If you contribute to a charity that increases by one part in a trillion the probability of mankind surviving the next century and >>if conditional on this survival mankind will colonize the universe and create a trillion times a trillion sentient lifeforms then >>your donation will on average save a trillion lives.

Alternately, if you do work that increases by one part in a trillion the probability of mankind surviving the next century...

=======

I think there is a lot of value in intelligent charity, but it's a mistake to assume that all careers have the same inherent non-monetary value to society (or to approximate the non-monetary value of all careers as zero). If I understand correctly, the underlying thinking is that the difference in salary between theoretical research and some sort of high-pay job (when multiplied by the value of donating that money to effective charities) outweighs the difference in non-monetary career value?