< Don't Believe the Hype

"AI vibe coding tools are amazing" and "AI vibe coding tools will be amazing" are two different statements. Right now, people are conflating them.

I use AI coding tools every day. I can feel the trajectory. What was impossible six months ago is routine now. But I also see what the "I built this in an hour" buzz on social media misses. I one-shot plenty of useful features, but these tools have (very) sharp edges and I get a cut or two every day.

Many demos spin up a website or app from nothing. Spinning stuff up from nothing is the easy (and fun) part. Software Engineers love greenfield projects; it's one of the reasons so many software systems get rebuilt over and over again. But building software is still about trade-offs, complexity, and decisions that hold up over time. The tools are getting better at all of that. They're not there yet.

The early adopters aren't necessarily wrong about where this is heading. I feel it day-to-day so it's hard to disagree. But it's an extrapolation building off an exponential trendline, and those assumptions need to be proven out in the real world. Being an early adopter isn't necessarily better than being a fast-follower.

And it matters which early adopters you listen to. The ones generating real signal are rethinking their workflow, not just generating code faster. They're treating AI as a multiplier on a small team rather than a replacement for skill.

So don't believe the hype. But don't mistake it for being wrong, either. It's just early.

I write about AI, organizations, and engineering leverage: find out about me and subscribe here.

Discuss and share via the meta page . Filed under AI, Code, and 100PR.