Yesterday I checked my 100% self-written text with an AI detector.

= According to the detector, 79% of it was AI-created.

Then I added a similar text, 100% AI-created.

= The detector said it’s 54% AI-created.

Transparency is totally fine. But I feel uncomfortable about the idea that I would have to destroy my voice even further to not look like an AI.

- A former mistress of em dashes & 3x structures

———

This post is not whining about AI. I use AI in almost every article, either for grammar improvements or partial or full translations. Yes, sometimes I talk to the mic and ask help to make some sense of it. Yes, sometimes I explain an idea and ask to phrase it.

Nope, not complaining about AI. I’m complaining about data that is incorrect.