We’re not replacing writers with AI. We’re replacing thinking with AI.
I’ve been spending some time on LinkedIn lately. Scrolling through posts, reading updates, trying to stay connected with what other devs are doing. And honestly? A lot of what I see feels… off.
Like, you can just tell when something’s been written by ChatGPT. The structure is too perfect. The wording too polished. The vibe too lifeless :(
LLMs are great at phrasing things beautifully. But when people rely on them to “write” instead of “express,” the soul of the content is lost. A LinkedIn post saying “I’m grateful to announce…” generated by AI, without a hint of real story or struggle, becomes just another pixel in the sea of performative posts.
Not everything needs to be co-authored by ChatGPT, you know?
I’m not saying using AI is bad. I get it. Writing is hard. Finding the right words, especially when you’re not used to it, can be frustrating. I’ve stared at a blank Notion page enough times to know that feeling. So yeah, if someone has an idea but doesn’t know how to put it together, I completely understand them getting help from a language model.
But here’s where I get a little uncomfortable, when people use AI to write about things they don’t actually understand.
Like I came across this blog post the other day about Docker. Nicely formatted, clear paragraphs, some decent diagrams. But then I saw a command that just… didn’t make sense. Not syntactically wrong, but wrong in the way that tells you the person who wrote it hasn’t really used Docker in the real world. It’s like when someone talks about “how to use git cherry-pick in teams” but you can tell they’ve never been in a messy merge hell on a Friday night.
BTW, never merge or deploy on Friday evenings. You’re welcome.
I’m not trying to gatekeep. Everyone starts somewhere. Everyone has a first blog post. But there’s a difference between learning something and sharing what you learned, versus copy-pasting a bunch of confident sounding text from an LLM and pretending it’s your own experience.
Because here’s the thing, people read these posts. Beginners especially. And they trust them. They try the commands. They absorb the explanations. And when something doesn’t work or doesn’t make sense, they blame themselves, not the blog post. That’s what bothers me the most.
I remember being that beginner. I’ve followed bad advice on Stack Overflow and random blogs. Wasted hours. Wondered if I was just stupid. It’s not a great feeling. And now, with LLMs churning out convincing but sometimes flat-out wrong content, I worry we’re making that problem worse.
Also, on a more personal level, what happened to writing as a way of expressing ourselves?
AI wrote it. You posted it. We all read it. Nobody felt it.
Not just proving you know stuff. Not just hitting publish to grow your personal brand. But to actually share what you’ve learned, what you struggled with, what surprised you. AI can assist with grammar and structure, sure. But it can’t replicate your story. It doesn’t know why Docker finally clicked for you after you nuked your environment for the third time and then rebuilt it from scratch at 2AM. It doesn’t know how you felt when your team shipped that feature and nothing broke in production. Those are the things that make writing worth reading.
I’m not asking for everyone to write like Hemingway. Just… be honest.
There’s a difference between writing with AI and hiding behind it.
Write what you know. Use AI if you want, but don’t let it replace your voice. Especially in tech, where clarity and accuracy matter so much.
If you’ve got real experience with something, even if it’s just a small win, share that. Those are the posts I stop to read. The ones that feel human. That remind me someone else is also figuring things out, one bug at a time.
Anyway, just something I’ve been thinking about. Maybe too much. But I figured I’d write it down myself.
No AI wrote this. Just me, with some tea, typing away on a quiet night.
