I want to talk about what code is about and what did to my life and mind, not as a career or hobby, but as a way of thinking that I didn't expect and won't let go of.

I've been building TokenGate and learned what I know from many prototype applications over 1000's of hours of code. I started working with LLMs - mostly Claude Sonnet, to help me understand things I didn't know, debug things I couldn't see, and make decisions about architecture I had no formal training for. And somewhere in that process, something shifted.

Code is not a language. It's a lens.

People talk about code like it's a skill, like typing or driving, or something you learn. But that framing misses something important. Code is closer to a way of perceiving. Once it starts working on you, you start seeing structure. Systems, dependencies, state changes, feedback loops, and contracts.

You start asking "what are the rules here?" about things that have nothing to do with a terminal, things that don't relate to feedback loops, things that are part of our life start to gain structure.

I'd call this structural truthiness. Not truth like a fact, but truth like: does this actually hold together? Is the logic real, or am I just convincing myself? Code is ruthless about this in a way that most thinking isn't. It either runs or it doesn't. The compiler doesn't care about your vibes.

That ruthlessness, weirdly, becomes freeing. Because once you trust the structure, you can go anywhere inside it. I've found myself reading things about physics, economics, biology, Lagrangian formulas - fields I didn't learn explicitly but got to "taste" the essence of because of code while actually following the logic because I learned to look for it. Code widened something in me that I had lost as a person: "Perspective".

Working with AI made it weirder and better

Working with an LLM like Sonnet changes the dynamic in an interesting way. You're not just writing code, you're describing what you want, and then negotiating with something that can tell you if your description is coherent. If you're fuzzy, the output is fuzzy. If you're precise, it's surgical.

That forced me to get better at knowing what I actually wanted. Not what I thought I wanted. Not what sounded right. What I could actually articulate as an intention with a structure behind it. That's a skill that transfers everywhere.

The thing that actually matters

Here's the point I keep coming back to:

A calculator computes what must be. A computer equates what you want.

A calculator is deterministic and bounded. You give it inputs, it gives you the only possible output. There's no room for intention, only execution of fixed rules. A computer is different. A computer runs your model of something. It doesn't know what you're trying to do. It just faithfully executes whatever you describe which means the quality of what comes out is a direct reflection of the quality of your thinking going in.

That's not a technical distinction. That's a philosophical one. Code is the medium where intention becomes testable. Where you stop saying "I think this is how it works" and start saying "let's find out". You're not looking up an answer, you're building a small version of your understanding and seeing if it holds.

For some people - myself included, that's the clearest and most satisfying way to think. Not because it's easier, but because it's honest in a way that feels like you're challenging your ability to see the truth in the structure.

If you're on the fence about going deeper into code: Don't think of it as learning a tool. Think of it as picking up a new way to think and challenge what you know.

It's worth it.

What's next for me: The Gemma 4 challenge kicks off today on dev.to. I'll be using the Gemma API to build something. Best of luck to everyone who enters, happy coding!