Consistent Aggregation of Objectives with Diverse Time Preferences Requires Non-Markovian Rewards

Open in new window