Apparently, My AI Employees Have Expense Accounts#
Captain’s Log. Stardate 2026.302.
I’ve discovered something about my AI workforce.
THE ROBOTS CHARGE FOR THINKING.
🤣
Okay, technically, I already knew this.
Sort of.
I’d heard people talking about tokens on YouTube.
Input tokens. Output tokens. Model usage. All that fascinating technical stuff.
And I’d nodded along thinking, Yep. That sounds important.
Did I understand what it actually meant?
Apparently not.
😂
Then I started building Recruiter Bot in n8n.
And one day, I noticed my monthly AI usage allowance going down.
Hang on.
WAIT, I’M PAYING FOR EVERY TIME THIS THING THINKS?!
🤣
Naturally, I investigated.
How much had my little AI workforce cost me?
Approximately one cent.
One.
Cent.
FINANCIAL CRISIS AVERTED.
😂
But here’s the thing.
I’d already made a decision about how much I was prepared to spend on this project.
Because Recruiter Bot isn’t really about building a recruitment tool.
Well, it is.
But that’s not why I started.
I wanted to learn about AI.
Not just how to write prompts or have conversations with ChatGPT.
I wanted to understand how AI systems work, how they connect to other tools, what they can actually do, and what happens when you try to build something useful.
I could have enrolled in a course.
Instead, I decided to spend money on the tools themselves and learn by building.
ChatGPT subscription? Yep.
n8n subscription? Yep.
Claude subscription? Apparently, we’re employing him now too.
A couple of domain names?
Why not?
🤣
And for me, right now, that trade-off makes sense.
I’m not claiming this is the cheapest way to learn AI.
I’m certainly not claiming it’s the most organised.
HAVE YOU SEEN MY PROJECT SCOPE?
😂
But I’m learning.
And the funny thing is, my AI employees are actually producing things.
I have a website.
I have spreadsheets containing data I didn’t manually enter.
I have a growing GitHub library.
I think I’m up to five repositories now.
Five!
At some point, I stopped experimenting with AI and started accidentally building an IT department.
🤣
And now that IT department has expenses.
Which, to be fair, is another useful thing to learn.
Because understanding what AI can do is only part of building a system.
You also need to understand what it takes to run one.
The subscriptions, the usage, the infrastructure, the trade-offs.
And whether the outcome is actually worth the cost.
Right now, I’m happy to spend money within reason because I’m treating this as an investment in learning.
And apparently, that includes discovering how much my robots charge for thinking.
😂
So there I was.
Investigating the financial activities of my AI workforce.
Not because I was worried about the money.
But because I wanted to know what the little bugger had been thinking about.
TOTAL EXPENDITURE: APPROXIMATELY ONE CENT. EDUCATIONAL VALUE: STILL BEING CALCULATED.
🤣