The Pen vs. The Prompt: You will be judged II

Back in 2020, I wrote a piece called You will be judged [1]. The premise was simple: fair or not, people assess your intelligence, competence, and integrity by the quality of your writing. At the time, the solution was traditional and straightforward: practice, edit, read aloud, and develop your craft. It underlined what I was taught as a kid: practice makes perfect. It’s a common idiom encouraging consistent, repeated action to achieve high proficiency. While it highlights that consistent effort is necessary for improvement, it is often noted that practice actually makes permanent: referring to how only correct, deliberate practice leads to true perfection, while practicing mistakes only solidifies errors.

Fast forward 6 years, and the landscape has transformed. Today, anyone with an internet connection can produce polished, grammatically flawless prose in seconds. The tools that might have been considered science fiction in 2020 are now ubiquitous. So, does my 2020 advice still matter? And if a machine can make your writing almost perfect, what exactly are we judging now?

For me, the answer is more nuanced, and more human, than you might think.

The New Double-Edged Sword

Let us start with an uncomfortable truth. If you are a professional in 2026 and refuse to use any AI assistance, you risk being judged as inefficient, or worse, lazy. Why should a reader invest their time in your communication if you could not be bothered to leverage the tools that would have respected their time? Nobody expects to see typos! There is an expectation of competence that now includes the judicious use of technology.

However, this is where the pendulum swings back. We are currently drowning in what critics have aptly termed "AI slop," generic, soulless content churned out by large language models [2]. Audiences are developing a keen, often subconscious, ability to detect when a document lacks a human soul. A 2025 study published in Nature Human Behaviour found that even when identical responses were generated by AI, people perceived human-attributed replies as more supportive, emotionally resonant, and caring [3]. If your audience suspects you are simply turning a crank, they will judge you more harshly than if you had written a flawed but authentic email.

The Frustration of Asymmetric Effort

Let me share a personal irritation (that I suspect many of you share). I will spend an hour crafting a nuanced communication to a colleague, weighing every argument, considering the emotional subtext, and structuring the flow. Ten minutes later, I receive a two-page response. They have barely had time to read and digest my original thoughts, let alone construct a meaningful reply. Did they even read email, or did they simply feed my text into an LLM and ask for a rebuttal?

When you reply to heartfelt, hard-won prose with an AI-generated wall of text, you are communicating a dangerous message: "Your time and thoughts were not worth my personal engagement." I believe that I am supported in this opinion from research reported by Ohio State University. Their 2023 study demonstrated that using AI to compose personal communications leads to profoundly negative reactions; recipients perceive the sender as having expended less effort, which reduces relationship satisfaction and increases relational uncertainty [4]. I often get the feeling that if I had wanted to talk to a robot, I would have emailed one directly. The work from Ohio State was completed in 2023 and many would argue that LLMs have got better at sounding more human… but why do I still feel so irritated?

The Cognitive Cost of the Shortcut

Beyond the social friction, there is a growing scientific concern regarding what we lose when we outsource our writing. Writing is not merely transcription; it is a tool for thinking. When we wrestle with syntax and structure, we are actually wrestling with ideas [5][6]. Today I sat through a Key Opinion Leader meeting where I was told I didn’t need to take notes as the session was being recorded. I feel I have learned nothing and simply struggled to stay awake for the last 8 hours.

Recent research presented by the MIT Media Lab warns of an "accumulation of cognitive debt" when using AI assistants for writing tasks. In their 2025 study, participants who used LLMs exhibited the weakest neural connectivity during writing tasks compared to those who wrote unassisted [7][8]. Even more concerning, when heavy LLM users were later forced to write without AI, they showed signs of neural under-engagement. You might say they had lost the muscle memory of thought. Furthermore, self-reported ownership of the content was lowest among AI users, and they struggled to accurately quote their own work [9]. The study concluded that while AI offers immediate convenience, the long-term implications for learning and critical analysis are deeply troubling.

We are not just ‘developing’ skills by using AI; I believe that in many cases we are actively losing the executive functions required to structure an argument from scratch.

Authenticity is the New Metric

So, how do we navigate the pressure to improve our productivity, deliver perfectly crafted documents and yet resist mind-muscle wastage ? First, we all have to accept that the standard for ‘good writing’ has shifted from pure grammatical perfection to authentic effort. In a world where anyone can generate a perfect sentence, the differentiator is the unique fingerprint of human experience and feelings.

At Niche, we are updating our guidance. Yes, use your spellcheckers and your grammar checkers. Use AI to defeat the blank page or summarize a complex topic. But do not let it take away your voice. A study comparing perceived empathy found that authenticity, believing that someone genuinely invested time and emotional effort, plays a critical role in how we experience a message [3]. We crave the human touch, even when the AI output is technically flawless.

The Verdict

Returning to the title of my 2020 piece: You will be judged [1]. You can be sure that this has not changed. But the criteria have evolved. You are no longer just judged on whether you can spell or use a semi-colon. You are judged on whether you care enough to engage your own brain and that of your reader.

And here is the message to all authors and Medical Writers in particular, if you outsource your thinking to a machine, do not be surprised when colleagues decide you are replaceable. Keep the 2020 habits: print your work, edit drafts, read out loud, and let your ideas ‘dry.’ But add a new rule for 2026: Write your own first draft. Let AI quarry and shape the stone, but polish it yourself.

References

1. Hardman TC (2020). You will be judged.
2. O'Connell M. We are up to our necks in a rising tide of AI-generated slop. The Irish Times. 2024 Nov 29.
3. Perry A, et al. Comparing the value of perceived human versus AI-generated empathy. Nature Human Behaviour. 2025 Jul.
4. Liu B, et al. Artificial intelligence and perceived effort in relationship maintenance: Effects on relationship satisfaction and uncertainty. Journal of Social and Personal Relationships. 2023.
5. Hardman TC (2019). Is Handwriting Better Than Typing for Note Taking and Memory?
6. Hardman TC (2026). Handwriting vs Typing vs AI: What Actually Helps Us Learn?
7. Rose J (2023). Your Friends Will Hate You If You Use AI to Write Texts, Science Confirms
8. Kosmyna N, Hauptmann E, Yuan YT, Situ J, Liao XH, Beresnitzky AV, Braunstein I, Maes P. Your brain on ChatGPT: accumulation of cognitive debt when using an AI assistant for essay writing task. arXiv:2506.08872. 2025.
9. Kosmyna N, Hauptmann E, Yuan YT, Situ J, Liao XH, Beresnitzky AV, Braunstein I, Maes P. Your brain on ChatGPT: accumulation of cognitive debt when using an AI assistant for essay writing task. MIT Media Lab Publications. 2025 Jun.