I don’t like writing.
I know that’s a weird way to start the first post on this blog, but it’s true.
In school, I absolutely hated writing assignments. I always felt like my brain operated faster than my words could keep up, so writing felt like a futile exercise in slowing down my thoughts. But over the past few years, I’ve begun to realize that slowing down my thoughts is not counterproductive; it’s actually one of the most important things I could do.
Slowing down is hard
We live in an age of cognitive outsourcing. In the 2010’s, the infinite scroll gave us the chance to passively consume content without choosing what’s next. In the 2020’s, LLMs are freeing us from the burden of thought. And we humans love the path of least resistance - if we can avoid making decisions or thinking, great.
Not only do AI agents give us an escape from critical thinking, but they’re very good at making us feel productive. Ideas that would take days or weeks to validate can now be evaluated and prototyped by an LLM in minutes. The iterative loop is getting shorter and the constraints are dropping. Not to mention voice-to-text models that allow you to express your thoughts without being slowed down by your keyboard.
Between these two highly-appealing draws - needing to spend less time thinking and getting more done per unit time - it’s hard not to be drawn in. And for this very reason, I find it dangerous.
The doomer implications of AI are well-trodden and dissected by people all over the internet (many of them smarter than I would ever claim to be). But this piece isn’t about those manifold risks. It’s about something more specific, and more boring: why it’s so important to slow down and think deliberately in 2026.
Thinking is hard
“Thinking” is a fickle, nebulous thing. We get easily convinced that we’ve “thought” about something, when in reality, our brains just reacted to the stimulus at hand. This reaction is a mirage - it feels like a complete, well-defined thought - but it often is a raw instinct, full of holes and inconsistencies. This isn’t anyone’s fault, it’s human nature.
What’s new is that the mirage is now industrialized. Hand a half-formed instinct to an LLM and it will return something polished, structured, and confident. The output feels like the product of careful thought, but you skipped the part where the thinking happened. You’re left with a finished artifact and no real understanding of whether the idea behind it holds up.
So how do you develop these kernels of thought into something more concrete? By challenging them. This is the definition of critical thinking - debating all sides of a piece of information in an attempt to distill some form of truth.
There are plenty of ways to do this. Talking through an idea with someone smart. Sketching it out. Sitting with it on a long walk. They all work, and they all share something in common: they force you to commit to a specific formulation of the idea, then expose it to scrutiny.
For me, much to my chagrin, the vehicle that works best is prose.
Writing forces you to formulate your thoughts in a (hopefully) comprehensive manner. There’s nowhere to hide; every sentence has to actually connect to the next one. The rough edges of your ideas get laid out clearly for you to evaluate. “Hmm, well that doesn’t make sense, does it?” That moment, more than the written product itself, is what’s valuable. The essay is a byproduct..
Slow down to speed up
As a software engineer in 2026, I use LLMs every day. In truth, I barely write code anymore. But the actual process of writing syntax in an editor was never the essence of building software. It always was about understanding the problem and the constraints, designing an approach, poking holes in it, distilling an architecture, and iterating. Implementation was just a small part of the job. It’s no different today. The important parts involve thinking, not doing - and that’s the part we can’t afford to outsource, no matter how tempting it gets.
So I’m starting this blog.
Not because I have some grand thesis to share, or because I think the world needs another tech blog. I’m starting it because I noticed that the more capable the tools get, the less I’m compelled to think. Ideas arrive half-formed, get handed off to an LLM, and come back looking finished and I’d get a little dopamine hit.
Writing, for me, is the friction that forces the thinking to actually happen. I still don’t like it. That’s sort of the point.