Topic 7: Uncertainties in EA

Objectives

This week’s theme is about exploring uncertainties in the wider EA community. It’s important to identify such uncertainties, as this helps signpost what topics need further research and consideration. The resources on the right are grouped into two distinct areas, namely uncertainties in cause areas, and uncertainties in research and methodology. The predicted total reading time is 45 minutes.  

Note that this week, you will discuss these readings in-depth at a 1:1 meeting with your personal mentor. Feel free to jot down your own personal uncertainties about EA to discuss during this 1:1.  

Essential readings (expected reading time 45min)

Uncertainties in Cause Areas 

To begin, read this succinct EA concepts post (3 min), which explains the ‘meat-eater problem’. In doing so, it articulates uncertain trade-offs between different cause areas.  

Next, read ‘On ‘fringe’ ideas’ (5 min), an EA Handbook article that discusses individuals who investigate speculative topics. While acknowledging that some of these ideas may seem surprising, the post ultimately applauds research attempts in the face of uncertainty.  

The final resource in this subsection is a short transcript of a speech (3 min), given by EA co-founder Will MacAskill. He elaborates on the idea of an unknown Cause X, and why identifying it is so important. 

Uncertainties in Research & Methodology  

This Aeon article (10 min) explores the strengths and limitations of randomised control trials (RCTs), which are often used by EAs to determine the effectiveness of different interventions. In particular, the post summarises some of the most common objections to RCTs, and the extent to which said criticisms apply to research methods more generally. 

This article (10 min) outlines how the charity evaluator GiveWell is shifting their approach. While their charity recommendations thus far have been based on robust, clear metrics and evidence (such as from RCT studies), GiveWell is also moving towards assessing other interventions that are harder to directly measure (such as influencing government policy). 

Optional readings

This 'FAQs and common objections’ post (8 min), originally from Week One, has been included again for reference.  It covers many questions people have, so it may be useful if you have any new uncertainties you would like to explore further.

This succinct video (3 min) by Julia Galef argues that we shouldn’t be hyper-critical of people who are trying to do good, even if their approach is far from ideal.  

This post (15 min) by Dr John Danaher, explores how EA’s focus on methodologies may lead to moral blindspots.