At the IT/OT Insider we use AI daily, and it helps us with research, coding and editing. What it will never do is write our articles for us.
Here we want to be completely transparent about how. Writing a good article takes a lot of time and effort: each line you read has been rewritten and reread multiple times. A single article takes us about 6-8 hours of writing, on top of days or weeks of research and discussions with specialists. So anything that supports that process is welcome.
Here is where we use AI, and where we don’t.
Where we use AI
Coding
Does this need explaining? We hope by now everybody uses, or at least experiments with, AI-assisted coding. We use it for our websites and, more importantly, for all the demos we build for our IT/OT Builds and our Academies.
Research
A huge time saver. What used to take days or even weeks now takes less than an hour.
Before tackling a new domain or subject, we have AI create a ±30-page discussion paper with a long list of references to kickstart our own process. We prioritise independent and academic research where possible, and we use these papers as a starting point, not as the end result.
The website
Is the time of hiring webmasters over? At least for small, engineer-led companies like ours, probably. We rely heavily on AI to maintain and expand our website and to improve layout, readability and searchability. However, all original content is still ours.
Podcast and video transcripts and summaries
Each podcast and most videos come with an article on our blog that summarises the discussion. These summaries are made with AI, based on automated transcripts. Of course, we read them ourselves and the people we interviewed vet them before we post anything.
Every video also ends up in our video archive. That archive is super helpful because we, and search engines, can easily find something discussed in a video. The entire archive is built with AI workflows, from transcript to summary. Luckily, it’s still us and our guests in the videos, with no plans to switch to avatars any time soon.
Text and article editing
This is one we picked up while writing the book. When writing, it helps to have an extra pair of eyes to check for issues and inconsistencies. We use AI in two ways:
- Structural editor: checks for gaps in the logic (e.g. we’ve cut so much text that it becomes confusing for the reader), balance, repetition, consistency in voice (am I writing as “we” or “I” to you, the reader?) and more.
- Copy editor: we’re not native speakers, so we use AI to check grammar, spelling, clarity and consistency.
We don’t simply accept what the AI proposes. We treat it as suggestions on what could be rewritten, but always in our own voice.
Where we do it ourselves
Our book
The IT/OT Handbook is written by us and edited by our publisher. Not by AI. However, as stated above, it’s extremely useful to use AI to help us in the editing phase. Every IT Revolution author – that includes us – works under their formal AI policy, which rests on a simple premise: Original human insight, expertise, judgment, and authorship are essential to every IT Revolution publication.
Articles
The words you read are ours, in our voice, even the em dashes. We try to keep each article focused on a single idea, at ±800 words. You might know the quote from Blaise Pascal: “I have made this one longer only because I did not have the leisure to make it shorter”. It’s absolutely true. Getting to 800 words means a lot of cutting and discussing what exactly we are trying to convey.
Our goal is to make the topic so clear that you don’t just understand and remember it after one pass, but understand it well enough to explain it to somebody else. That’s the only way these ideas spread.
We admit, we tried it. But in the end, the best way to be noticed in a sea of AI slop is not to write AI slop. So yes, we add those stupid emojis intentionally and take a hot shower afterwards to feel clean again.
Research and reading
Wait, didn’t we just say we use AI for research? We use it to get started, but we still check the references and read them, often printed out on paper with lots of notes and discussions between ourselves and others (fun fact: that still works so much better).
Images
We don’t use many images in our articles, and when we do, they are normal stock photos. Our Academy certificates are made by real artists, without AI. Maybe the odd infographic, but only to save some time (we’re not graphic designers, after all…).
Why we believe it’s important to share this
In a world where we are drowning in content, knowing which sources are worth your time is more important than ever. Credibility is key, and being transparent about where we use AI is part of that. That’s why we clearly label every article as AI-generated or not.