Dzone iconDzoneSep 8, 2026 ~1 min source read

Optimize an AI Agent to Sound Human, Judged by an AI Detector

Some platforms down-rank content that reads as, so teams publishing at scale have a real stake in prose that clears a detector, even when a human wrote it. The "I hope this email finds you well" opener, the three polite paragraphs answering a one-line question.

Optimize an AI Agent to Sound Human, Judged by an AI Detector

Share this story

Send the public story page.

Useful takeaways from this story.

The "I hope this email finds you well" opener, the three polite paragraphs answering a one-line question.

I wanted a reply-drafting agent that didn't do that, and "don't sound like an AI" turned out to be hard to put in a prompt.

The rest is judgment, and a single prompt that holds across a friendly dinner invite and a recruiter cold email took more iterations than I'd guessed.

Building the complete brief

The page is ready to read now. The fuller skim-friendly version will appear here automatically.

The useful part

The "I hope this email finds you well" opener, the three polite paragraphs answering a one-line question. I wanted a reply-drafting agent that didn't do that, and "don't sound like an AI" turned out to be hard to put in a prompt. The rest is judgment, and a single prompt that holds across a friendly dinner invite and a recruiter cold email took more iterations than I'd guessed.

How it works

  • Some platforms down-rank content that reads as, so teams publishing at scale have a real stake in prose that clears a detector, even when a human wrote it.

Keep reading in the app

Open the app view to save this story, compare related coverage, and continue from the same source.

Open in app