What do the authors Tetlock and Gardner (2016) mean by saying that superforecasters should be in 'perpetual beta?' Is this a state of working that comes naturally to you? (Why or why not? Can you give a personal example?) I found the discussion regarding prompt feedback in this chapter interesting. Are there ways we can avoid problems like hindsight bias? Do you experience other problems related to making the best use possible of feedback (not just with forecasts, but in general)?

TYPE IN COMPUTER AT LEAST 250 WORDS