Interlude: "Good Enough" Drafts, AI Guidelines, and DOI-First Referencing
I've talked to many researchers who submit "good enough" drafts of their manuscripts. By that point, they're burned out from months—or even years—of collecting data, analyzing results, and writing the manuscript. And they just want to get the draft off their desk.
But when readers have to trudge through dense or confusing prose, they lose interest—and with it, they lose meaning.
The writing might meet the standard for scientific rigor, but it doesn't meet the standard for reader engagement.
Yet engagement is the real threshold for impact.
An engaged reader is more likely to:
Grasp the significance of your findings.
Retain your message.
Apply your insights in research or practice.
Cite, share, and build on your work.
None of that happens unless your reader connects with your message. And connection starts with writing that feels clear, purposeful, and alive—that respects the reader's intellect and attention.
A “good enough” draft might get the job done, but your work deserves more than that.
Great science deserves great writing.
Now onto this week's round-up...
Round-up
From My Desk
5 Essential Guidelines for Using AI in Scientific Writing
AI tools like large language models can be really useful. But they also have risks and limitations—and you are the one who is accountable for how they’re used. In this second video of my series on how to use AI responsibly for scientific and medical writing, I share some best practices so you can confidently and responsibly use these tools for your own work.
Reading
Towards a DOI-First Referencing Model: Opportunities, Limitations and Implications for Scholarly Publishing
“This article proposes a DOI-first referencing model as a simplified and identifier-centered approach to scholarly citation. It discusses the opportunities of DOI-based referencing for improving efficiency, interoperability, and integration with digital research infrastructure while acknowledging disciplinary diversity and identifier limitations. The paper argues that prioritizing persistent identifiers can modernize citation practices in an increasingly digital and AI-assisted scholarly publishing environment.”
Correction to a retraction highlights tortured phrases have been around longer than LLMs
“While large language models are taking the blame for hallucinations, punctuation and all manner of language choices these days, turns of phrase were being tortured well before the arrival of LLMs. . . ‘Tortured phrases are produced by ‘text spinners’: online websites running a basic algorithm that replaces some words with synonyms, using a thesaurus."
Training
From AI Curious to Confident – May 4, 2026
Most AI training for medical writers offers generic advice that doesn’t reflect the reality of working with clinical data, peer-reviewed literature, and regulatory standards. In this 2-week bootcamp, Núria Negrão will give you practical, real-world AI workflows with hands-on exercises, real use cases, and live Q&A in just 35 minutes a day. Early bird rate ends April 15.
Quote
"We tend to think of writing as the act of assembling words, but it's a deeper experience than this. Words may be symbols, but they are not abstractions; they are the method by which we express our ideas." – John Warner
Thank you so much for reading.
Warmly,
Crystal
