Something Is Happening

Does anyone else get that feeling that something is happening but you can’t quite put a finger on it?

Freud had a word for this. He called it das Unheimliche, often translated as the uncanny. It’s that unsettling sensation when something familiar suddenly feels strange. Not foreign. Not new. Just… off. The beauty of the German language is that it is more precise than English allows: heimlich means homely or familiar, but it also means hidden or concealed. The uncanny is what happens when the hidden part of the familiar reveals itself.

After more than a decade of doing something related to computers and computing, I can’t shake the sense that we’re in the middle of exactly this. The tools I’ve used my whole career are still here. The terminal is still the terminal. Code is still code. Bugs too. And yet something underneath has shifted. On one side, it’s painfully obvious. On the other, it feels almost hidden, like a distant buzzing, an off-frequency just at the cusp of perceptible.

01 // Origins

My origin story starts in early 2000s Serbia, where widespread internet access was still a novelty. Kudos to my dad for pioneering wireless broadband in our bubble. He gave me access to the vastness of the internet before I had any idea what to do with it. We look up to the people nearby, and I wanted to be like him: know cool stuff, interact with the terminal, be the guy who could fix things.

So began my first dabblings with MikroTik routers, network switches, and servers. I was brute-forcing my way through. Much like an LLM, really. I didn’t understand what I was doing. I was pattern-matching. Imitating. Copying iptables rules from forums without grasping netfilter chains. Running chmod 777 because it fixed permission errors. Pasting .htaccess rewrites and later nginx configs that felt more like incantations and tribal magic than configuration. The sacred ritual of sudo make install followed by hope. First installs of Fedora. Bricking boot, recovering, asking for help, and ultimately having to start from scratch. Many times.

Fast-forward through mIRC channels, MSN, and early YouTube. It feels like eons have passed. Those first steps feel like they were executed by someone else.

02 // The Friction That Taught Us

One might ask: why the lament for the past?

It’s not a lament. It’s an observation. And like any observation, it needs grounding.

With the newest workflows, vibe coding and the rest, we’re witnessing the end of something we won’t see again. And genuinely, that’s equal parts sad and scary.

One of the first lessons I got from more senior colleagues was simple: read and be patient. There was a certain passive-aggressive elegance to receiving a Let Me Google That For You link. It stung. It was meant to; perhaps it was even needed. Stack Overflow became the great equalizer, the place where you could watch strangers argue about the correct way to center a div or whether tabs were morally superior to spaces (a war that auto-formatters mercifully ended). I even landed a gig and eventually moved to Germany through a connection I made there.

That era had friction, and the friction was the teacher.

The question of how many times can I make the same mistake, debug it, and swear that I won’t make it again… until it happens again.

Now? I genuinely don’t know how people new to technology learn and interact with it anymore. The answers come faster, but I’m not sure they stick the same way. When the solution arrives in seconds, fully formed and confidently explained, does it build the same intuition? Does it ring a bell of “oh, this might not be good”? Or does it just… dissolve? The hours lost in man pages built something that no autocomplete can replicate: patience and resilience.

The data tells a story too.

Stack Overflow’s traffic has collapsed since ChatGPT launched in November 2022. By December 2024, monthly questions dropped to ~25,000, levels not seen since the site launched in 2009. That’s a 70% decline from March 2023. The 2025 Developer Survey reveals an uncomfortable paradox: 84% of developers now use AI tools, yet only 33% trust the accuracy of AI output. 46% actively distrust it.

And here’s the kicker: the biggest frustration, cited by 66% of developers, is dealing with AI solutions that are almost right, but not quite. When asked if AI can do most coding tasks, what would still make them ask a human? 75% said when I don’t trust AI’s answers.

03 // The Trade-Offs

It’s not all rainbows, as some would have you believe. But it’s not doom and gloom either.

This is not an argument against the tools. I use them constantly. They’ve made me faster, more capable, far more than I thought they would. But it’s a reminder that we’re building on ground that’s still settling. The GitHub issues pile up with variations of the same story. People trusted, got burned, learned. The new friction. The new way of working.

Lewis Mumford wrote those words ninety years ago. Different era, different machines, but the same question we should be asking now: what are we trading, and is the exchange worth it?

What I know is that the advances of the last five years are just the beginning. Like anything that disrupts, we need to stay open to change. But we should also be paying attention to what we’re trading away in the process.

Something is happening. I can feel it. I just can’t tell yet whether we’re the ones driving, or if we’ve already become passengers.

Further Reading