I Used AI to Explain Why I Like Using AI
I read an article about AI.
I had opinions.
This will shock absolutely fucking nobody.
I explained those opinions to AI in the usual format: half an argument, three detours, some profanity, and a better version of the opening point arriving near the end.
Then I asked it to turn the whole thing into a comment.
It did.
I read it and thought: yes. That's what I meant. That would have taken me fucking forever to arrange properly.
Then I noticed what had happened.
I had used AI to explain why I like using AI.
The demonstration was included in the fucking complaint.
Lucifer pulled up a chair.
“Let me understand this.”
“Dangerous opening.”
“You disagreed with an article about AI.”
“Yes.”
“You told the AI why.”
“Yes.”
“And it helped you explain the disagreement.”
“Beautifully.”
“Was that cheating?”
“At what? Having a fucking opinion?”
He considered this.
The Expensive Middle Bit
I already knew what bothered me.
I didn't need the machine to issue me a position. I needed help turning the position into something another person could follow without first attending a guided tour of my nervous system.
There's work between thinking something and communicating it well.
Pick the opening. Remove the repeated point. Put the example next to the claim it explains. Find the ending. Notice that paragraph four belongs before paragraph two and paragraph three should be taken outside and quietly disposed of.
I can do that work.
I don't always want to spend that much time doing it.
The useful part was the transformation cost dropping. The thought could reach someone else while I still cared enough to send it.
Still My Problem
Of course I have to read the result.
If it changes my meaning, adds a claim I don't believe, or sands the thing into a corporate LinkedIn suppository, it goes back.
Polished is only useful if it's still what I meant.
The opinion remains mine. So does the responsibility for the version I choose to put my name on.
Then I did it again while explaining context and model economics to somebody else.
Apparently I was conducting a repeatability study.
Lucifer looked at the comment, then at the draft of this post.
“And now you're using AI to help explain that you used AI to explain why you like using AI.”
“Correct.”
“Does this end?”
“Ideally before somebody suggests a framework.”
He closed the notebook.
For once, we agreed.