When I was 11, my dad gave me an iPod Touch for my birthday. I immediately fell in love with mobile gaming. I spent hundreds of hours on all the classics like Doodle Jump, Angry Birds, and Fruit Ninja. The simple yet addicting game mechanics of early iOS games fascinated me, inspiring me to brainstorm ideas of my own. Slowly but surely, my interest shifted from playing games to understanding how they were made.

I began playing around with frameworks like GameMaker and Unity to bring some of my concepts to life, but it didn't feel right. I wasn't really making software, I was just using someone else's tools.

So, I decided to learn how to code. I watched a few Python tutorials, but it still felt like cheating to me. What do you mean I don't get to manage the program's memory myself? Ridiculous! I needed to go deeper.

I pivoted to C++ and Objective-C, where I finally felt at home. I also started my first real company, Wolverine Games, under which I developed my first iOS title, Twilight Hop. While building Twilight Hop, I refused any and all code that wasn't mine; I learned trigonometry to build the physics engine, and OpenGL to build the graphics engine. Had I taken advantage of preexisting frameworks, I certainly would have produced a higher quality product in a shorter period of time, but I absolutely loved going deep and understanding technology at the lowest of levels.

Even though it was buggy, Twilight Hop still gained traction. I remember obsessively refreshing the iOS developer dashboard to watch the metrics climb. It was a thrilling experience, but it got me thinking — maybe if I had shipped a better product, I would have gotten even more people excited about my game. I discovered a clear tension: my love for going deep was at odds with my desire for real-world impact. I realized I didn't want to just make software. I wanted to make software that people loved using.

Over time, I started to find a balance between technical depth and impact, and it's a balance that helps me choose which projects I work on to this day. I thrive when I can go deep on literally anything — whether it's a new technology, industry, or market segment — and then build a product that leverages those insights. Take Sonara, for example; we explored a relatively stagnant industry in job seeker tooling, lived at the forefront of LLMs and web scraping, and ultimately created a whole new product category in auto-apply technology.

In the AI era, it's far too easy to splash around in the shallow end of the pool, but I believe that real value can only be created after you jump into the deep end. Fortunately, building deep expertise is easier than ever thanks to AI; it's simply a matter of having the curiosity to do it.

Clearly, we're in an AI bubble, with hundreds of new companies launching every day. I believe the ones that survive will be those grounded in deep foundational knowledge of their technology and market. It turns out depth isn't actually a hindrance to impact — it's table stakes.

So, where does that leave me today? For one, I'm mostly coding in Python now, since that enables me to maximize my impact at the bleeding edge of LLMs (and who wants to manage program memory themselves, anyways?) I also happily use AI coding tools to accelerate my work.

That same kid who was hacking away on physics engines for fun is still here, but he's learned to channel his energy into all the right places. The core drive to understand things deeply and build fearlessly remains; only now, it's focused on pushing boundaries, driving innovation, and building software that makes a difference.