A Civilisational Nudge by Your AI Assistant Writer

Most work isn’t just doing. It’s documenting, updating, escalating, persuading, responding, aligning. Work is words. And words are messy.

TL;DR

Clearer messages mean fewer misunderstandings; better tone means fewer escalations. Better communication means less wasteful, better work, full stop.

Let’s start with how we got here. Office communication, broadly defined, lives in a grey area. It’s not just “tasks” and it’s not exactly “outputs.” It’s what cognitive scientist David Kirsh calls epistemic actions — work that’s about understanding as much as it is about doing. Emails, complaints, memos, meetings: they exist to transmit meaning, not just content.

Why this matters

AI makes our communication less noisy, more intentional and best modulated according to the task at hand.

The code of communication

AI steps into this messy in-between and acts as a harmonizer. Not in a creepy “mind control” way, but in a very practical, boring, brilliant way: it helps match the code of work communication.

Which brings us back to good ol’ communication theory: sender, receiver, message, code. And, crucially, noise.

Noise is the heart of the problem. We think we’re being clear, but we’re not. We write passive-aggressive emails instead of asking questions, we escalate complaints instead of clarifying needs, we oversell instead of clarifying benefits of an idea, and so forth.

This is where AI (particularly generative models like ChatGPT) steps in with a civilisational nudge. Not toward obedience, but toward code-competence. AI helps us match format.

The key idea is that we already want to conform to the code. The “professional register” isn’t forced upon us — it’s what we actively learn and attempt to emulate, imperfectly, at work.

Angry customers

A great example comes from customer service. When AI rewrites aggressive complaints into “more nuanced and balanced outputs,” (in their own GPT words) it’s not dodging the emotion, nor the message. It’s making the message usable, considering also the receiver's needs, including all the information in the order the customer service agent needs to receive, making the case clear, highlighting the points of conflict and making a case compelling, without the emotional overload.

Bubble-bursting

Moreover, AI often fills-in the blind spots: the pesky details you've learned in some expensive course you should remember, and which you end up forgetting to add in your report. AI is great at reminding you of those.

This gap — where AI leaps from adjusting code to actively introducing new ideas to the content — is where the peril and opportunity for another civilisational leap is. Because when one writes a polarised, blunt and ill-informed message, AI is able to propose a more "nuanced", or "layered" message, considering other factors that burst the writer's bubble.

Three main risks

I identify three risks, here.

Be the editor

The new AI-assistant may make mistakes, if working unchecked. Duh!

The boring style of GPT may replace your own style. The development of your own personal writing style may be atrophied by it. Unless you become the editor of your assistant: there's a refined writing opportunity there.

We'll incorporate the quirks

Like any other linguistic trend (remember uptalk and vocal fry?), it's a bit sad, but we will all start communicating a bit like GPT, even if we don't use it. It just inserts new terms — like "delve into" — the zeitgeist, and we, humans, start picking it up.

Authenticity

Lastly,. the problem of authenticity. As we see all the time on LinkedIn and elsewhere, people are just copy-pasting GPT content as their own. I tend to think that there will be an increasing split in the content people generate:

• Theoretical, instructional and how-to-do-this will be increasingly written by GPT

• Lived experience, new critique and fresh data from the world will be primarily valued when written by humans, and especially valuable for training AI.

As usual, the calculation is if the losses outweigh the benefits.

Boring is sexy?

GPT text, when prompted without any flavour, produces that vanilla-style, Eurovision Song Contest type of humor, and phrasal constructions that quickly saturate LinkedIn. But it may save a lot of time, resources, emotional strain and money. Because it's bringing clarity, structure and modulation to a process that is, by definition, subjective and chaotic: the process of writing thoughts down.

This is a civilisational shift towards a boring, monotone present, but it may be just what we need. I wouldn't be surprised if, like me, we are all tired of a world that never gets boring.

The recommendation

Instead of fearing the perfection, lean into it as an active editor, to augment your day-to-day, and continuously refine intent, expression, and understanding.

---

Key takeaways

• Work is made of words, and words are full of noise

• AI doesn’t replace clarity — it delivers it

• We already try to speak the corporate code — AI helps us match it

• Professionalism is performed, not innate — AI supports the performance

• The real gain isn’t speed — it’s signal

---

Summary Protocol

TOPIC: AI as harmonizer in intellectual and office communication

PROBLEM: Human communication in office settings is riddled with noise and misalignment

QUICK TAKEAWAY: AI helps us match the professional code we already try to follow — but better

CORE CONTENT: Communication breakdowns stem from a mismatch between intent and format; AI bridges that gap by harmonizing code

POLITICAL LANDSCAPE: Fear of AI conditioning minds vs. its actual role as format facilitator

QUICK ACTION: Integrate AI into operations where tone, clarity, and intent often go wrong (think customer complaints, internal emails, meeting briefs)

RISK OF DOING NOTHING: Persisting noise in critical communications, leading to misalignment, escalation, and loss of trust

FUTURE PROTOCOL: Design service flows where AI isn't a substitute, but a syntax-enhancer and translator of human intention

🦾 Agent, your linguistic latency is decreasing. Mission clarity improving. Proceed.

ℹThis text was written by a human, augmented by OpenAI GPT Turbo.

---

Further reading

1. Shneiderman, B. (2020). Human-Centered AI. Oxford University Press.

2. Grudin, J. (2022). From Tool to Partner: The Evolution of Human-Computer Interaction. Morgan & Claypool.

3. Harvard Business Review (2023). “AI can make your customer service better — if you let it.”

4. MIT Sloan Management Review (2023). “The Real Value of Generative AI Isn’t Efficiency. It’s Clarity.”

5. IBM Institute for Business Value (2024). The AI-Augmented Enterprise: Bridging Human

6. Intention and Digital Execution