It Says All the Right Things. It Means None of Them. How other’s read your AI email

Jun 13, 2026

You’ve felt it. The email that’s flawless and hollow. The condolence note that hits every right word and lands on none of them. The song that’s pleasant, competent, and leaves you completely cold.

Perfect. And somehow cold.

There’s a name for this feeling, and it’s over fifty years old.

In 1970, a Japanese roboticist named Masahiro Mori noticed something strange. As robots became more human-like, people warmed to them — right up until they got almost human. Then affinity collapsed into unease. A face that’s nearly right but slightly dead in the eyes. A hand that looks real but feels wrong. He called it the uncanny valley.

For fifty years we assumed this was about bodies. Skin, faces, movement.

It isn’t.

We’re now feeling the exact same drop in places with no body at all — in language, in music, in the written word. The most intimate channels we have. The eeriness used to be triggered by something almost looking human. Now it’s triggered by something almost sounding human. Almost feeling human.

Almost human is the worst place to be.

And here’s what I find fascinating: that discomfort isn’t a flaw in your perception. It’s a signal. We are exquisitely tuned to tell the difference between competence and presence — between something that performs humanity and something that actually shows up.

Which is the real question for anyone leading, writing, or creating right now:

When the polished version is free and instant, polish isn’t the scarce thing anymore.

The scarce thing is intention — the deliberate work of shaping something so it carries your essence, your voice, your humanness.

Not because it slipped in. Because you put it there.

That’s the part the valley can’t fake.

#HumanModeAlways

Update: Here is an exchange I had with one reader that brought to mind additional insights and resources related to these ideas:

 

Enterprise Architect | Data Scientist | Communication Professional |

 

I find this very interesting. And when reading your post, I do automatically scan for the signs of AI writing: The em dash, antithesis/reframings and epistemic hedging. That way of reading texts, call it a bias, has slippes into my reading of almost all LinkedIn posts. I often find myself stop reading if those signs are there. I continued with yours because I assume you already is aware of this, and that it might be the point. Maybe the use of the distancing language found in AI texts is contributing to the feeling of distance? Curious to know your thoughts from a behaviour psychological perspective.

Jeffrey Sanchez-Burks

 

Margrethe Blomsø Your instinct about distance language and interpersonal distance is consistent with work I’ve recently read by the Belgium scientist Francis Heylighen and linguist @Jean-Marc Dewaele. They write,”…deep formality
is defined as avoidance of ambiguity by minimizing the context-dependence and fuzziness of expressions. This is achieved by explicit and precise description of the elements of the context needed to disambiguate the expression. A formal style is
characterized by detachment, accuracy, rigidity and heaviness…It is proposed that formality becomes
larger when the distance in space, time or background between the interlocutors increases…” Here is the link to the paper: https://pespmc1.vub.ac.be/Papers/Formality.pdf Thanks for priming me to make the connection to their work.

 

…and to be fair, research by Ethan Kross and his colleagues do show situations where such distancing language can be helpful, specifically for self-talk when mentally working through challenges. A review of that work can be found here: https://sites.lsa.umich.edu/emotion-selfcontrol-psych/wp-content/uploads/sites/1322/2017/09/Self-distancing-theory-research-future.pdf