The olden days
Back in 2013, I decided I was going to get out of the software development business. I was running a one-man consulting gig, and I was burnt out.
While I was winding down the business and figuring out what to do next, I suddenly had some free time on my hands. So I decided to use it to make YouTube videos sharing the things I’d learned over the previous decade.
That went better than expected. Through a series of events directly linked to those videos, a few things happened:
- I got pretty popular on the interwebs.
- I landed my first actual software engineering job working for an actual employer.
- I got very busy raking in large salaries and stopped posting to YouTube.
Ye Old YouTuber
Fast-forward to August 2025. Freshly laid off, I found myself with even more free time on my hands.
For the first time in a long time, I wasn’t immediately concerned with finding my next paycheck. I had some breathing room, at least for a while. So I decided to kill some time making YouTube videos again.
Only this time, everything was different. The things that used to work weren’t working anymore, and I seemed to have lost whatever knack I’d once had for this.
The internet had changed or I did. Probably both.
Multiple Restarts
I started by doing what I’d always done: learning new things and uploading coding tutorials.
Nothing. No views. No subscribers. No momentum.
Apparently, somewhere along the way, people had decided they didn’t want to code anymore. Or at least, they weren’t interested in watching me do it.
The thing I knew how to do was take complicated ideas and make them more digestible. But for most of my career, that meant explaining and demonstrating code. If people weren’t interested in watching code tutorials, what exactly was I supposed to make?
I tried talking to the camera about the state of the industry (ala Mo Bitter). That picked up a few people, but it wasn’t generating any real momentum. Another dead end.
Then I tried detailed walkthroughs showing people how to build AI agents from scratch. That got a little traction, but nothing that really moved the needle.
So I can rant about things, I can’t just code, I was kinda lost.
A Glimmer
At the end of April 2026 I just stopped caring about looking dumb and uploaded some videos on things I was just messing with. A New MLX inference engine, TSRX, Deerflow… just open source stuff I was exploring. Views rev’d up just a bit.
The funny thing is, this is exactly what I was doing in 2013.
Somewhere along the way, I’d become too experienced to talk about things I wasn’t an expert in. I had spent so much time building a career, accumulating experience, and trying to be the person who knew EXACTLY what he was doing that I’d forgotten what made the original channel work.
Back then, I wasn’t necessarily an expert in everything I talked about. I was curious. I’d find something interesting, explore it, figure out what I could, and hit record. That was the whole thing.
I’d been trying to rediscover my old YouTube channel. What I actually needed to rediscover was the mindset that created it. And just like that, a match was once again made.
A Hard Conversation
Explaining to my wife that I was going to focus on YouTube and just stop looking for a job was a tough conversation to get through.
It took several days to get through and just when things had finally settled down enough for us to talk about it, I had to explain the next part of the plan. My immediate goal was to reach 4,000 watch hours before the minimum requirement increased to 8,000 in February 2027. At the time, I had about 150 hours and was gaining roughly eight hours a day.
You could practically hear the enthusiasm draining out of the room. But she eventually found a minimal amount of faith in the plan. We’d been here before. I’d done this before.
Clean house
One thing became increasingly clear: my old content was competing with my new content.
My channel had spent years building an audience around coding tutorials. Now I was posting about local AI, open-source projects, and whatever else caught my attention. I was trying to take the channel in a new direction while leaving a hundred or so old videos sitting there, telling YouTube who it thought my audience was.
In June, I made the difficult decision to unlist about 100 videos. Then I waited.
Would the algorithm figure out what I was doing? Would it find the people who actually wanted to watch this new kind of content? Or had I just thrown away years of work for nothing? By July, I started seeing signs that it was working. My videos had previously been topping out around 500 views. Now I was starting to see 1,000, sometimes 2,000.
October 2026
So after 6 months I have been monetized for almost 4 weeks. I made $250 from YouTube Adsense so far which is a far cry from the $250,000/year I was making a little over 14 months ago. But I’m excited again. I’m building, creating, learning.
I try to avoid the formulas, schools, plans to “EXPLODE ON YOUTUBE” I am now inundated with. I’m not trying to pretend I know everything. And I’m not trying to turn every passing interest into a business opportunity.
I currently have two rules for myself:
- I post when I have something to share
- I schedule time to find those things
That second rule is tricky. You can’t just sit around waiting for inspiration to strike so I need to make time for exploration. I try an endless amount of things that just dont seem to work. I am sometimes tempted to make a video about all the bullshit projects I have struggled through setting up only to realize they don’t actually work as advertised, but I’m probably not going to do that any time soon. It seems petty and I’m not interested in manufacturing content I don’t personally think has value.
The cost of doing business
There are a lot of people in my inbox offering me money these days. Sponsorships, partnerships, and other opportunities. I haven’t taken anyone up on those offers yet, but eventually I will.
To date I have spent:
- $35: Lights purchased from Facebook Marketplace.
- $84: Twitter Blue, which I probably won’t renew.
- $27: Hugging Face.
- $18: Printing a T-shirt I thought people might buy (I’ve sold 2)
I’m using a decade-old camera, an equally old microphone I got working with egghead.io a dozen years ago, and the laptop my former employer let me keep when I was laid off. I almost exclusively share things that cost little or no money. This is all perfect for me, the “Look at my $50k stack of LLM Inference” niche is just not a treadmill I want to be on.
The investment has been time. And even that time has been spent doing things I’d want to do anyway: learning, experimenting, and building stuff.
Onward
4,000 hours seemed impossible 4 months ago, I get that in a week now. 5k views is my current floor. 10,000 subscribers wasn’t even on my roadmap, but as I write this I am at 9,983 and it’s getting late, it will probably cross over sometime while I am sleeping tonight.
There’s probably a lesson here, but mostly I am writing this to remind myself:
HAVE FUN.