The Influence Triad and Future Suffering

Original text Luciano Carlos Cunha · Volume IX, p. 88 of the PDF edition · English translation

13.1. Introduction

The emergence of new technologies may vastly multiply the potential for generating suffering¹¹⁴. This is because they can give enormous power to individuals who do not care about those who would be negatively affected by such technologies, or even to individuals who aim to harm them. However, new technologies can also be tools for minimizing suffering¹¹⁵, provided there is the motivation to do so and the appropriate knowledge.

This is an example of how an increase in power has very different results depending on the goal the agent is pursuing. But, as we will see below, the agents' goal and their degree of power do not by themselves determine whether the outcome will be one thing or another. There is at least one other crucial factor: the degree of knowledge agents have for achieving the goal they pursue.

In this chapter, we will explore the relationship between these three factors (goal pursued, knowledge, and power), imagining various possible combinations of them and investigating how these combinations influence the probability that the outcome will be more negative or more positive for those affected by the agents' decisions.

13.2. The influence triad

How can we know whether a decision will have outcomes with a positive or negative balance for those affected by it? How can we estimate the size of that balance? What factors does this depend on? Although there are many factors, in this chapter we will explore the relationship between three factors that seem to exert great influence: the goal pursued by agents, their degree of knowledge for achieving that goal, and their degree of power to obtain the means necessary to achieve it. We will call these factors the influence triad.

In section 3, these three factors are related in a table, as follows:

• Intention: concerns the goal the agent intends to achieve and how it relates to the degree of consideration the agent gives to those who will be affected. It is divided into:
• Harming those affected.
• Indifferent to those affected. In this case, the agent does not aim to harm, but would not bother to avoid harm if it were a side effect of pursuing its goal, and would cause harm if doing so were a good means of achieving its goal. Likewise, it does not aim to benefit, but would not avoid benefiting if that were a side effect of pursuing its goal, and would benefit if doing so were a good means of achieving its goal. In short, with this type of intention, whether the agent will cause more harm or more benefit, and to what extent, depends on what is most efficient for achieving its goal.
• Benefiting those affected.
• No agent (natural processes that occur without a rational decision-maker).
• Knowledge: concerns the degree to which the agent knows how to use the existing means to achieve the goal it seeks. It is divided into low or high.
• Power: concerns the degree to which the agent has the means to achieve the goal it seeks. It is divided into low or high. This includes, for example, the amount of resources at its disposal and the degree of influence it exerts over other people's decisions or within institutions.

Obviously, these classifications are crude simplifications. For example, degrees of power and knowledge, as well as types of intention, do not have just two or three stages. Instead, they vary along a gradual spectrum, with many possibilities. Even so, crude simplifications can give us some very useful insights, at least as a starting point for investigating this question.

Important: in what follows, when we speak of how far from or close to the goal someone would be, this applies both to goals that have a limit and to those that are unlimited. For example, a goal such as "watching the 2026 World Cup" has a limit, since once the tournament is over, the goal is complete. By contrast, goals as different as "maximizing profits from animal exploitation" and "benefiting sentient beings" are unlimited goals, since it is always possible to increase the results. Thus, when we say "means A is more efficient than means B for achieving goal X," we may be saying either that this means achieves the same result while spending fewer resources (money or time, for example) or that it yields greater results while spending the same amount of resources.

13.3. Relating the factors of the influence triad

The following table explores the possible combinations of the factors of the influence triad. The results concern the degree of probability that the consequences will be positive or negative for those affected. The reasons why the results are as they are for each possibility will be explained in sections 13.4, 13.5, 13.6, and 13.7.

* "No agent" refers to processes that occur without a deliberating agent, such as natural processes. This category was added to show that the table allows us to assess how much more negative/positive the agents' decisions would be compared to the outcomes of natural processes.

In sections 13.4, 13.5, 13.6, and 13.7 we will see the explanations for each outcome, as well as examples.

Important: throughout the explanations and examples, we will speak of "harming/benefiting a lot/a little," "small/large harm/benefit," "slightly/very negative/positive outcome," and so on. This should always be understood in proportion to the size of the goal. For example, if an agent intends to affect (whether to benefit or to harm) a quintillion beings but manages to affect only a million, that is little (compared to its goal), but it is still a large number of individuals affected.

13.4. When the intention is to harm

In the initial table, we had the following possibilities:

These possibilities are explained and illustrated below:

(1) Suppose an agent that aims to harm but has little knowledge of how to achieve its goal and little power to achieve it. It is likely that it will not manage to cause very great harm. Since it has little knowledge of how to achieve its goal, it may even unintentionally produce a benefit while trying to cause harm. However, if this happens, since it has little power, it will probably also be a small benefit.

(2) Suppose an agent that aims to harm and has a lot of power to do so, but little knowledge of how to do it. It is likely that it will not manage to cause very great harm. As in possibility 1, since it has little knowledge of how to achieve its goal, it may even unintentionally produce a benefit while trying to cause harm. This time, however, if this happens, since it has a lot of power, it may even unintentionally cause a large benefit.

Let us look at an example that illustrates possibilities 1 and 2:

Suppose an agent has the following goal: whenever it is possible for an animal to be born into a predominantly positive life, to try to prevent that animal from being born. Let us also imagine that this agent mistakenly believes that most animals born in the wild have predominantly positive lives (in reality, exactly the opposite is the case¹¹⁶). This agent then decides to try to ensure that fewer and fewer animals are born in this situation. Given its lack of knowledge about how to achieve its goal, the more power this agent has to prevent the birth of animals in the wild, the greater the positive effect it will cause (since it will in fact be preventing animals from being born into lives full of suffering).

Now let us look at possibilities 3 and 4:

(3) Suppose an agent that aims to harm and has a lot of knowledge for achieving its goal, but little power to do so. Given its lack of power, it is likely that it will not manage to cause very great harm. However, since it has considerable knowledge for achieving its goal, it will probably not cause a benefit unintentionally, given that its goal is to cause harm.

(4) Suppose an agent that aims to harm and has a lot of knowledge about how to achieve its goal and a lot of power to do so. The probability that the outcome will be very negative for those affected is extremely high.

Let us look at an example that illustrates possibilities 3 and 4:

Suppose an agent has the following goal: to maximize suffering throughout the universe. Let us also imagine that this agent knows that the higher an animal's reproductive rate, the greater the chances that most of its offspring will be born into lives full of suffering¹¹⁷. This agent then decides to try to make the animals that maximize the number of offspring reproduce at ever higher rates. Since this agent knows very well what the means of achieving its goal are, the more power it has, the greater the negative effect it will cause. For example, if in the era in which it lives it is already possible to colonize other planets or even galaxies¹¹⁸, and it has the power to do so and to send such animals to reproduce in those other places, the negative effect will be astronomically greater than if it had the power to do the same only on planet Earth.

13.5. The indifferent agent

We will now address the possibilities related to what we will call the indifferent agent. This type of agent has the following characteristics:

• The goal it seeks is not to cause harm, but if pursuing its goal has the side effect of causing harm, it will not refrain from pursuing it for that reason.
• The goal it seeks is not to cause harm, but if causing harm is instrumental to achieving its goal, it will seek to cause that harm.
• The goal it seeks is not to benefit sentient beings in general, but if pursuing its goal has the side effect of benefiting them, it will not refrain from pursuing it for that reason.
• The goal it seeks is not to benefit sentient beings in general, but if benefiting them is instrumental to achieving its goal, it will seek to bring about that benefit.

In the case of decisions by indifferent agents, in order to analyze the probability that the outcome will be positive or negative (and to what extent), it is crucial to know how much the agent's usual way of pursuing its goal causes suffering (instrumentally or as a side effect), and whether there is another, more efficient way of achieving its goal that causes less suffering (or causes no suffering at all, or even prevents suffering). For the sake of simplicity, we are assuming here rational indifferent agents, that is, agents who seek to achieve their goal in the most efficient way possible. In real life, people do not always choose the most efficient means¹¹⁹.

A typical example of an indifferent agent is people who exploit animals¹²⁰. In the vast majority of cases, these people do not aim at harm to the animals for its own sake: what they aim at is maximizing the profits from their exploitation. However, the way they consider most efficient for maximizing this profit has the side effect of maximizing the harm to the animals. Sometimes, the animals' suffering is also instrumental to the goals they aim to achieve (and not merely a side effect). For example, causing animals to suffer is sometimes instrumental to gaining information (in certain pain research, for instance).

This also applies to consumers of products and services derived from animal exploitation. For example, most people who consume animal products do not do so because they aim at harm to the animals for its own sake, but because they like a particular taste. What happens is that, for these people, the harm this entails for the animals is not considered sufficient for them to change their practice. In this case, at least, they too are indifferent agents.

Whether we are talking about producers or consumers, since we are talking about indifferent agents, it is not enough, in order for them to change their conduct, to show that substitute methods for the use of animals are already available or can be developed (such as cultivated meat or research methods that do not use animals, for example). For the indifferent agent to change its practice, it must believe that the method that does not harm animals is more efficient for achieving its goal.

For the sake of simplicity, the explanations and examples below deal with cases in which the way the indifferent agent pursues its goal causes suffering (as a side effect and/or instrumentally). Let us see how this relates to the degree of knowledge and power the indifferent agent has in pursuing its goal:

In the initial table we had:

Let us look at the explanations of each possibility, along with examples:

(5) Suppose an indifferent agent with little power to achieve its goals and little knowledge of how to do so.

If the indifferent agent's usual way of pursuing its goal causes suffering (as a side effect or instrumentally), since it has little knowledge of how to achieve it and little power to do so, the effect is likely to be less negative than in the following possibilities (since here it would probably achieve these goals to a lesser degree than in the next possibilities).

(6) Suppose an indifferent agent with little knowledge of how to achieve its goals, but with a lot of power to do so.

If the indifferent agent's usual way of pursuing its goal causes suffering (as a side effect or instrumentally), despite its great power to do so, since it has little knowledge of how to achieve it, the negative effect is likely not to be maximized (since it would probably achieve these goals to a lesser degree than if it had more knowledge). However, given its high degree of power, the effect is likely to be more negative than in possibility 5.

(7) Suppose an indifferent agent with a lot of knowledge about how to achieve its goals, but little power to do so.

In this case, if the way the indifferent agent chooses to pursue its goal causes suffering (as a side effect or instrumentally), it is because it knows this is an efficient way of achieving it (since it has a lot of knowledge about how to achieve it). However, despite its great knowledge, since it has little power, the negative effect is likely not to be maximized either (since it would probably achieve its goal to a lesser degree than if it had more power). But, again, given its high degree of knowledge, the effect is likely to be more negative than in possibility 5 (since in both possibilities it has little power, but in possibility 7 it has more knowledge).

Now, it is uncertain whether the negative effect would tend to be greater or smaller than in possibility 6, since this would depend on: (1) whether what weighs more in producing suffering is the indifferent agent's degree of power or its degree of knowledge, and (2) what the relationship between degree of knowledge and degree of power would be in each example of cases of types 6 and 7.

(8) Suppose an indifferent agent with a lot of knowledge about how to achieve its goals and a lot of power to do so.

In this case, since we are talking about an agent with a lot of knowledge, the chances are high that it will discover the most efficient way of pursuing its goal. Since it will also have a lot of power to do so, what is crucial here for determining how much suffering the outcome will contain is what the most efficient way of pursuing its goal is, and how much suffering it results in.

It is also possible that an efficient way for the indifferent agent to achieve its goal also prevents suffering (and perhaps prevents more suffering than it causes). For example, when humans build houses, buildings, and highways in wild areas, they do not aim at the suffering of wild animals for its own sake, but they usually do not care about the harm caused to the wild animals that live there either (in this situation, such agents are therefore indifferent agents). Many animals will probably die because of urbanization. However, because urbanization also reduces the available plant biomass, it also drastically reduces reproductive rates, and thereby also prevents a gigantic number of animals from being born only to suffer. Despite the harm this practice causes, because of the gigantic harm it also prevents, it is likely to result in less death and suffering than having a forest in the same place. This is not to say that such a practice is necessarily justifiable¹²¹. The point is only to show that the practices of indifferent agents sometimes also prevent suffering, owing to the side effects of the way they pursue their goal.

13.6. When the aim is to benefit

In the initial table, we had the following possibilities:

These possibilities are explained and illustrated below:

(9) Suppose an agent that aims to benefit but has little knowledge of how to achieve its goal and little power to do so. Given its little knowledge and little power, it is likely that, if it manages to bring about any positive outcome, it will be only slightly positive. Given its lack of knowledge, it may even unintentionally cause a negative outcome, but since it has little power, this will probably be less negative than if it had more power.

(10) Suppose an agent that aims to benefit and has little knowledge of how to achieve its goal, but a lot of power to do so. Since it has little knowledge, it is likely that, if it manages to bring about any positive outcome, it will be only slightly positive (but since it has more power, it will probably be more positive than in possibility 9). Given its lack of knowledge, it may even cause a negative outcome, and since it has a lot of power, if this happens it may even cause a very negative outcome.

Let us look at an example that illustrates possibilities 9 and 10:

Imagine that a group of people wants to benefit wild animals. Imagine that these people decide to try to bring about the birth of as many animals as possible in the wild, because they mistakenly believe that most animals born in the wild have predominantly positive lives (as we saw, in fact exactly the opposite is the case). The resulting negative effect will be greater the more power these people have to bring animals into existence in the wild.

Now let us look at possibilities 11 and 12:

(11) Suppose an agent that aims to benefit and has a lot of knowledge about how to achieve its goal, but little power to do so. Since it has little power, it is likely that, if it manages to bring about any positive outcome, it will be much less positive than if it had more power. However, since it has considerable knowledge about how to achieve its goal, the chances of causing a negative outcome are low.

(12) Suppose an agent that aims to benefit and has a lot of knowledge about how to achieve its goal and a lot of power to do so. The outcome is likely to be very positive.

Let us look at an example that illustrates possibilities 11 and 12:

Imagine that a group of people wants to benefit wild animals. Suppose also that these people have a great deal of knowledge about which kinds of interventions in nature would tend to have a positive balance and which would tend to have a negative balance over time for the animals that live there¹²². The resulting positive effect will be greater the more power these people have (and the more they increase their knowledge over time).

13.7. Comparison with situations in which there is no decision-making agent

The framework developed here allows us to compare any of the first 12 possibilities (in which there is an agent with an intention to achieve a goal) not only with one another, but also with possibility 13, in which there is no decision-making agent. In possibility 13 we are talking about outcomes that result from natural processes.

To compare any of the first 12 possibilities with possibility 13 (in terms of whether the outcome would be worse or better for those affected, and by how much), we have to weigh the magnitude of the natural harms that would occur against what would probably happen if the risks associated with each of the 12 possibilities materialized.

For example, the victims of animal exploitation are gigantic in number: many trillions per year. However, this number of victims, already enormous, becomes small compared to the victims of natural processes. In turn, if in the future humans gain access to space colonization technology or to the technology for creating sentient beings in digital media, it is quite possible that current suffering (adding together that caused by humans and that caused by natural processes) will become "a grain of sand" compared to the suffering that would be created¹²³.

13.8. What can we try to do?

Below are some suggestions about what to try to do in terms of influencing each type of decision-maker, with the aim of making the outcome tend toward the positive (or at least making it less negative).

13.8.1. Regarding those who already aim to benefit

Trying to broaden the goal

There are already many people trying to bring about benefits, but usually not to all sentient beings. For example, most people who devote themselves to a cause do so only for human causes, even though the vast majority of the suffering in the world is endured by nonhuman animals.

Moreover, even within animal activism, many animals are left aside. For example, when it comes to animals used for consumption, the vast majority of those used (crustaceans and other invertebrates) are relegated to the background or not even mentioned. There is even greater neglect of the situation of animals in the wild as a result of natural processes, even though they are far more numerous than exploited animals.

So it is worth providing these people with information, both about the facts (for example, informing them about which animals are used the most, about the situation of animals in the wild, about s-risks, etc.) and about the arguments that explain why we should consider all sentient beings and take into account all the forms of harm to which they might be subject.

Trying to increase knowledge

We can try to help agents who aim to benefit gain more knowledge about how to achieve their goal, especially if they already have more power (whether in terms of resources, influence, decision-making power, etc.). This would increase the probability that their actions will have a more positive balance but, above all, it would prevent them from being counterproductive.

For example, we can try to provide these agents with information about longtermism and s-risks, about how natural processes affect animals, about which problems affect the largest numbers of animals, about criteria for deciding which causes and problems to prioritize and for evaluating strategies, and so on.

Trying to increase power

We can suggest that agents who already have more knowledge about how best to achieve the goal of benefiting all sentient beings keep increasing that knowledge, and that they also strongly consider trying to increase their power.

For example, they can try to build careers in which they are well paid, seek a position with decision-making power, or increase their degree of influence (which can be done by publishing books and articles, running social media channels, etc.).

13.8.2. Regarding indifferent agents

Trying to make them stop being indifferent

It is possible to try to get indifferent agents to stop being so. Many are indifferent because they have never stopped to think about this issue. For example, the vast majority of animal advocates were, up to some point in their lives, indifferent toward animals used for consumption, and most animal activists are still indifferent to the situation of wild animals and to the long-term future¹²⁴. Of course, many people will not stop being indifferent even after learning the arguments, but at least some of them may come to change.

It is worth focusing on people we perceive as having the potential to change (above all if they have many resources or occupy positions of power, including the power of influence). For example, we can send these people materials that not only show the reasons for considering all sentient beings, but also offer knowledge about how best to benefit them.

Trying to make the most efficient way of achieving their goal a beneficial one

We have seen that whether indifferent agents will harm or benefit other sentient beings depends crucially on what the most efficient way of achieving their goals is. That being so, we can try to make the most efficient way of achieving their goals one that does not involve harming sentient beings. For example, research could be done on ways

of producing products and services to replace the use of animals (such as cultivated meat and research methods that do not use animals).

However, as should be clear, the mere existence of these other possibilities is not enough for indifferent agents to stop harming sentient beings: these other possibilities must be more efficient for achieving the same goal, or allow it to be achieved to a greater extent.

For example, if cultivated meat becomes cheaper to produce than raising and killing animals for the same purpose, producers will probably choose to use it. However, the degree of public acceptance will probably depend on whether the product offers advantages over the animal-based one (for example, being tastier and/or cheaper).

In this regard, we can draw a historical parallel: carts were only gradually replaced by cars because the car offered the public advantages in terms of speed and convenience. Even so, because speciesism remained in force, this did not completely eliminate the exploitation of animals as a means of transportation. Similarly, even if cultivated meat comes to reduce the use of animals for consumption by offering advantages to both producers and the public, activism in defense of consideration for all sentient beings remains necessary. However, if both strategies are combined, there is good potential for change.

Trying to limit power

It is also possible to try to limit the power that indifferent agents have to harm other sentient beings. For example, one can try to create legislation that guarantees legal rights for nonhuman animals¹²⁵ and for possible future sentient beings.

13.8.3. Regarding agents who aim to harm

Is it possible to get them to aim to benefit?

One might think there is no way to get agents who aim to harm to adopt the goal of benefiting instead. Of course, this is much harder (perhaps impossible) if we are talking about sadistic or psychopathic agents, for example. However, it is important to remember that not all agents whose aim is to harm someone do so because they have these tendencies. Many of them have simply absorbed and reproduce the set of values prevailing in their culture without reflecting on it. In addition, the harm they aim to cause is usually directed at specific groups or individuals, and the reasons why people try to cause it vary. They range from feelings of retributive justice to hatred stemming from prejudice against specific individuals. Whatever the motivation, in a good number of cases the individuals who do this believe they are doing the right thing. Thus, if they come to believe, in a given case, that what they are doing is not right, this would give them a motivation to change their conduct.

However, agents who aim to harm are usually more resistant to change than indifferent agents. Thus, our impact per unit of effort and time spent is probably greater if we focus on people who show themselves to be more open to thinking about these issues.

Trying to limit power

It is also possible to try to limit the power agents have to harm other sentient beings. Again, one can try to create legislation that guarantees legal rights for nonhuman animals and for possible future sentient beings.

Other measures that could contribute to such prevention would be: (1) trying to reduce polarization in society (since it increases the chances of conflict and decreases the chances of reflection in society) and (2) strengthening political systems in which power is less concentrated¹²⁶.

13.8.4. Regarding events without agents

In the case of events without agents, we can always try to make the outcome of natural processes less harmful to animals. For example, by developing the field of welfare biology, we can research in depth how to mitigate the negative effects of natural processes on animals living in the wild.

We can also try to increase concern about this issue in society. For example, we can provide information both about how natural processes typically harm animals and about the arguments supporting the proposal to help them and the responses to objections to that proposal.

13.9. Limitations of the analysis presented here and how to correct them in later analyses

Obviously, our analysis has not covered all the variables that influence the outcomes and, as already stated, the tables were devised on the basis of crude simplifications. Below is a list of the simplifications that were made and how they could be refined in later analyses in order to increase accuracy.

(1) The degrees of knowledge and power were set in terms of "low" or "high." There are infinitely many degrees of these variables. Another possibility would be to carry out an analysis in terms of "the less" and "the more."

(2) The analysis assumed completely indifferent agents. Indifference is usually a matter of degree. For example, a moderately indifferent agent may accept paying the price of not achieving its goal in the most efficient way possible, if there is an alternative for achieving it that causes less suffering or that prevents suffering. However, when it comes to harms caused to nonhuman animals, it seems that most people, at least with regard to the consumption of animals, tend to be almost completely indifferent agents.

(3) The analysis assumed agents whose intention is only to harm or only to benefit. There are many situations in which agents intend to harm some individuals and benefit others.

(4) The analysis presupposed that the agents in question are rational, in the sense that they will choose, among the various alternatives, the one they believe to be the most efficient for achieving their goal. In reality, however, agents often act irrationally¹²⁷. There are situations in which an agent, even while believing that a certain strategy is more efficient for achieving its goal, prefers to keep applying the less efficient strategy, not because it considers the more efficient strategy unethical, but because of some bias (such as the simple fact of already being used to the less efficient strategy). One possibility along these lines would be to investigate which biases frequently affect our decisions.

(5) We worked with outcomes of the type "slightly/very positive" and "slightly/very negative." In reality, most of our decisions have both negative and positive effects. Another possibility would be to carry out an analysis in terms of a positive/negative balance and try to quantify the size of that balance.

(6) We did not compare the 13 scenarios with one another here in terms of which would tend to be worse or better, and to what extent. A table ranking them in this way could be attempted as a further development. One way of doing this would be, first, to list which of the 13 possibilities tend to have a positive or negative balance, and to try to estimate how positive or negative the balance would be in each of the possibilities. Obviously, this may be very difficult to compare. For example, as we have seen, one and the same possibility runs the risk of being only slightly positive or even negative, and may, in certain circumstances, be very negative. All of this would have to be taken into account in a comparison of this kind.

Despite these limitations, however, as mentioned earlier, the analysis carried out here offers some useful insights for our decisions. These insights are listed below.

13.10. Conclusions

To produce gigantic harm, an intention to produce harm is not necessary. Indifference, for example, is enough. Moreover, it is possible to produce gigantic harm even with the intention of reducing harm. This is the case if we are mistaken about what will lead to that goal. This harm may be greater than the harm that results when there is the

intention to produce harm. This will depend on how far the degree of power of an agent who aims to benefit exceeds its degree of knowledge for achieving its goal.

All of this converges on the conclusion that people who want to make the world a less bad place need to focus on: (1) gaining more and more knowledge about the most efficient ways of achieving this goal and (2) trying to have more and more power (in terms of resources, influence, holding public office, etc.).

In the absence of either of these two things, the probability of failing to bring about effective change is already high. Knowledge without power is probably inefficient for achieving a positive outcome. Power without knowledge, on the other hand, runs the risk of causing great harm, even with good intentions. We can sum up these two ideas with the phrase "knowledge without power may be useless, but power without knowledge can be disastrous."