Here, we test if planning strategies discovered automatically through our reinforcement learning pipeline from this work can improve human decision-making in naturalistic decision-tasks. We find that it is the case for roadtrip planning, and mortgage selection.