Marc is telling it like it is, even if it is a little difficult to hear.

A Ph.D. is a huge opportunity cost with an extraordinarily high risk/reward ratio. I have often had undergraduates ask me if they should pursue a Ph.D. I first ask them if it is their desire to teach at a research university. This desire is a little like becoming an NBA player especially in certain technical fields like physics. A horde of Ph.D. candidates is good for tenured Professors and research universities, but not necessarily for the students. Post-doc hell is left as an excercise for the reader.

If they don't have a burning desire to be a professor or happen to dislike trees, I don't really see the point of a Ph.D. I tell them to at least take a hard look at the costs of their decision either way: real financial, opportunity, time, etc. An objective tally will rarely be in favor of a Ph.D.

I share Marc's (and your) skepticism about the value of a Ph.D., which is part of why I said that graduate degrees are overrated. I also counsel undergraduates to consider the large opportunity costs of graduate school. But I was taking issue with the sloppiness of the statement "most of the people who have a huge impact on the world . . . do not have PhD's", which would seem to be true whether or not having a Ph.D. is particularly valuable. It's sort of like saying "most of the people who have a huge impact on the world are not left-handed". Well, yes, but that's true even if you replace "the world" with "baseball", and yet lefties have proportionally more impact in baseball despite still being a minority. The question is, can the same be said of Ph.D.s (in the world, I mean, not in baseball)?

I think this is software-centric thinking.

If you look at hardware, you really need a Ph.D. to make a serious contribution to theory, and a lot of important founders (Intel springs to mind) had doctorates.

[Off Topic]: mnemonicsloth, please put your email address or some contact info in your news.yc profile, if you don't mind. Or shoot me an email.