Forty-three businesses. this one stuck.

Chapter 02·2017

One book, then fifty

The book that did the damage was Thinking Machines by Luke Dormehl. I was supposed to be studying corporate finance. Instead I was reading about founders using machine learning on real problems, and I hit a concept I'd never heard of: affective computing. Teaching machines to read human emotion.

That's where I found Rana el Kaliouby. Her company, Affectiva, came out of the MIT Media Lab and used computer vision to recognize emotion in faces. The use that stopped me cold was helping people with autism understand what the faces around them are feeling. I got so distracted from the exam I was supposed to be studying for that I went down the full rabbit hole of her story, and for the first time I started picturing what it would look like for me to break into AI and help people too.

One book became five, then ten, then fifty. And for the next two years I went back to that same coffee shop, almost every day, with a notebook, teaching myself to write Python from absolute zero. Nobody photographs you becoming obsessed. It just happens, one flat white at a time.

Chapter 04·2020

The degree with no ceremony

I did the Master of Management in Artificial Intelligence at Smith School of Business at Queen's, specializing in natural language processing, as one of the youngest in the class. I paid an unbelievable amount of money to learn from some of the best in the industry, back when you couldn't get that knowledge anywhere else. Would I do it today with YouTube existing? Probably not. Did it change everything? Yes.

It was the COVID year. No ceremony, no gown, no photo of the moment. So I made the photo that should exist: same degree, same campus, same pride, rendered from my own likeness.

What the diploma didn't get me was a job. It took about eighty interviews to land my first data science role, and I kept losing to physicists. Over and over. I was still the finance guy in a room full of PhDs. Eighty rejections is a lot of chances to quit, and I logged every single one as data instead.

Chapter 06·Since high school

Forty-two attempts

To get to forty-three businesses you basically need an allergy to working for people. A visceral one. I had the allergy and the business mind, but no idea what the right vehicle was, so I tried everything. Shopify clothing brands, and I don't care about clothing. Dropshipping. Selling widgets. Dashboards as a service, where I'd build small businesses their Tableau and Power BI reports for next to nothing.

The last one before this was a handyman business. I can barely fix things myself. But I'd met some newcomers to Canada working on porches who needed steady legit work, so I put them on payroll, bought the branded t-shirts, ran the ads, the whole thing. It made money. They all made a little money. What none of them had was the trifecta I kept chasing: something I'm passionate about, something I'm actually good at, and something I have real authority in.

I call what happened next the Slumdog Millionaire strategy. In the movie, every brutal chapter of the kid's life turns out to be the answer to one question. That's what forty-two businesses were. I'll be on a sales call today and close it with a line I learned from a door slammed in my face during failure number thirteen. Nothing was wasted. It just wasn't assembled yet.

Chapter 08·2023

The ten-dollar era

Then people started hearing about ChatGPT, and the gig woke up. What that actually meant: I was selling prompts for ten dollars. I'd put four or five hours into a ten-dollar order, then spend more hours in disputes with buyers trying to shake free work out of me. I lived in fear of the review bar, protecting a perfect five stars from people who were not always fair. Hours away from my family, for almost nothing. Soul-crushing is the honest word.

The tools didn't help. This was the davinci era: four-thousand-token context windows, no memory, nothing agentic. Everything anyone called an AI app was a chatbot in a trench coat, and I Frankensteined every single one together by hand.

Month four, I closed my first ten-thousand-dollar deal and told everyone I knew. Then I went shopping for developers to build it, and they politely explained the thing I'd sold cost twenty thousand to build. I spent six weeks begging for help before refunding the client. I felt like a failure and a sham. But the lesson I hung my hat on was this: somebody had decided, in less than a day, that my idea was worth ten grand. The market was real. My order of operations was wrong. Fix the order, keep the market.

By the end of that run the gig ranked number one in the world for the keyword prompt engineer, pulling fifty to a hundred thousand impressions a month. The ten-dollar era had bought me the thing no shortcut sells: reps.

To make sure there would never be a business number forty-four, I did something a little unhinged. I ordered a sweater that said 43rd time's the charm in every color, and I started filming myself on YouTube wearing it, low quality and all. Putting it on the internet was the whole point. If the world could watch me try, I had no quiet way to quit. This one had to work, so I gave myself no other choice. I still wear the sweater.

Chapter 10·October 25, 2023

The podcast that became a partnership

Taha El-Harti had seen that video, and he drove down to meet me because of it. But the interview that mattered most, I didn't know mattered yet.

He had a brand-new podcast and I was episode one. First time we ever met. Somewhere in that conversation he gave me one piece of advice: start the YouTube channel. Stop waiting. I listened.

That channel is how most people reading this found me. And Taha didn't stay the guy who interviewed me once. He became my business partner. We built the Early AI-dopters community together, and two years after episode one we were standing on a stage in Berlin at our own sold-out event. Some podcasts are content. That one was a hinge.

Chapter 12·2025

Built in public

The YouTube channel Taha told me to start sits at 82k subscribers as you read this. Early AI-dopters grew to 1,300 practitioners with twelve coaches. And we threw our own sold-out event in Berlin, where I gave the keynote. The photos on this site are from those two days.

Remember the trifecta I chased through forty-two businesses? Passion, competency, authority. This is what it looks like when all three finally show up in the same place.