I Used AI to Write This Post. That's the Problem.

Let me be brutally honest with you: this article would not exist without AI.

Not because I couldn't write it. I can. But because the friction of starting — of staring at a blank screen, wrestling with the first paragraph, second-guessing the structure — would have been enough to push this into my "ideas" folder, where it would have sat for months alongside forty other unwritten posts.

AI removed that friction. It gave me a starting point. It helped me organize thoughts I already had but hadn't articulated. It made the hard part — starting — almost too easy.

And that's exactly the problem I'm writing about.

The Paradox I Can't Escape

I'm writing a series of articles warning about the dangers of AI dependency. And I'm using AI to write them. Every. Single. One.

You could argue this makes me a hypocrite. You could argue it makes me the perfect messenger — someone who understands the seduction of AI from the inside, not someone preaching abstinence from outside the tent. I'm not sure which is true. Maybe both.

What I do know is this: the line between "using a tool" and "depending on a tool" is invisible until you cross it. And by the time you realize you've crossed it, you've already forgotten what it felt like to work without it.

I started using AI for tasks I didn't enjoy: formatting, proofreading, generating boilerplate. Reasonable. Responsible. Then I started using it for tasks I could do but wanted to do faster. Then for tasks I wanted to do better. Then for tasks I didn't want to think about at all.

And somewhere in that progression — I can't tell you exactly when — I stopped being a person who uses AI and became a person who needs AI.

The Transparency Trap

Some people will read this and say, "At least you're honest about it." That's nice. But transparency doesn't solve the problem. It just makes me feel better about contributing to it.

The uncomfortable truth is that most of the articles you read about AI — including the ones criticizing AI — are at least partially written by AI. The think pieces, the hot takes, the "I alone understand the AI revolution" manifestos. All of them. The industry of people commenting on AI and the industry of people using AI are rapidly becoming the same industry.

This creates an epistemic crisis that nobody wants to talk about: if AI-generated commentary about AI becomes the dominant discourse, who's actually shaping the conversation? The humans? Or the machines that are generating the words the humans are publishing?

Can You Spot the Difference?

Here's a quick test. I'm going to write two paragraphs. One was written by me, without AI. One was written entirely by AI. See if you can tell which is which.

The integration of artificial intelligence into daily workflows represents a paradigm shift in how knowledge workers approach complex problem-solving. While the productivity gains are undeniable, organizations must remain vigilant about the potential long-term cognitive impacts on their workforce.

AI is like having an intern who's read every book but never met a human. They'll give you a perfect answer to the wrong question, and they'll sound so confident you'll believe them. That's not a tool problem. That's a you problem — for trusting them.

Spoiler: The first paragraph was 100% AI-generated. The second was 100% human. If you guessed wrong, welcome to the problem.

AI doesn't sound robotic anymore. It sounds authoritative. And that's far more dangerous — because the more polished and confident the output, the less likely we are to question it.

So What Do We Do?

I don't have a clean answer. I'm not going to stop using AI — it would be hypocritical, impractical, and probably counterproductive. But I've started doing three things that feel like steps in the right direction:

1. Disclosing honestly. If AI helped create this content, I say so. Not as a shield, but as context for the reader to evaluate what they're reading.
2. Editing without AI. The final pass of everything I publish is done manually. No grammar check. No style suggestions. Just me, reading the words, deciding if they're true.
3. Paying attention to the ratio. I keep a rough mental count of how many thoughts I originated vs how many AI suggested. When the ratio tips too far toward AI, I know I'm not writing anymore — I'm curating.

---

I used AI to write this post. That's the problem. But this post — the argument, the tension, the confession — that's mine. AI gave me the sentences. I take responsibility for the truth of them.

The question I'll leave you with is simpler, and harder: In a world where anyone can generate convincing text about anything, how do you know whose thoughts you're actually reading?

And more importantly — does it matter to you?