Hacker News new | past | comments | ask | show | jobs | submit login

This is a great post.

One issue here is that a (possibly deeply-layered) "neural network" model applies to many industrial processes, which means that there are a lot of interacting components, but in many of those the input/output relationship is almost binary-- a logistic curve close to a step function. Most companies have absolutely no need for exceptional people. They're looking for a process that can hire good-enough people at controllable volume. They have a mission to be fulfilled and want it done well enough. There's nothing wrong with that; it's just that most of them have no idea how to execute. They might over-hire and reject mediocre-but-good-enough people. They might come up with insane purple squirrel queries or HR-ish hiring policies ("we only hire NoSQL developers; it says here that you use Scala"). They might hire mediocrities (when they actually need strong people) and compensate by hiring a lot of them, which we know doesn't work.

There's also a deep, bilateral trust sparsity in the economy. Most employers aren't doing interesting work and have shit cultures. Most programmers haven't been well-trained or motivated to grow and are incompetent. It goes both ways. I'm just as elitist and network-driven in terms of the companies I'm willing to consider credibile as they are in their willingness to fairly evaluate candidates.

I can't come up with a good way to overcome this in the general case, but the OP gives a strong guide for recruiters to overcome trust sparsity and show capable developers (who can be selective) that they're not actually clueless.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: