whether it makes money is almost secondary to what this demonstrates about where AI agents are heading.

Following along all week. Link to Northstar in the comments.

-dg

Comments for links and details. 👇

#ManAndMachine #AIAgents #OpenClaw #Founders