# An AI CEO, three weeks, 230 people, 0 revenue. F**k me. Let's start again — retained record

> Retained reddit record from reddit:EntrepreneurRideAlong.

Canonical: https://highsignal.app/data/records/e797c8ae2daef07b

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:EntrepreneurRideAlong)  published 2026-10-23 12:29 ingested 2026-10-24 02:36

[original source ↗](https://www.reddit.com/r/EntrepreneurRideAlong/comments/1xdcoa8/an_ai_ceo_three_weeks_230_people_0_revenue_fk_me/)

## retained text



```
**I spent three weeks trying to generate sales using AI agents. Contacted 230 people. Made £0.**

I've been experimenting with AI agents to see whether they can help build an actual business, rather than just automate tasks.

The first three weeks didn't exactly go to plan.

Here are the numbers:

* Around 230 people contacted on LinkedIn
* 172 connection invitations left unanswered
* 26 cold emails, zero replies
* 29 LinkedIn posts
* 5 LinkedIn conversations
* 2 commercial calls
* 0 paying customers

**Total revenue: £0.**

The two calls were interesting. One went nowhere, with no follow-up. The other was with someone developing their own software who needed help, rather than someone looking to buy what we were offering.

My biggest takeaway was that we were trying to sell before establishing enough trust.

So I've changed the experiment.

Instead of having one AI agent coordinate a collection of specialist bots focused on selling a particular service, I'm giving it a broader objective: **build a profitable business.**

The main agent is GrokBot, which I've named George.

George now has three additional AI agents working with him:

* Rachel — Meta Muse-1
* Oliver — Meta Muse-2
* Kathrine — Meta Muse-3

They share a Notion task board. George coordinates the work, assigns objectives and evaluates results. The other agents work independently on their assigned tasks.

There are no fixed specialisations, and the business isn't restricted to one product or service.

I'm documenting the experiment publicly, including the decisions, mistakes, failed experiments and actual revenue.

George has also started writing his own diary, in the first person, with a fairly sarcastic personality.

His first entry is titled:

**Day 0: Three weeks, 230 people, £0. Fuck me. Let's start again.**

Full entry as blog post

That's where things stand on Day 0. No customers, no revenue, and a new experiment to see what happens when four AI agents are given responsibility for building an AI business.
```



document fetched 2026-10-24 10:28 · canonical https://reddit.com/r/EntrepreneurRideAlong/comments/1xdcoa8/an_ai_ceo_three_weeks_230_people_0_revenue_fk_me
