Comment by plg | Hacker News Reader

Comment by plg | Hacker News Reader

Parent Full thread

The story raises an interesting dilemma that many of us in the tech and academic realm are faced with quite often, when recruiting / hiring people to join our teams (whether at a company or in a lab or in an academic dept). Personality and indeed intelligence is a vector. How much of a priority do you put on the length of the projection of the vector onto the dimension(s) that are of direct relevance (e.g. expertise and creativity in theoretical physics) and how much weight do you put on the other dimensions (of which there are many)?

The risk of hiring a mad/strange/unstable/unusual "genius" is that dealing with the social aspects of their presence is not without its cost, even if there is a high probability they can generate the magic sauce. The risk of hiring a more well rounded set of team members is that you will end up with a pod of best friends who never end up generating the magic sauce.

Personally I favour the latter approach, I strongly believe that a more cohesive, well rounded team ends up being greater than the sum of the parts, and that quite often, a team containing a mad genius devolves into chaos.

Reminds me of the problems with Cowboy Coders. Many people would agree with your second approach.

http://c2.com/cgi/wiki?CowboyCoder http://c2.com/cgi/wiki?CowboyCoding

A private investigator I know was asked by a startup to look into the background of a prospective engineer. She found what she thought were some personal red flags, including a lot of online forum unpleasant rants, and some financial ones, including a previous bankruptcy filing. They hired him anyway because he seemed like a really good engineer. It didn't work out (one problem was that he apparently wasn't a very good engineer) and they let him go, and now he's threatening to blackmail/extort them.

I think businesses and engineering teams would be wise to consider sociability a big factor in choosing talent. But I think in academia, that's less of a concern if the scientist can show that he/she can publish papers and create monetizable research.

Was hiring a PI for perspective hires standard procedure or was there something with this candidate that required it?

I worked for a large pharmaceutical firm that had a whole department involved with just doing background checks on their prospective scientists. Diploma fraud could be a big problem for them.

Perhaps this startup was doing the same thing but just outsourced the work?

No...she is friends with one of the startup founders. She wasn't hired, but did it as a favor. And the investigating was limited to electronic records...not the garbage-snooping, surveillance-from-car type :)

I would have to agree with the later approach as well. I have some limited experience hiring and I would say that I've learned some valuable lessons. One is that the "other dimensions" you list, which I assume are things like ability to work in a team, humility, dedication, work ethic, etc., are far more important than "smarts". To go along with that, having a ton of "smarts" does not compensate for a lack of the other dimensions. I'd say it can in some cases even go the other route and make things worse as a "smart" but non-team player is capable of doing more damage than a less intelligent non-team player.

I feel that the only relevant factor in hiring should be competence. How do you define being well rounded, and how do you measure that when you're interviewing? Is that just a code word for being someone like you?

So a competent employee who is a bore, is rude to coworkers, makes sexist/racist comments is fine. Okay.

...unless you define 'competent' correctly.

you can be a competent solo developer and still leave a team strictly worse off with you than without. i've seen it happen - people who subtract so much by the sum of their bad interactions with people that it outweighs any value their work adds.

I guess it depends on how outstanding your stuff must be. If it is ok to be merely very good, then the later approach should be the best, I think. If you want one of a kind stuff, then you might want to be a little bit more lax about the other aspects. But often there isn't a conflict. I've found that the brightest people are often very nice, too.

Reply on news.ycombinator.com