Rediscovering love for writing code

I’ve recently become more fond of using AI tools in my everyday work — whether that’s understanding legacy code, learning unfamiliar libraries, writing complex behaviour, or just brainstorming new ways to build what I want to build.

But sometimes I get this feeling that AI has drastically changed the way I think about starting new things. I don’t get as “scared” when a project looks super complex; if anything, it’s the opposite. It’s exciting to be able to architect software and feel like I’m doing “real engineering”, because the manual part of writing code is no longer the bottleneck.

And yeah — I still love my Neovim bindings and feeling super cool typing at 100wpm. But that’s not the thing that matters most anymore. If you need to ship fast, there are better ways than copying answers from Stack Overflow… and better ways than copying stuff from Claude, too.

What I’m rediscovering

• Speed isn’t the point: typing fast is fun, but shipping is about clarity, constraints, and good decisions.
• Framework debates matter less: SSR vs not, React vs whatever — those are tools. The craft is in how you structure the system and how you think.
• Engineering gets to be engineering again: AI lowers the cost of “try it and see”, which makes room for more ambitious ideas and cleaner designs.

I genuinely like where the job is going. It enables more ideas to come to life, and it enables talented engineers to actually engineer — to focus less on time constraints and code limitations, and more on building something that makes sense.