Mr. Mark Kaskin, I am curious about your claim that it "is not possible even from a utilitarian perspective to judge the outcome of the death of five people as worse than the death of one different person". What, precisely, do you mean by this? Is your claim one about the practicality of the situation - the cost of accessing information, making a decision, dealing with errors, etc. - or are you making a stronger claim that even in theory no decision is possible? Because part of your counter-example raises doubts about this claim. You say the fat man may be a scientist who will develop a cure for cancer which will save thousands of people every year. Yet there is something odd about this claim, since you are actually implicitly are making a life-to-life comparison here (the lives of thousands of people on an ongoing basis vs. the lives of 5 people). A utilitarian might well respond that life-to-life comparisons can be made as long as all of the relevant costs associated with the calculation are taken into considerations (so that the value of the fat man's life is adjusted so that it's worth includes the expected value of the lives he may save if his research bears fruit).