Article
All you wanted was a little help with your work. Now you're ten tutorials deep, building a second brain, and wondering whether you're heading in the right direction.
Claude is doing things for you. Scripts are running. The organization looks sort of all right. But the whole thing keeps getting more complicated.
I've been a developer for 20 years, and I made the same mistake.
The work around the work
When AI first came along, I went down the rabbit hole. I built my own tools for running AI agents, then kept adding to them. Getting started was so easy that it became intoxicating.
There was always something else I could build. It felt productive.
The problem was where I was putting that effort. I was getting further into the tools without focusing on the important things in my business. Building a more complicated AI setup had become its own job.
If you're in that position, another tutorial may take you further down the same path. I'd start by asking what your business needs right now.
More sales? More connections? Better delivery?
Pick one.
Write down one job before choosing the tools
Once you've chosen the business need, work through one job that serves it.
- Gather the sources needed to do that job.
- Write down the instructions and the steps in the workflow.
- Look at each step and decide whether you need it.
- For the steps you keep, decide how the work should happen.
That last decision has several possible answers. You can use ordinary automation. You can delegate the work. You can use AI where it makes sense.
You don't have to make every step an AI step, and you don't have to keep a step just because you've figured out how to automate it.
Before opening the next tutorial, choose the job your business needs done. Gather its sources and write down how the work should happen.