One important assumption of the trolley problem is that I have perfect unbiased knowledge of the situation.
However, in many important cases, like should I kill one person now to save 10 people in the future, the two paths have different level of uncertainty. Killing a person now is almost certain, but saving 10 people in the future is much less certain (for example, if I donate to a charity which will save them, there could be many errors in estimation of the charity efficiency).
One important assumption of the trolley problem is that I have perfect unbiased knowledge of the situation.
However, in many important cases, like should I kill one person now to save 10 people in the future, the two paths have different level of uncertainty. Killing a person now is almost certain, but saving 10 people in the future is much less certain (for example, if I donate to a charity which will save them, there could be many errors in estimation of the charity efficiency).