If you wonder the three criteria with three discrete levels you get a cube that's something like:
Very unfriendly, somewhat friendly, very friendly.
Very untrustworthy, somewhat trustworthy, very trustworthy.
Incompetent, average, very competent.
Now if you take these and model the interactions between people who exhibit a three-pair (e.g. (somewhat friendly, somewhat trustworthy and average) against (v. unfriendly, s. trustworthy, v. competent)) I would imagine the "world" would eventually remove all friendly people (depending on the assumptions).
If you imagine friendliness and competence is correlated I would also imagine you'd get rid of a lot of the competent people as well.
Unfortunately more research would be needed to assert or refute this hypothesis. Anecdotally I'd say friendliness and competence are positively correlated, but not very strongly.
---
To put it another way, if you have a team of people. Will you team as a whole receive more benefit from someone very competent but ultimately toxic and ruins the culture or somewhat incompetent but very friendly?
If you imagine friendliness and competence is correlated I would also imagine you'd get rid of a lot of the competent people as well.
Unfortunately more research would be needed to assert or refute this hypothesis. Anecdotally I'd say friendliness and competence are positively correlated, but not very strongly.
---
To put it another way, if you have a team of people. Will you team as a whole receive more benefit from someone very competent but ultimately toxic and ruins the culture or somewhat incompetent but very friendly?