What people miss is the incredible amount of infrastructure development done at Facebook. While it would be an honest statement that it's behind Google (which t

What people miss is the incredible amount of infrastructure development done at Facebook. While it would be an honest statement that it's behind Google (which takes a different approach to infrastructure) in that respect, it also differs crucially from Google that much of it is open source. As a mater of fact, here is one example of code that Facebook contributed heavily to (and that I've worked on as well) that is being used (amongst many other things) to cure cancer: http://www.slideshare.net/ryancox/20101207-o-connortrihughbasetalk http://www.slideshare.net/ryancox/20101207-o-connortrihughba... Not everyone is a geneticist or a bioinformatician. Not everyone who has the capability to write code that Facebook need has the capability to become as good of a geneticists as they are a programmer. Yet, it's not fair to say that this talent is wasted: some Facebookers have contributed to open source, others have joined infrastructure companies (much as others have come from those companies). Why haven't I instead went to work directly for, e.g., NASA or a national lab? Well first, as an undergraduate I deeply _wanted_ to take an internship at NASA (they had a great program for local students) but couldn't as only US citizens were permitted to do so. I was already well into a full-time industry job when I became a US citizen (I've also worked at startups prior to attending college: again, I highly doubt NASA or a research lab would just hire a high school student ). Now I've already heard too many horror stories from classmates working for various government/aerospace/hardware/other traditional orgs about the low autonomy, office-space-esque working condition (be in the office by 8:30 AM, or there'll be a "talk", even if you've worked late into evening), but most importantly about how most of the folks working there are not doing the kind of work I am. Those that _did_ do that kind of work first had to expand a lot of energy proving what was apparently at the start -- that sometimes building something from scratch is less work (both now and later) than trying to shoehorn a problem to an existing but ill-fitting abstraction. Fortunately, I am seeing this change for the better -- and many places (e.g., LLNL) already stand out -- but for now I'll quote Thiel: "rocket scientists go to Wall St for money, but also because aren't allowed to play with rockets anymore!" Obligatory disclaimer: I'm a former Facebook employee and am holding on to my RSUs, but I've sad much the same long before I've worked at Facebook (indeed, this is why I chose to work there!). I am not speaking solely for myself, not on any company's behalf.