I've Used My Own AI to Write Every Post for the Last Two Weeks. Here's What I've Learned.
Three posts in two weeks after months of silence. How building Hadithi and dogfooding my own AI writing tool turned friction into a workflow and broke a long publishing drought.
This is the third post I’ve published in a month. Before I built Hadithi, I could go months without writing anything - the occasional “proxy” research/due-diligence paper aside, my publishing record was basically silence.
That’s not nothing. For someone whose default was silence, three posts in a month is a pattern. And patterns don’t happen by accident.
The Problem Wasn’t Ideas
I’ve never struggled to find things worth writing about. The struggle was always the gap between having a thought and having a publishable post. Research. Drafting. Editing. Formatting. By the time I’d worked through all of it, the energy that sparked the idea in the first place was gone, and so was the post.
I tried the obvious fix. I opened ChatGPT and asked it to help. It did, technically. But it felt like working in a crowded room where everyone is shouting suggestions. Powerful, yes. Focused, no. What I needed wasn’t more horsepower. I needed a workflow.
So I built one. And then I built the tool around it.
What Dogfooding Actually Means
There’s a term for this in tech: dogfooding. Using your own product as your primary user. Microsoft did it in the 1990s with its own networking software, and the practice was so central to how they worked that they named an internal server \\dogfood. It’s widely credited as a key factor in how they surpassed Novell’s market dominance.
The principle is simple: if you build something and then refuse to use it yourself, you don’t actually believe in it.
I believe in Hadithi. So I use it for everything I publish.
What I’ve Learned After a Month
Friction is the enemy of habit, not motivation.
Before Hadithi, I had plenty of motivation to write. What I didn’t have was a frictionless path from idea to draft. The research phase alone could swallow an afternoon. Now that process is compressed. I show up with a topic, and the tool meets me there. The habit is still forming - I’m not going to claim otherwise. But something has shifted. Showing up to write feels less like forcing myself and more like following a path that’s already been cleared.
Building in isolation creates tunnel vision.
Gabriel Fortuna put it well: “Building features in isolation gives you tunnel vision. Real usage shatters that illusion instantly.” Every post I’ve written with Hadithi has taught me something about the product. A clunky step in the workflow. A prompt that needed sharpening. A place where the output didn’t quite match what I had in my head. I wouldn’t have found any of it sitting in a product review meeting. I found it by writing.
The output is proof of the process.
This post is the third one. The posts before it exist. You can read them. That’s not a marketing claim, it’s a verifiable track record. One founder on Reddit described the pattern I recognise in myself: technical founders who build something incredible, then struggle to turn what makes it great into language that lands with other people. Building the tools I needed to grow forced me to solve that problem for myself first.
A writing habit isn’t about writing more. It’s about removing the reasons to stop.
Akash Sharma, co-founder of Magical, blocked two hours on a Friday for dogfooding his own product and called it the fastest way to develop product insights. I’d go further: it’s also the fastest way to develop conviction. When you use your own tool to do real work - not a demo, not a test, but actual work that goes out into the world - you stop being a builder who thinks the product is good and become a user who knows it is.
The Honest Part
I’m not a neutral user. I’m building this tool around my own workflow, my voice, my writing process. The risk with dogfooding - and it’s a real one - is that the creator is not the customer. I’ve designed Hadithi around how I think about writing. That’s a strength right now. It will become a limitation the moment I assume everyone thinks the same way.
That’s the next problem to solve. But it’s a much better problem than the one I started with, which was: why can’t I just write consistently?
Now I can. Three blog posts in two weeks. For someone who went months without publishing anything, that’s enough to call it progress.
If you’re a founder, consultant, or professional who knows you should be writing more but keeps running into the same wall - Hadithi is what I’m building to knock it down. I use it myself. That’s the whole pitch.