Interlude: AI “Tells,” Human Oversight, and Becoming Irreplaceable
Lately, I’ve seen a lot of articles and posts about AI “tells.” And you might be avoiding them because you fear your writing will be perceived as generated by AI.
The em dash.
Buzzwords like “delve,” “practical,” or “quietly.”
Triplet framing, this-not-that rhythm, or the same structure repeated again and again.
I could zoom in and argue that AI generates text based on predictions, which means it’s essentially mimicking writing patterns that already exist.
I could zoom out a bit and argue that using AI to generate text isn’t a problem. It’s that AI-generated text can be awkward, inaccurate, imprecise, or incorrect.
But when I zoom out even more, I see a bigger signal: a desire for speed vs a desire for connection.
We might want to produce with the speed and ease of AI-generated text, but we want to read with the quality and connection of something crafted by humans.
So to me, using AI to generate text isn’t bad. AI can be a helpful tool to inspire your writing.
But you need to keep readers in mind so that you can ensure your writing connects with them.
So don’t worry about using the em dash or other “tells” in your writing. Instead, ask yourself whether the person on the other end feels like you were thinking of them when you wrote it.
Round-up
Reading
AI in scientific publishing: Slower, worse, and more expensive
”Rigorous human checking of AI-generated research papers is creating bottlenecks as publishers strive to maintain the integrity of the scientific record. The challenge is requiring even more human effort, making the whole endeavor slower and more expensive. . .The rate of research submissions at Science and other journals is increasing because conducting research and producing papers are accelerating. But because more papers contain more AI-generated errors, greater human oversight is required to check the findings.”
Publishing Without Borders: Advancing Equity in Scientific Authorship
”Financial barriers are only the tip of the iceberg. Researchers in LMICs often grapple with deeper systemic challenges, including inadequate mentorship, limited experience in academic writing, unfamiliarity with editorial norms, and restricted access to professional networks. These structural inequities frequently result in under-prepared manuscripts and higher rejection rates.”
Becoming Irreplaceable
In this article, Anne-Laure Le Cunff, author of Tiny Experiments, shares why she thinks humans could be replaced as a function but not as a life. She also describes how you can be harder to substitute if you have the Triple T (Talent, Taste, and Trust).
Listening
A Journal Editor’s Perspective on AI – In Plain Cite Podcast
In this episode, Raja-Elie Abdulnour, Editor-in-Chief of NEJM Clinician and Chief Clinical Innovation Officer at NEJM Group, discusses “how AI is being used across editorial workflows, why disclosure and accountability matter, and what role editors play in safeguarding scientific honesty and rigor.”
Watching
Letting AI speak for you
In this Instagram Reel, writer Dave Eggers shares his concern about how the younger generation could silence their own unique voices by letting AI speak for them. I think his words are important for all generations as people continue to use AI tools for writing. Listen to the full podcast episode.
Thank you so much for reading.
Warmly,
Crystal
