There is a new internet habit that is beginning to drive me insane. You disagree with someone in the comments. You write a few sentences explaining why. And five minutes later you receive 1,400 words copied directly from ChatGPT.
Sometimes there’s even an introduction:
“Interesting. Here’s what my AI agent thinks about your comment.”
Mate. I don’t care what your AI agent thinks. I have one too.
If I wanted an encyclopedic explanation of the subject, I could ask it myself. Probably with a better prompt.
I want to know what you think.
Increasingly, though, people seem happy to become MCP interfaces for an LLM with three available tools:
Ctrl+C
Ctrl+V
Send
This is not the AI-enabled future I was hoping for.
Use AI. Just Actually Use It.
There is nothing wrong with using AI to help formulate an argument. I do it constantly. But there is a difference between using AI to improve your thinking and outsourcing the entire act of thinking and forwarding me the receipt.
Let it draft something. Then read it. Delete 80%. Rewrite what’s left in your own words.
If one paragraph genuinely answers the question, send that paragraph instead of the other twelve.
If the model found a useful fact, link to the original source. “Claude says…” isn’t a source.
If you want to quote something particularly useful from the model, fine. But tell me why you agree with it.
And please stop using the model as an authority:
“I asked GPT and it agrees with me.”
Congratulations. Your opponent can ask GPT the same question with slightly different framing and receive an equally confident explanation of why they are right.
LLMs are very sophisticated mirrors. They are not Supreme Court justices.
Ask It to Attack You Instead
There is a much better use of AI during an argument.
Don’t ask: “Write a response proving I’m right.”
Ask: “Find the holes in my argument.”
Give it your position. Give it your opponent’s position.
Ask which assumptions are unsupported, where you’re oversimplifying, what evidence would falsify your claim and which part of the other person’s argument is actually stronger.
Now the machine is doing something useful. It is increasing the quality of your thinking instead of replacing it.
One Simple Test
Before posting an AI-assisted reply, ask yourself: Could I defend every sentence here if someone asked, “Why do you think that?”
If the answer is yes, send it. If at some point you’d have to reopen ChatGPT to discover why “you” supposedly believe paragraph six, perhaps don’t. And if you genuinely don’t know enough about the subject?
“I don’t know.”
“I need to read more about this.”
“I don’t have anything useful to add.”
All perfectly respectable answers. Certainly more respectable than sending someone 1,400 words of machine-generated confidence and pretending you’ve participated in a conversation.
AI should give us more leverage to think.
It shouldn’t turn humans into the copy-paste layer of the stack.
I completely agree. We need to maintain intellectual honesty.👀
I wonder if some people comments only to please the algorithm :)