Can the reasoning not be:
1) Our brains are evolved prediction engines
2) It is more beneficial and urgent to predict danger (e.g. approaching predator) than benefit (e.g. approaching prey)
3) There are always more negative outcomes than positive ones (entropy)
4) (1), (2), (3) => the more powerful the prediction engine, the more negative scenarios it will consider