Not long ago, a client asked me to rewrite an article so it wouldn’t “sound like AI.” Fair enough — except the piece in question was something I had written six or seven years ago. Long before ChatGPT, Claude, Jasper, or any of the generative models existed.

In fact, I’ve tested curricula I personally wrote for education over the years — many, many thousands of pages, blood and sweat in every line — and those too often come back flagged as “AI-generated.”

It’s not just technical work, either. I’ve run some of my own fiction drafts — character sketches, early chapters, even bits of dialogue for novels — through these detectors, and they flag high percentages as AI too. Imagine spending late nights writing a scene where two characters fall in love, only to be told later that a machine must have written it. If that’s not insulting, I don’t know what is.

Why Content Detection Is a Mess

Today, dozens of companies claim they can detect AI-generated text:

• GPTZero

• Originality.AI

• Copyleaks

• Turnitin

• Undetectable.AI

• Sapling AI Detector

Run the same document through five of them, and you’ll often get five different answers. One says 90% AI, another says 20%, another says “mixed.”

The truth is, most detectors don’t actually “see” AI — they look for patterns in sentence structure, word predictability, and rhythm. And guess what? Both technical writing and creative writing can easily fall into those patterns.

• Technical writing needs to be precise.

• Novel writing leans on rhythm, pacing, and motifs.

A love story using short, emotional sentences? Predictable.
A fantasy epic that repeats lore names? Predictable.

To a detector, predictable equals AI.

My Own Use of AI (The Honest Bit)

Here’s where I stand: I do use AI — mostly to correct spelling and grammar. Think of it as the modern spellcheck. But I always try to tell it: keep my words intact unless they’re embarrassing.

I don’t want it rewriting my sentences into something bland and soulless. I want it to polish, not ghostwrite. That goes for both my technical work and my fiction drafts. The little grammar fixes help, but the voice — the humor, the pacing, the personality — has to stay mine. Otherwise, it’s not writing.

SEO: The Real Battlefront

For years, Google has insisted that it ranks content based on quality, expertise, and helpfulness — not whether it’s written by a human or a bot. Their official Search guidance even says:

“Appropriate use of AI or automation is not against our guidelines.”

Yet here’s the paradox: many SEO firms and marketing departments now insist that “AI-free” content ranks higher. I’ve had clients flat out request rewrites to “beat AI detectors” before they publish.

And whether or not Google is actually penalizing AI text, the perception has become reality: if it’s flagged as AI, many businesses won’t risk it.

The irony? I’ve run some of Google’s own help pages through detectors — and they flag as AI too.

The Problem With “Humanizing” Tools

A cottage industry has sprung up around “AI humanizers.” Services like Quillbot, Paraphraser.io, and others promise to rewrite your content so it slips past detection.

But here’s the catch:

• In technical writing, these tools often break accuracy. The math gets fuzzy, compliance terms get butchered, and key steps vanish.

• In creative writing, they sand off your style. A sharp, witty line of dialogue turns flat. The eccentric rhythm of a fantasy passage becomes bland.

They make your writing “safer,” but they also strip away its humanity.

Creative vs. Technical: Why It Matters

If you’re writing a funny blog post about AI crypto coins or a personal story about Adobe PDFs, AI tools (or humanizers) can shuffle synonyms without much harm.

But if you’re writing:

• Medical curriculum

• Government forms

• Educational syllabi

• Legal compliance documents

• Novels, short stories, or poetry

…the stakes are higher. Accuracy and originality matter. A wrong dosage note could be dangerous; a flattened story could kill its spark.

And unfortunately, these are the kinds of writing most likely to be flagged because of their repetitive structure, rhythm, or precision.

So What Should Writers Do?

There’s no perfect fix. But a few strategies help:

• Mix sentence length for a natural rhythm.

• Add personal anecdotes that no bot detector can fake.

• Use contextual links (Adobe, Wired, Pew Research, etc.) for SEO trust.

• Focus on value — if your writing helps people, Google doesn’t care who (or what) wrote it.

• Educate clients — often the battle isn’t with Google, but with how people think Google works.

The Bigger Picture

At the end of the day, this obsession with “AI-free content” is both funny and frustrating. Writers are now being asked to make their human writing less human-looking just to please a machine. And it’s worth remembering: the very detectors we’re trying to please are themselves powered by AI.

That includes fiction. If detectors keep flagging novels and stories as “machine-made,” how will authors prove originality? Will publishers demand AI-detection reports alongside manuscripts? Absurd, but possible.

The real danger: if companies overreact, they’ll waste time rewriting good content, chasing a moving standard that may not even affect SEO. Meanwhile, the winners will be the ones who focus on clarity, authority, and originality — whether their text scores 10% or 90% “AI.”

So maybe the real skill for today’s writer isn’t just writing well. It’s writing well enough to survive the machines judging the machines.