ChatGPT said I'm in the Top 3% of Users. I feel like a total mess.
February 22, 2026
I got my 2025 ChatGPT Year in Review... and it genuinely surprised me. It says I'm in the top 3% of users - but most days I still feel awkward and messy using it. In this video, I talk about why that "messy" way of prompting might actually be the point. From debugging code to troubleshooting pool chlorine on a family walk, I'm using AI less like a tool... and more like a third person I can think out loud with. In this video: - What "Top 3%" actually means (and what it doesn't) - Why depth matters more than volume - How I use ChatGPT day-to-day (work, life, random problems) - Getting past the awkward phase most people quit in - Using AI to get unstuck, not just move faster
Transcript
00:00 - This Is Not a Tutorial
Welcome to Slow Builds.
I want to start by saying I'm not an expert.
This is not a tutorial.
This isn't a "how to write better prompts" video.
I'm not here to teach you how to use AI properly.
I'm just trying to make sense of something I noticed.
I got my 2025 ChatGPT Year in Review.
And it said I'm in the top 3% of all ChatGPT users.
Also... top 3% of first-time users.
Both surprised me.
And honestly?
I didn't feel proud.
I felt uncomfortable.
Because most days I feel messy using it.
01:30 - Early Tech, That Part I Understand
The "early user" part makes sense to me.
I've always gravitated toward new tools.
Early Twitter.
UserVoice when it was brand new.
South by Southwest during the early SaaS explosion.
We were from the early wave of startups and internet tools growing up in small-town Canada.
So being in the first few percent of users?
I can grasp that.
But top 3% overall?
That scared me a bit.
Because I still feel like I have no idea what I'm doing.
03:00 - I Feel Like a Hot Mess
When I use ChatGPT:
- I rephrase constantly.
- I correct it.
- I correct myself.
- I restart mid-sentence.
Half my chats start with:
"This might be dumb, but..."
And end with:
"Wait. That's not what I meant."
So part of me wondered-
If I'm top 3% and still this unsure...
What does that mean for everyone else?
04:10 - So I Asked It
Instead of guessing, I asked ChatGPT directly:
"You put me in the top 3%. How?"
If you look at volume alone - new chats started -
I'm nowhere near the top.
I don't use it like Google.
I don't fire off quick one-off searches.
The difference, it said, was depth.
I go deep inside conversations.
I stay in them.
I follow up.
I push back.
I clarify.
I ask for alternatives.
I ask it to go deeper.
It values sustained dialogue over quick surface usage.
That made sense.
05:30 - How I Actually Use It
I don't delete chats.
I return to them.
Trips.
Car purchases.
Family texts.
Recipes.
Side projects.
Work integrations.
I'll stay inside one thread and keep building.
Even on walks with my wife.
We once spent an entire 30-minute walk troubleshooting why the pool chlorine kept dropping.
It became a third voice in the conversation.
We could interrupt it.
Correct it.
Ask for clarification.
It didn't replace thinking.
It extended it.
07:00 - Code, Quality, and Going Deeper
I use it heavily for code.
But not to blindly generate more.
I use it to:
- Refactor.
- Re-examine.
- Challenge assumptions.
- Ask if there's a better structure.
There's a narrative out there that AI makes developers ship bloated code.
Maybe.
But I think that's a usage issue.
The tool isn't the problem.
The intent is.
09:00 - It Wasn't Volume. It Was Staying in It.
Being top 3% didn't mean I mastered AI.
I don't have:
- A giant prompt library.
- An elite workflow.
- Fancy automation chains.
It was about staying in the conversation long enough to reach clarity.
Not accepting the first answer.
Rephrasing.
Refining.
That repetition taught me better prompting over time.
11:00 - Messy Is My Default
I use it the way I think.
Midstream.
Switching directions.
Throwing in random thoughts.
In one chat I might ask:
- Who's the shortest NBA player?
- How long to roast a chicken?
- Would an eSIM work for an app idea?
It's chaotic.
But it mirrors how my brain works.
I don't use it to move faster.
I use it to get unstuck.
There's a difference.
12:30 - The Awkward Phase
Most people quit during the awkward phase.
When answers feel generic.
When you don't know how to ask better questions yet.
When you feel like you're "doing it wrong."
I didn't get better because I mastered it.
I got better because I stayed awkward longer.
I kept opening it.
Kept asking.
Kept refining.
14:00 - Context Changes Everything
Over time I learned something simple:
If I need something serious,
I give it context.
I tell it who to be:
- Senior developer
- Therapist
- Investor
- Chef
- Life coach
Not because it's magic.
But because clarity creates better output.
The constraint improves the result.
15:30 - "It's a Time Waster"
I see people saying ChatGPT kills productivity.
And I understand how that can happen.
But that's true of any powerful tool.
Your phone is incredible.
Navigation. Banking. Work. Communication.
It's also three hours of scrolling.
The tool is what you make it.
If you use it to escape,
it becomes escape.
If you use it to build,
it becomes leverage.
Most days, I use it to build.
18:30 - You're Probably Not Behind
If your prompts feel messy...
If you constantly rephrase...
If you restart mid-chat...
You're not behind.
You're probably early.
As long as you're opening it,
trying,
experimenting-
you're ahead of the people who refuse to touch it.
20:00 - It Learns You Slowly
By staying in longer conversations,
it starts learning how you think.
My recap reflected:
- My work as a developer
- My family dynamics
- My travel preferences
- My personal goals
That didn't happen from one-off prompts.
It happened from continuity.
21:30 - A Slow Build on a Fast Tool
This isn't about mastering AI.
It's about staying in the conversation long enough
for it to compound.
It's a slow build
on top of a tool that's moving fast.
And that feels very on-brand for this channel.