My Story
I started with LEGO and math. Nothing deep just liked how it felt to build something and solve problems. Nobody pushed me to do it. I just wanted to make something bigger than me. At one point I told people I wanted to work at NASA, not because I understood what that meant, but because I wanted to be around big ideas.
When I was 11, I got curious about how websites worked. My school
blocked right-click and inspect element, so I Googled my way into
view-source://. It felt like peeking behind the
curtain. That was my first real "hack"; not illegal, just curiosity
paired with stubbornness.
At 12, I made a spacebar clicker using JavaScript. It saved your score and everything. I threw it up on a free domain from Freenom and shared it with friends. Was it dumb? Yeah. But it worked.
In 2022, I made a cookie-clicker-style game. But mine had a whole material system: wood -> stone -> iron -> steel -> diamonds -> rubies. The upgrades followed that logic too from axe, wood farm, pickaxe, stone miner. I didn't look up a tutorial there was non but there were tutorials teaching diffent parts of the puzzle. I wanted to build it all myself and figure out how logic stacked on top of logic.
That same year, I found out about AI. I didn't just want to use ChatGPT. I wanted to build my own. I started learning Python and built a console-based chatbot using the OpenAI API. It could only answer one question and forgot everything after. From there, I made a multi-question version using Flask and chat completions. It broke constantly. The API would timeout, memory would leak, frontend wouldn't update, but I kept pushing until it worked.
In 2023, I got into home labbing. The school's computers didn't have the power I was seeking, so I used RealVNC to remote into my PC at home. It worked, but was slow. So I got smarter. I hooked up a Raspberry Pi with a proxy to the school network, which let me securely route traffic to my Chromebook. Nothing shady just wanted to access tools the school didn't have.
In 2024, I found ChatGLM-6B, a Chinese open-source LLM that mimicked OpenAI's API format. Cheaper, local, and actually decent. I started running it on my own server, experimenting with how far I could push it on a budget.
That year I also started circuit bending. Cheap toys would break, so I'd open them up, rewire them, add FX chips, and turn them into something new. Then came the ESP32 experiments, sensors, buttons, microcontrollers all mixed up into anything I wanted.