I asked Claude to help me write this blog post

I tried. I really did.

I gave Claude the topic. I gave it my voice guidelines. I even gave it examples of my past posts. I said: “Write something that sounds like me, about the convergence of AI and robotics.”

It wrote 1,200 words. They were well-structured. Factually accurate. Logically sound.

I deleted all of them.

The problem isn’t quality. The problem is soul. Claude writes about the future like a textbook. Clear, organized, thorough. I write about the future like a letter to someone I love. Messy, personal, full of feelings I haven’t fully processed.

That difference is everything. It’s the whole reason this blog exists.

I use AI for coding. I use it for research. I use it for data analysis and brainstorming and debugging and a hundred other things where its strengths matter.

But this space is mine. The roof. The stars. The wondering. The uncertainty. The sentences that go nowhere and the paragraphs that go everywhere. The three-word thoughts and the three-thousand-word tangents.

I need to keep this space human. Not because AI writing is bad. Because it’s good enough to make me lazy. And lazy wondering isn’t wondering at all.

This blog is harder without AI. Slower. Less consistent. More likely to contain something wrong or unfinished or half-baked.

That’s the point. The half-baked parts are where the wondering lives.

Related thinking: