make clean: sizeof(a lifetime of experience) = my(whole life)

2007/04/05

sizeof(a lifetime of experience) = my(whole life)

Today, I had the fortunate experience of getting to spend 30-40 minutes talking to a man who worked at Cray for 26 of the last 27 years. A year (and two weeks) ago, he left them to start at Microsoft. Keep in mind that this is a man far more intelligent than I (we're talking pretty clearly like 99.99 percentile or so).

During the course of the conversation, I asked him about his reasons for leaving Cray. His response was primarily financial - MS fairly obviously has a larger market share than Cray, and he felt like he could do more good for more people by working at MS. Philosophically, I find this very interesting. After some thought, I came up with three major categories to describe approach to such a decision:

Needs of the many over the one -
A form of gross help; a blanket approach. Clearly lots of people could be helped in lots of different ways by someone with the kind of experience he has. Applying that kind of skill to a large number of small problems will definitely make a dent.

Needs of the one over the many -
The entire HPC industry is a niche market and Cray only has a small share of that. Obviously the amount of help supplied will reach a smaller number of people. This set of people, however, tend to have disproportionate weights to the amount of influence they have in the world. Anywhere funded enough to have supercomputers tends to have analysts, which usually feed information to decision-makers. Therefore, some sort of trickle-down effect.

Needs of details, rather than trying to make large generalizations -
Possibly every situation is utterly unique, and must be examined on a case-by-case basis. In this case, he's working with MS' CCS division, which only shipped it's first product in August, and has yet to make a substantial dent in the industry.

That, at least, is my whole collection of feelings on the subject. If you have more/others, I'd love to hear them.

I have little doubt that philosophers have studied this question at length, and probably even formed some sort of general answer to it. I could probably even find it, if I bothered to look. I am not, however, going to. I'm more interested in hearing what you (esteemed and assorted readers) think about this (even if it is just a statement of agreement, with said mythological philosophers).

**Side note: every single Cray engineer I've ever met has been single.