Longtermist Strategies: Broad and Targeted

Original text Luciano Carlos Cunha · Volume IX, p. 64 of the PDF edition · English translation

9.1. What broad strategies and targeted strategies are

To achieve the best impact for all sentient beings, we need to take the long-term future into account. In this regard, there are two main groups of strategies⁷⁸:

• Broad: they try to have the best possible impact across a wide variety of scenarios.
• Targeted: they try to have the best possible impact in specific scenarios.

9.2. Advantages and disadvantages of each kind of strategy

Each targeted strategy, because it has a very specific focus, can have a very large impact if the scenario for which it was designed occurs. However, strategies of this kind can also have a very small or even zero impact if other scenarios occur.

Broad strategies, on the other hand, will have a positive impact in a great many scenarios, perhaps in almost all that occur. However, this impact will be smaller than the impact targeted strategies have when the scenario for which they were designed comes about.

Below, we will look at some examples of broad and targeted strategies.

9.3. Examples of broad strategies

9.3.1. Increasing consideration for sentient beings

One example of a broad strategy consists in achieving changes in attitudes at the societal level. To do this, it is essential to spread antispeciesism and consideration for all forms of sentience. In addition, institutional changes can be promoted that reinforce the changes in attitude achieved and make future regressions more difficult. The idea is for societies to progressively accept consideration for all sentient beings.

It is also important to emphasize that consideration for sentient beings involves: (1) truly considering all sentient beings (and not just some) and (2) caring about all the forms of harm that sentient beings may suffer (and not just some of them). Accordingly, consideration for sentient beings includes:

• Advocating moral consideration for all sentient beings (including those usually disregarded, such as aquatic animals and invertebrates⁷⁹, and the new forms of sentience that may come to exist in the future⁸⁰).
• Addressing all the causes of harm that may affect sentient beings (that is, not only the harms that result from human practices, but also the harms that result from natural processes and other causes of harm that may arise in the future).

This is a broad strategy because the more widespread consideration for all sentient beings becomes, the greater the chances of preventing or minimizing s-risks, regardless of which scenarios occur in the future.

9.3.2. Increasing our capacity for action in the future

Another example of a broad strategy consists in acting now with a view to increasing our capacity for action in the future. Here are some examples of how this can be done:

• Spreading concern for the long-term future and for the risks of future suffering, since this will lead more people to work on this topic.
• Trying to increase the resources of people interested in preventing future suffering.
• Increasing our knowledge about the long term by conducting research, since this will allow us to:

(1) Better understand how to plan broad strategies.

(2) Identify new targeted strategies not considered so far. (3) Clarify which of the available strategies have the greatest potential.

Again, this is an example of a broad strategy because, whatever the future scenarios may be, the greater our knowledge and resources, the greater the chances of preventing or minimizing s-risks.

9.4. Examples of targeted strategies

Targeted strategies deal with specific scenarios. They can therefore be classified according to the negative scenario they address. Here are some examples.

9.4.1. Some strategies focused on preventing animal exploitation

• Preventing new forms of animal exploitation (for example, by prohibiting the development of new forms of exploiting invertebrates for consumption⁸¹).
• Developing proteins and materials of nonanimal origin to replace animal products.
• Trying to prevent artificial intelligence from being used to make animal exploitation more efficient.

9.4.2. Some strategies focused on preventing the suffering of wild animals

• Trying to create and consolidate the field of welfare biology⁸². This field would study animals in their ecosystems from the standpoint of what could affect them positively or negatively as individuals who have well-being (and not as specimens of species or components of ecosystems). Research in welfare biology could guide programs to reduce the suffering of wild animals.
• Establishing regulations on space exploration so as to prevent the intentional or accidental spread of life beyond planet Earth (even nonsentient life, since it could, over a long time, evolve and give rise to sentient beings⁸³).

9.4.3. Some strategies focused on consideration for new forms of sentience

• Trying to prevent the engineering of new forms of sentience, both biological and nonbiological.
• Trying to secure legal rights for nonbiological sentient entities.
• Pursuing a non-anthropocentric alignment of artificial intelligence (that is, ensuring that artificial intelligence does not have anthropocentric values).

9.5. Precautions when choosing longtermist strategies

Above, we looked at some examples of strategies for preventing long-term suffering, both broad and targeted. Listed below are some precautions we should take when deciding on strategies:

(1) We should not underestimate the complexity of the question of which strategies are best. Given the uncertainty about the future, there is probably no single, conclusive answer. We should always remain open to revising our beliefs about which strategies are best, and be prepared to change strategy if necessary.

(2) Cognitive biases can lead us to place too much confidence in certain approaches and to dismiss others prematurely. That is why it is important always to remember that we may be biased in our assessments, to cultivate epistemic humility, and always to seek new information.

(3) It is important to avoid approaches that carry a high risk of causing people with other priorities to reject the proposal to reduce s-risks. We can reduce this risk by advocating goals that can be accepted by a wide range of perspectives⁸⁴.