Thinking Is Slower. That's the Point.

AI isn’t new. I’m not early to this. There are countless blog posts filled with smart takes on AI. This isn’t one of them.
It’s just what I’ve been dwelling on lately, especially over the last year as I’ve started letting AI sneak into my workflow more. I used to be stubbornly principled about avoiding it, but I realized that sticking to my guns might mean falling behind.

For me, it started with straight-up skepticism. This thing? It’ll never be useful. It’ll never touch what I can do. Then curiosity kicked in, and it was weirdly impressive how it could predict text, or even fake being a Linux shell and pull it off.

But after a bit, it showed up in a different way. It crept in slowly, not with some big bang, but through this gradual, undeniable shift. A tool that started out as a fun party trick has suddenly become incredibly useful.

It’s there when I’m deciding stuff. When I’m searching. When I’m writing, building, or trying to puzzle something out.
Always whispering: go faster, be smarter, skip the tough bits.

The nudge builds up. Even without fully buying in, I find myself using it all the time. I start handing over work. First the boring stuff, then the meaty parts.

I noticed I’d reach for it on autopilot, not to wrestle with a problem, but to dodge it. Eventually, I realized my thinking wasn’t happening when it used to. And that? That might be an issue.

Loops, Detours, and Accidents

My brain’s always been a bit of a circus, Thanks ADHD. Thoughts loop, interrupt, derail, and double back. It’s messy as hell, but it works in its own chaotic way.

Most of the useful things I’ve built or figured out came from that chaos. I tried dumb ideas, hit dead ends, broke things, and often stumbled onto something better.

AI doesn’t play that game. It mostly smooths out the road, giving you clean, predictable, useful stuff.

But getting the right answer isn’t the same as really understanding what you’re doing. And creating something is not the same as generating it.

Lazy Is the New Easy

Despite my doubts, I still use it. For code, writing, thinking. It’s handy.
The models speed things up, more than I’d like to admit sometimes.

But when I lean on it too hard, my thinking gets fuzzy. The idea feels done before I’ve even spent time with it, like the answer popped up before I finished asking the question.

There’s also this outside pressure that doesn’t get talked about enough, especially in IT, folks expect everything fast, polished, automated. Not using AI can feel like you’re choosing to be slow on purpose.

And for someone whose brain already spirals and derails, that pressure isn’t helpful.

I worry we’re training ourselves to produce, not understand.
A new generation of copy-paste machines.

Talk. Quack. Think.

As someone with dyslexia, I often toss sentences at a model for spell check, or sometimes to rewrite them. The spell check is spot-on, but I rarely go with the rewrites, which are often better structured, maybe even objectively better, but they feel stiff or off. I like the ones that sound like me.

Still, it helps. It clarifies what I’m trying to say, and I can weave that back into my own words. But when I treat it like Google, it usually just regurgitates what I already knew. Or when I try to take a shortcut and ask it for details in a spec, it just hallucinates a new section that contains exactly what I asked about.

So I’ve stopped seeing it as the source of just the answer and started seeing it as a source for questions too.
Like a sparring partner, a personal interactive rubber duck that pushes back and keeps me thinking.
Something to bounce off, not defer to.

Because of that, my detours are back in action, with more loops, more derailments. And honestly, that feels good. What I feed into the model shapes what comes out, so when I ditched the easy answers and started rambling, the model joined the ride.
It even reminds me what I was about to say… assuming one of us hasn’t hit the context limit first.

So I’m Keeping the Mess

I don’t hate AI. I’m blown away by it every day. It’s a killer tool. But I’m trying to be more intentional with it.

Whenever I pass something to a model, I check: Was that mine to hand over?
Did I skip the learning part?
Did I take a short-term step forward, but a long-term step back?

Sometimes it’s fine. Other times, not so much. There’s always this tug-of-war between speed and depth, ease and owning it.

If I stop making space for friction, I stop thinking clearly. And when that happens, I lose the part of me that actually wants to make things. So sure, I will use AI, but I also care about how I get there, not just where I end up.

I’m deliberately keeping the mess. That’s where the real shit happens.