Posting this while technically on company time, but I just needed to get it out somewhere. I’ve been a developer in India for ~20 years, and I’ve seen hiring hype cycles before. But the AI engineer interviews we’re doing right now feel different.
A lot of candidates walk in thinking the job is about building or training models, working on “advanced AI systems,” or doing something close to research. But in reality, most of the work we actually need is much less glamorous and way more chaotic.
In interviews, I keep seeing the same theoretical talk, but the candidates break down completely when I ask how they’d handle real-world unpredictability.
It is so easy to build something that looks like an AI system now. But production is a different game entirely.
I don’t really have a conclusion here. It just feels like the gap between “can build a demo” and “can ship something reliable” is getting misunderstood more and more.
Curious if others hiring right now are seeing the same thing.
[link] [comments]