I read a piece recently from a fellow blogger and MVP, Elio Struyf, that asked the question a lot of us have been quietly avoiding: is blogging still a thing, now that AI can answer most questions before anyone clicks through to your site? It stuck with me, mostly because I’ve been seeing the same signals on my own site and hadn’t really sat down to think through what they mean.
Short version: I think blogging still matters. But not for the same reasons it did five years ago, and not in the same way.
The bots are already reading my blog
If you’ve never looked at who’s actually requesting your pages, do it. I did, and it’s not subtle. A growing share of what shows up in my logs isn’t a person on a laptop, it’s an AI crawler pulling in content to answer someone else’s question in a chat window somewhere else entirely.
That’s not a hypothetical for me. I’ve seen it happen on this site. Organic search traffic from actual humans has flattened while bot requests keep climbing. My reaction to that isn’t panic, though, it’s more like: okay, so who am I actually writing for now?
Why I’m not blocking the bots
I could lock this down with robots.txt rules and bot-blocking at the edge. I’m not going to, at least not for this blog.
Here’s my reasoning: I don’t blog for ad revenue or to fund a product. I write to share what I’ve learned about Microsoft Teams, Copilot, automation, and whatever else I’m knee-deep in that week, and to build a personal brand around that. If someone gets the answer they needed through ChatGPT instead of through my article, and that answer traces back to something I figured out and wrote down, I still consider that a win. The knowledge moved. That was the point.
That calculus changes completely if your blog is your product, if the content itself is what pays the bills, or if it’s the funnel into something commercial. In that case, AI quietly eating your direct traffic is a real business problem, not a philosophical one, and blocking crawlers is a legitimate defensive move, not being anti-progress. That’s just not the position I’m in.
The stuff I refuse to publish
The flip side of “AI can write anything in seconds” is that a lot of what gets published now is obviously, immediately AI-generated, and readers can tell. Emoji stuffed into every heading, the same three-bullet-and-a-dash rhythm on repeat, posts that read like nobody actually decided what they wanted to say before hitting publish.
I stop reading the moment I spot that pattern. Not because AI-assisted writing is bad, I use it too, more on that below, but because there’s a difference between AI helping you say something and AI having nothing to say and dressing it up anyway. Readers can smell the difference even if they can’t always name it.
The only thing I’ve got that a model can’t fully fake is that this stuff actually happened to me. My opinions on a tool, a migration, a feature rollout, those come from having done the thing, not from summarizing what other people said about it.
How I actually use AI when I write
I’m not precious about this, so I’ll say it plainly: I use AI in my writing process, and I said so directly in my Hugo-to-Astro migration post when I used Codex to help with the technical cleanup work. Same principle applies here.
What that looks like in practice: the ideas, the opinions, the “this actually happened to me” parts are mine. What AI helps with is structure, turning a pile of half-formed thoughts into something that reads cleanly, catching the awkward sentence, formatting a code block properly. It’s not doing my thinking for me. It’s doing the parts of writing that used to eat an hour of polishing after I’d already said what I wanted to say.
If I ever let AI generate a post from a one-line prompt with no input from me, I’d expect readers to notice, and they’d be right to stop reading. That’s the line I try not to cross.
What I think actually matters now
A few things I keep coming back to, for anyone else weighing whether to keep blogging in 2026:
Write what only you could write. Tutorials and how-to’s get commoditized fast, AI can reproduce the mechanics of most of them. Your specific experience, the thing that went wrong that nobody documents, your actual opinion on whether a feature is worth the hype, that’s harder to replace.
Don’t chase length for its own sake. Say what needs saying and stop. Readers and bots both prefer that.
Blogging supports the rest of what you do, it isn’t the whole thing. For me that means it sits alongside speaking, community work, and just being useful to people who ask questions directly. I wouldn’t want to build a personal brand on blog traffic alone, bots or not.
Check your own logs before you have opinions about this. I wouldn’t have known any of this was happening on my site if I hadn’t actually looked.
I still write because it’s how I figure out what I actually think about something. That hasn’t changed, and I don’t expect AI to replace that part, it can help me say it better, but it can’t do the thinking for me.