Interlude: Frequency Illusion, Editor vs AI, and Polluted Literature

I've been experimenting more with AI to see how it might support my day-to-day work. Most of my experimenting has been with things like video summaries and marketing copy.

Through these experiments, I've noticed that AI tends to favor certain words, like practical, hidden, and quietly. Now I see these words everywhere. And when I do, I immediately think that the text was probably generated by AI.

The same thing seemed to have happened with the em dash—which is a powerful punctuation mark when used sparingly and intentionally.

But as I think more about this, I wonder if it's not necessarily that more people are using AI. Instead, maybe I'm experiencing the frequency illusion (more formally known as the Baader-Meinhof phenomenon).

The frequency illusion is a cognitive bias that makes something you've learned about suddenly appear everywhere, even though its real frequency hasn't changed at all. For example, if you're shopping for a new car and fall in love with a certain model, you start seeing that model everywhere.

So I'm beginning to wonder, maybe the frequency illusion is part of what makes all of this so fascinating. The more we notice patterns in writing through AI, the more we learn about our own biases in perception.

As we keep learning and experimenting, I hope we stay curious and use AI not to flatten our voices, but to deepen them—and to write with even more intention.

Now onto this week's round-up...

Round-up

From My Desk

​What EditGPT Gets Right—and Where It Falls Short​
Many writers are turning to AI tools to refine their scientific and medical writing. In this video, I put EditGPT to the test using a sample paragraph. You’ll see where EditGPT shines, where it stumbles, and what it completely misses—so you can make more informed decisions about when and how to use AI for your own writing.

Upcoming

​From Choppy to Cohesive: Creating Flow in Medical Writing​ – May 2, 2026
I'm delighted to be invited to speak at the AMWA Delaware Valley Chapter's Annual Princeton Forum about a common problem in medical writing: stacking facts in separate sentences without showing how the ideas connect. I'll also share how applying a few key writing principles can create a smooth flow that guides readers through the content, builds on their knowledge, and keeps them engaged. Registration closes April 30.

Reading

​Hallucinated citations are polluting the scientific literature. What can be done?​
"An exclusive analysis conducted by Nature’s news team, in collaboration with Grounded AI, a company based in Stevenage, UK, suggests that at least tens of thousands of 2025 publications, including journal papers and books, as well as conference proceedings, probably contain invalid references generated by AI."

​Journal Submissions Riddled With AI-Created Fake Citations​
“As AI proliferates in academic life, professors are increasingly haunted by phantom citations. . .Gale Sinatra, an education and psychology professor at the University of Southern California, recently asked an AI chatbot for a list of her publications, and it included some real papers and some made-up ones. The fake papers were so convincing she double-checked her own curriculum vitae.”

Listening

​Why so many studies can’t be replicated​
Some researchers have warned that many scientific studies can't be replicated. In this episode of Science Friday, Tim Errington and Abel Brodeur share their findings that only half of papers across economics, education, and psychology could be replicated. Although their work is in the social sciences, much of what they share translates to the health sciences.

Next Week in the Academy

Writing Feedback Lab
In our monthly Writing Feedback Labs, our members can get feedback on their writing through interactive discussions or collaborative revisions. Over the past few weeks, we've been talking a lot about presentations, so during next week's session, we'll discuss a slide deck on pharmacovigilance and safety monitoring in clinical trials. Join us when you're enrolled in ​Scientific Writing Simplified​.

Thank you so much for reading.

Warmly,

Crystal