Hacker News Reader

Comment by ammon | Hacker News Reader

Yeah (author of the blog post here), I think it's a powerful idea. I don't know of any other public attempts to do this (most ideas about hiring stay locked up inside companies). I'd love to hear any ideas you have. Email me at ammon@triplebyte.com

View on HN

You might not be able to share it, but I'd love to see a covariance analysis on that matrix. You have this hypothesis that companies hire based on some sort of culture. Is that culture random or are there types of culture? Product-focused v tech-focused. test-driven v code-review-drived. If so, the companies hiring practices should cluster.

This would be very useful for triplebyte to know. It would tell you very quickly who this or that company would like, without just throwing a bunch of candidates at them and seeing what sticks.

Out of curiosity, on the taxonomy of programmers, is that supposed to be mostly exclusive, or can they overlap?

Not the author, but I know about ontologies generally. Non-overlapping is great, but rarely possible because whatever you're trying to organize might not divide up that way. Next best is eigenvectors. You would find the "pure types" of programer such that each real programer would a linear composition of the pure types. This is really handy for doing statistics. But you might run in to trouble if programmers don't have interaction free types. Beyond that you can use more elaborate interaction models, but those are more of an art then a science. They require a lot of domain expertise to design.

Hope that helps.

Bit of a nitpick, really, but you want a basis, not necessarily just a set of eigenvectors.

Reply on news.ycombinator.com