Two types of builders. Opposite feelings.

I'm sure you've already heard about this little thing with a medium impact on our lives: AI.

Everyone is busy studying the latest model, sharing their newest skill.md, or trying the latest way to integrate it with R. Well, at least I am! For those of us working on computers, AI has a massive impact on our everyday lives. It's an important and fascinating topic. But there's something I feel nobody talks about much:

The love for coding

15 years ago I watched my internship supervisor build a graph in front of me with R — the first lines of code I had ever seen. It blew my mind.

A few months later I built a Christmas tree in R during a genetics class instead of listening to the lecture. It made all my classmates laugh.

Then, across 12 years as an employee, I spent my nights and weekends exploring package vignettes, hunting for a chart variation that wasn't yet in my graph galleries.

Writing code has always felt like playing a video game to me. You struggle hard, 100% of your brain locked on a single puzzle. And when you solve it and something beautiful appears on your screen, the endorphin hit is marvelous. It always felt like playing and learning — never like working.

2 types of pleasure

When you use code to build things, there are two types of pleasure.

The first is the struggle itself. Crunching code is like playing chess, and I love that feeling of brain sweat. Every great developer I've met shares this love for the act of programming itself.

The second is reaching the end result. Turning a thought into something alive is deeply satisfying, especially for us dataviz people. Finally revealing what was hidden in the data, or landing on something beautiful, is such a joy. But…

AI killed #1

I barely write code anymore. Used properly, and with a bit of knowledge about your field, AI does it all for you.

Take D3.js. It is, without question, the best tool for data visualization — unlimited customization and interactivity. The pinnacle, but also notoriously the hardest to learn. It took me 10 years to get decent at it.

Yet two months ago I launched an online course on it. Students who started from zero already build stunning graphs and dashboards. Here's an example by Bernadett Piros:

You still need general knowledge to communicate with the AI and steer it in the right direction. But the act of coding is gone. Or almost.

2 types of people

I see a lot of people excited about this shift, and I get it! So many were frustrated by code — they can finally build their dreams in minutes.

But I'll be honest: I feel the opposite. I miss pleasure #1. And without the struggle before it, pleasure #2 feels almost empty too. 😔

On top of that, my days have become strangely boring. The AI takes somewhere between 30 seconds and 3 minutes to answer a prompt — too short to multitask, too long to stay focused. If you've found a workaround, I would genuinely love to hear it.

The future

This issue is a very personal one. For the first time in my life, my days at work aren't as fun as they used to be, and I'm thinking hard about what my next move should be.

I do know I'll keep pouring energy into my D3 loves React project. I'm certain many people will move to this technology soon — there's no reason to use anything else nowadays. And watching my students gain these superpowers is the most satisfying feeling. On the data-consulting side, I need to think. The good news? I'm expecting a daughter in a few weeks — so I'll soon step back from tech for a while and think slowly about what comes next.

If any of this resonated, hit reply. I'd love to know I'm not alone in this boat.

Talk soon,
Yan