I’m Not Anti-AI. I’m Anti-Hype.

I am well aware of the fact that some people think I am opposed to AI. And it is easy to see why: I love clean architecture, I enjoy writing code, I don’t think tests are a necessary evil and I strongly believe that code was never the bottleneck - much of this hard earned wisdom is apparently about to fly out the window.

I have made these points again and again - on my blog, in my podcasts, at meetups and conferences, in watercooler talks and more. But if that is all you have taken away from that, then I am afraid you have missed the point.

Here’s the part that should matter to you: I’m not skeptical from the cheap seats. I’ve led engineering teams, I’ve been the architect people escalated to, and every week on the podcast I have to dig into something new and be right about it fast. I know this craft in my hands, not just in theory. That’s exactly why my skepticism is worth something - I’m not afraid AI will expose me, I’ve simply seen enough cycles to know hype when it walks in.

Yes, I am very critical. Not just about AI, either - about most things. But never out of rejection or spite. Quite the opposite really. I criticise things that I am invested in - because I care. I believe that not criticising and merely blindly following is the greatest disservice anyone can provide. I criticise not to dissuade you from anything, but to contribute to a joint thought-process.

I’m not against AI. I’m against hype - and AI hype is the most dangerous kind, precisely because the thing underneath it is real. How real - only time can tell.

Yet, multiple things can be true at once. I stand by the fact that productivity 10x/100x claims are often shaky at best, the FOMO economy rewards shovel-sellers and the job-title churn is marketing wearing a lab coat.

I firmly believe that we have opened Pandora’s Box and long since passed the point of no return. AI simply does too much well enough already - and there is no indication it is stopping anytime soon. I run local models, I dig into where agents actually earn their place, I use LLMs as writing and sparring partners. The excitement is real; that’s why I want it framed right.

Software development is a craft. The risk of the hype isn’t that AI is bad - it’s that in the stampede we throw out the baby with the bathwater: the judgment, the rigor, the caring about how the thing is actually built. My skepticism is in service of protecting the craft, not resisting the tool.

All of this goes to say: You can always be certain that I’m fair, not dogmatically anti-anything. And here’s the deal that makes it worth it: because I don’t hand out enthusiasm cheaply, when I do back something, that endorsement carries weight a permanent hype-man’s never could. Skepticism isn’t the opposite of being an asset - it’s the collateral behind my yes.