Our current approach to aid tends to prefer to target the bottom – how to lift those who need help to a point at which they no longer need help (and starting from those who need to be raised the furthest). That point, however, is arbitrary – it’s relative to how well everyone else is doing. What if instead we focused on helping people reach a point at which they too can help others (starting from the people closest to reaching that point). The latter would snowball our ability to help in a way that the former would not.
While meaning well, the ‘un-smartness’ of aid is mostly rooted in its focus on the relief of the greatest immediate suffering. But this is like the boy with his fingers in a dyke.