What Does the Focus on S-Risks Consist Of?

Original text Luciano Carlos Cunha · Volume IX, p. 56 of the PDF edition · English translation

7.1. Introduction

In this chapter we will address what Baumann (2022, p. 36-40) calls the s-risk focus. We will discuss how this focus relates to the way suffering may be distributed over the future, and what the difference is between an s-risk focus and a narrow focus.

As we have seen, s-risks are the risks that events will occur in the future resulting in suffering on a gigantic scale, exceeding all the suffering that has existed on Earth so far. It is sometimes argued that the most efficient way to prevent suffering in the history of the world from now on is to focus on preventing s-risks⁶⁶, since the magnitude of suffering in these cases may be so astronomical that it would make up most of future suffering. This is the so-called s-risk focus.

However, someone could have the goal of reducing future suffering as efficiently as possible but reject the s-risk focus because they believe that most future suffering will come from the sum of scenarios that are bad, but not bad enough to be s-risks. For this reason, we will now examine the probability of future suffering being distributed in one way or another.

7.2. How will suffering probably be distributed over the future?

Baumann (2022, p. 37-8) observes that many phenomena have what is called a heavy-tailed distribution (in reference to the curve formed on the graph). For example, if we add up all the deaths from wars or earthquakes throughout history, most occurred in a small number of the worst cases. The same may hold for future suffering: a good percentage of it may be concentrated in the most extreme s-risks.

A distribution can be heavy-tailed to a greater or lesser degree. One reason for thinking that future suffering will be heavy-tailed to a high degree is that the scale of future suffering may vary by many orders of magnitude (for example, owing to the potential for

space colonization⁶⁷ or digital sentience⁶⁸). So it could have extremely sharp peaks.

7.3. The s-risk focus does not imply a narrow focus

Focusing on the worst cases does not mean focusing on just a few s-risks that someone considers the worst of all. This is so for the following reasons:

• Even if future suffering has a heavy-tailed distribution, this does not mean that most of the expected future suffering arises from a few specific cases of s-risks. It may arise from the sum of all or many of the s-risks.
• In phenomena with a heavy-tailed distribution, the most extreme 0.1% of cases rarely contain 99% of the total distribution. Thus, it makes sense to focus, for example, on the worst 10% of outcomes rather than on the worst 0.1%.
• Even the most severe 1% of s-risks may still occur in very diverse scenarios, with very diverse sources of suffering.
• Even if most future suffering occurs in a limited number of very specific worst-case scenarios, we may be mistaken about which scenarios these will be.
• Interventions against the worst risks may not be available, or may not be more efficient than interventions that could help across a wide range of risks.
• There may also be a positive tail in the future, which would come from the opportunity to reduce, for example, natural s-risks⁶⁹. Thus, we should try to reduce the negative tail and increase the positive one and, for that reason, prefer interventions that cover many kinds of scenarios.

Baumann (2022, p. 39) concludes that all these observations concern what we should do as a group. Individually, it may make sense to specialize in a narrow range of risks, as long as our collective efforts cover a broad range of risks.