As I understand Bill Gates is a member of the effective altruism movement, which is a movement that argues that we should support the most effective charities that are actually proven to do good by measurable metrics, not just assumed to do good. The idea is to study different interventions and see how effective and how expensive they are (e.g. let's say we want to pay for textbooks for kids in Sierra Leone to help them complete school. Sure, seems like a good idea. But does that actually result in a greater percentage of students completing school? or is the reason they don't finish school mainly because they can't get enough to eat and they're suffering from tuberculosis, in which case the money would be better spent on tuberculosis treatment and food or just giving money directly to the families?). As a baseline, it compares how effective it is to just give money directly to the world's poorest people. And it only supports charities that are more effective than that, per dollar. This ends up supporting a few niche charities because if charity A proves they can save a life for $100 and charity B proves they can save a life for $200, you may as well dump as much money into charity A as they can handle, and only give money to charity B if charity A runs out of capacity/ability to save lives at that price.