By Hiran de Silva
A recent comment from Casper Badenhorst — in a thread discussing voice, authenticity, and AI — gave me pause. It was a fair question: why use AI if you can speak well and express ideas clearly yourself? This reflection, recorded on a walk through the British Library, is my answer. And as it turns out, it touches on a much bigger theme: there are two fundamentally different ways to use AI — and they couldn’t be more opposite in their relationship to truth and creativity.
The Two AI Archetypes
Let me paint the contrast.
Type 1: The Pretender’s Prompt
This is where someone writes a single sentence like:
“Write me a 4,000-word article on nuclear physics that makes me sound like an expert.”
They might not know the first thing about the topic — but they’re now armed with an impressively researched article, ready to pass off as their own.
This, to me, is the inauthentic use of AI. You’ve probably seen it: content that sounds polished, packed with jargon, but lacks any lived experience. It’s the digital equivalent of hiring a ghostwriter to craft a memoir you didn’t live. Social media is already awash with this kind of content, especially in places like the Excel training community, where surface-level knowledge often gets mistaken for depth.
Type 2: The Brain Dump Refinery
Now, this is what I do.
I walk. I talk. I respond in the moment to something I’ve read or seen — often something that’s triggered an idea, a disagreement, or a sense of “hold on, that’s not quite right.” I’ll pick up the phone and record my thoughts — sometimes for five minutes, sometimes for an hour. Over the last six years, I’ve done this literally thousands of times.
What comes out isn’t a polished article. It’s a torrent of insight, frustration, history, critique, memory, and metaphor — all muddled together. It’s unfiltered. It’s real. And until recently, it was also unmanageable.
Here’s the problem: turning these recordings into publishable content would take months. I tried transcribing with tools like Otter.ai, but even that left me staring at a wall of text that I had neither the time nor energy to clean up. I needed an editor. A team. A ghostwriter who understood my style, knew my voice, and could cut through the repetition, tangents, and filler to find the insight.
Enter ChatGPT.
Now, I simply paste in my messy, real, heartfelt transcript with one line at the top:
“Please clean this up.”
And it does. Not as some generic, glossy rehash, but in my voice. In a style that sounds like me — just more organized, with proper paragraphs, headings, and punctuation. Like this article you’re reading now.
So Why Not Just Write It Myself?
Because writing, for me, interrupts the flow.
The minute I start typing or trying to structure my thoughts into neat lines on a screen, the spontaneity vanishes. My train of thought derails. I’ve tried handwriting, I’ve tried typing, but the only thing that captures my ideas in their raw form is talking out loud.
AI has finally enabled me to retain the integrity of my thought process without getting bogged down by formatting, editing, or second-guessing the structure. It’s not replacing my ideas — it’s amplifying them.
The Real Difference: Input Size vs. Output Pretence
So here’s the key distinction between the two AI types:
- The first uses tiny input (“make me sound smart”) to produce fake expertise.
- The second takes in huge input — real thinking, lived experience, messy emotional insight — and makes it presentable, digestible, and useful for others.
One is a shortcut to appear smart.
The other is a tool to help communicate genuine thought.
That’s why I use AI.
A Note on Camera, Voice, and Presentation
Some people have asked: why not just film yourself more? You’ve got a good voice, they say. You come across well. You could make these points directly on video.
And I could — but it’s not about how I show up, it’s about what I’m trying to deliver. Sometimes showing your face adds value. But not always. My priority isn’t being seen — it’s being understood. And that often takes structure, iteration, and polish.
Final Thought: Ideas Are Worthless If You Can’t Communicate Them
Without AI, most of my ideas would have stayed trapped in my phone — dozens of hard drives filled with rants and reflections that nobody would ever see.
With AI, I can share them.
With AI, I can teach.
Right now, I’m building a learning series on enterprise-level Excel under the working title “Hiran: Enterprise Excel.” It’s not just me delivering lectures. I’m creating simulated conversations between a curious student and a knowledgeable guide — because dialogue is the best way to explore complex ideas.
That’s another example of how I’m using AI: not to fake authority, but to unlock new ways of sharing what I’ve lived and learned.
So to Casper, and to anyone else wondering: this is my voice. Just cleaned up, structured, and brought to life with a little help from technology.
And to anyone using AI just to sound smarter than they are: I see you. And we all know the difference.
Transcribed by Otter.ai. Cleaned and structured by ChatGPT.
Add comment