[adamchalmers]

The point of paradoxes like this is to demonstrate that even in our incredibly simple toy models of agents, we still run into issues. These paradoxes help point out weaknesses of the models and act as new desiridata for better models.