Thoughts after reading PG’s “How Universities Should Prepare Founders”

I was genuinely excited to read Paul Graham’s new essay because it feels unusually relevant to my own experience.

I strongly recommend it to my peers and professors. It says much of what I have been trying to express.

Three months ago, I graduated from Technology Innovation Management (TIM) at Carleton University. The honest reason I applied was immigration. I did not come to Canada because I wanted to build a startup or become an entrepreneur.

I still remember telling a professor in my first class in 2024: “If I really wanted to build my own product, I would not have come here. I would have gone to San Francisco.”

The funny thing is that I now have my own product.

I started building Joe Speaking last November as a side project I wanted to exist. AI made me realize I could build it, and I gradually became obsessed with improving it.

That is why PG’s essay feels so personal. I am a live example of learning by building, not learning to pitch a hypothetical company.

My program partly focuses on entrepreneurship, but what I valued was not a founder formula. I found a supportive community and room to be myself. AI helped with the execution; the program gave me space and confidence.

That gave me room to build intensely while still enjoying nature and life in Canada.

A professor once asked how I stay consistent. For me, it starts with trusting myself.

I believe I am building one of the strongest products in my category. Most people may not see it yet, and that is fair. Claims are cheap. I have something to prove, and the product and users must be the evidence.

People tell me to focus more on distribution or paid ads. They are not wrong, but the product is not where I want it. I would rather improve what users feel than send more people into an experience I do not consider good enough.

Before Canada, I was not a typical technical person. Now I manage AI agents like a small technical team and work hands-on with voice-to-voice models, WebRTC, latency, and VAD.

The best way to learn is by building. I may pay a price for lacking a traditional technical foundation, but that gap pushes me to learn faster and go deeper.

Friends ask why I do not pitch more. Why optimize for pitching when I am not fundraising? I would rather polish the product so users feel the difference, not investors hear a better story.

The irony is that building my startup alone may not give me a path to stay in Canada. Last week, I announced I am looking for a full-time permanent role, but most of my energy still goes into improving the product. 😅

Two years ago, I did not have this mindset. Now I believe that as long as I keep building a great product, I can find the right role, or the right role can find me, and I can solve my immigration status problem too.

I still need to dream big.

Whatever your situation is, you can still build.

Trust yourself.