I Didn't Know What a Loop Was. Now I Build Systems.

There is no catch

Everyone thinks they can teach themselves AI.

Here is what it cost me. Over two years on ChatGPT. Over 250 hours in Claude in the six weeks before I recorded this Fargo Walks clip. That is my own count, and it is the honest price of the sentence that follows.

I am now building systems that automate the things you hate. I am coding in Python to close loops.

What the hours bought

I did not know what a loop was. A loop is the instruction that tells a system to keep going until the job is done. It is the difference between a tool you have to push and a system that runs without you.

I did not know what slop was. Slop is the generic output AI produces when nobody gave it anything real to work with. It reads like a machine wrote it, because one did. Now I know what slop is, and I know how to stop it. I wrote about the fix in How to Make ChatGPT Emails Sound Like You.

Neither of those came from a course. They came from hours, most of them on days I did not feel like it. The Python part of that story is in From 9th Grade Algebra to Coding Python With AI.

I do not recommend this path

I mean that. I do not recommend you go down the path I went down. Go down this path instead, with me. It takes patience and time either way. The difference is whether you pay the 250 hours yourself or walk with someone who already did.

That is not a pitch for shortcuts. There are none. It is a pitch for not wandering.

What the path looks like with a guide

1. Start with the thing you hate

Not the flashy use case. The recurring task that steals your evening. That is the first loop to close.

2. Build from your own records

Your email, your calendar, your transcripts. A system that reads your real work beats any template.

3. Close one loop at a time

One task, running without you, tested on real inputs. Then the next one. Systems compound. Dabbling does not.

4. Learn to see slop before it ships

Give AI your voice and your facts. Review what comes back. The habit takes weeks to build and pays every day after.

5. Budget patience and time

The first system will frustrate you. Keep going. That is the part most people quit at, and it is the part that separates the owners who have AI systems from the ones who have a subscription.

Why I exist now

Aside from the undying love I have for my wife, my children, my dad, and my close friends, this is why I now exist. Building the systems, and walking the path with the people who decide to take it.

What this does not mean

It does not mean you cannot learn this. It means the self-taught route costs more than most owners can spend, and the cost is hidden until you are 100 hours in.

The decision

Stop assuming you will teach yourself AI on the side. Pick the first loop to close, and walk the path with someone who has already paid for the map.

Sources

• Jeff Fargo, Fargo Walks: I Didn't Know What a Loop Was. Now I Build Systems. (recorded 2026)
• Jeff Fargo, Fargo Factor Insight: From 9th Grade Algebra to Coding Python With AI (August 2026)
• Jeff Fargo, Fargo Factor Insight: How to Make ChatGPT Emails Sound Like You (August 2026)

Transcript

Verbatim from the Fargo Walks clip, as recorded.

There is no catch. Everyone thinks they can teach themselves AI. Over two years on Chat GPT, over 250 hours in Claude over the past 6 weeks, I'm now building out systems that automate things that you hate. I'm coding in Python to close loops. I didn't even know what a loop was or stopping slop. I didn't know what slop was. Now I know what slop is. I do not recommend you go down this path. Go down this path instead with me. It takes patience and time. Aside from the undying love I have for my wife, my children, my dad, and my close friends, this is literally why I now exist. Go forth.

Jeff Fargo is the founder of Fargo Factor and host of Fargo Talks. Fargo Factor advises established businesses on AI Search Authority, AI Decision Systems, and selective AI-enabled implementation.