Trying to find interventions that accomplish the most good is extremely difficult. Sometimes we have to make uncertain and subjective judgement calls about the quality of some body of evidence, or make speculative forecasts about some unknown variable. It’s very important, therefore, to make sure that we use the best reasoning processes we can to try and improve the accuracy of our judgments.
This Topic focuses on a range of different tools that we might be able to use to increase our good judgment and ability to forecast well.
Key Material
Superforecasting Chapter 1, and Chapter 3 segment (60 mins.)
Estimation and forecasting: an overview | Amanda Ngo | EA Student Summit (Video - 25 mins.)
Stop saying "I can't understand" (Video - 7 mins.)
The Cognitive Science of Rationality (15 mins.)
An epistemology for effective altruism? (5 mins.)
Sequence thinking vs. cluster thinking (40 mins.)
The Question of Evidence (Interactive - 35 mins.)
Further Reading
Forecasting and predictions (Video - 1hr.)
Predicting what intervention will produce a positive impact using only expert consensus or deductive reasoning can be hard. Prediction markets and forecasting tournaments pose a potential solution.
Other topics to check out
Cluelessness
Cluelessness
What should we do given that we can't evaluate the vast indirect effects of our actions?
Expected value and fanaticism
Expected value and fanaticism
Is maximising expected value the right approach for trying to do the most good? How should we approach tiny probabilities of doing vast amounts of good?
Expected value of the future
Expected value of the future
Is the expected value of the future positive or negative? A crucial consideration for the value of extinction risk reduction.