Pragmatically reframing ethics: Impact without judgment

Reflections on ethics and consciousness touch on some of the most profound questions we can ask about existence, including what matters and why, who matters, and how to act. The field of ethics is often framed as an attempt to answer the question “How should we act?” The implicit assumption is that this is an unambiguous question, with potentially clear answers—especially if ethics were ever considered to be “solved”—that could be applied to all spheres of activity, including scientific research and healthcare. This and other long-standing and widely held assumptions behind that framing are rarely questioned. But similarly to how societies have been shifting away from the conventional reliance on purely economic measures of value, such as gross domestic product (GDP), as metrics to be maximized, and toward more direct measures of well-being,Footnote ¹ the field of ethics could benefit from a deeper consideration of human biases about morality and value, including how they affect the very notion of well-being itself.

The language and more descriptive, naturalistic approach of science can itself bring valuable perspectives to ethics, offering reframings that may lead to clearer questions and answers—including the conclusion that precise, purely rational answers are often impossible. In this paper, I describe a broad framework for thinking about ethics centered on sentience and, in particular, the capacity to suffer. By “sentience” I am referring to the capacity to have feelings and sensations—essentially, phenomenal consciousnessFootnote ²—although in animal ethics and welfare, the term is often used more narrowly to mean the ability to perceive experiences as attractive (hedonically positive) or aversive (hedonically negative).Footnote ³

A pragmatic starting point is to think of ethics, not in terms of casting moral judgment, but as determining how and where to act in the world to improve it, or to prevent it from becoming worse. This can be understood as responding to situations that have some inherent degree of urgency to be changed, or preventing such situations from arising. The actions taken may be direct interventions, such as helping others, or indirect ones, including the establishment of rules and guidelines, or improving education to instill concern and promote effectiveness. But the ultimate concern is impact, without which ethical reflections have little practical relevance. This still leaves open the question of what to impact.

This reframing also allows us to avoid the traditional language of morality, including words such as “ought” and “should,” which, when reflected on, can be understood as having narrower, more precisely defined meanings than intended or at least suggested by moral statements. These can include the expression of feelings rooted in moral intuitions (which is what we often mean when we use this language),Footnote ⁴ attempts at persuasion that appeal to such feelings (sometimes used as a form of manipulation, even if justified by the intent), alignment with specific normative frameworks (perhaps the most precise use of such language), or an indication of the most apparently “reasonable” actions, by which I mean some pragmatic combination of rationality and intuition (less precise, but probably the most useful application of this language).Footnote ⁵

Prescriptions in a loftier moral sense can be replaced with logical statements, which may explicitly attribute rationality and compassion to certain actions without implicitly appealing to moral intuitions. And importantly, the focus shifts to ideas and actions that can improve the world, rather than exercising judgment on the character of individual human beings. This means employing some form of consequentialism without a moral realist framing—focusing on impact without the concern of trying to label actions into tidy categories of “right,” “wrong,” “permissible,” “obligatory,” and so forth.

What matters: The urgency of suffering

With this pragmatic, outcome-oriented framing and approach, we can then turn to the question of what matters, and specifically, what makes situations better or worse. This is where sentience plays a central role, because nothing can be said to truly matter independently of sentient beings who can experience it, be affected by it, or attribute value to it. (If value were attributed conceptually rather than just affectively, it would also require higher-order consciousness and not just feelings and sensations.) All the things we care about matter because of their effect on us or other sentient beings. Remove sentient beings from the scenario, and there is just mindless physical matter. Even if that matter is organized in complex arrangements and can self-replicate as living beings, if it is not sentient, then there is no one there for anything to matter to, no perceptions, and no feelings—the phenomenological equivalent of a rock. Even if we personally value objects and systems that are not sentient, there is a fundamental distinction between the intrinsically important experiences of a sentient being and things or situations that are instrumentally important for their effects on sentient beings but do not matter in and of themselves.

But what aspects of sentience actually matter? What experiences are we concerned with in terms of situations being better or worse? To explore these questions more deeply, I suggest that we draw a clear distinction between potential situations or experiences that we value or desire, and actual experiences that we wish to escape. The former can include happiness and meaning, or goods of various sorts that we may believe will lead to these experiential states. The latter typically refer to unpleasant experiences that we are subjected to involuntarily and wish to be rid of. This is what is often meant by suffering, and it is the definition that I will use here. Central to this definition is the subjectivity of suffering as an internal phenomenon, even though different forms of suffering can feel very different, such as intense physical pain, medical conditions such as akathisia or severe tinnitus, or grief over the loss of a loved one. In cases where the experience of an autonomous being is consensual or voluntary, or at least perceived as such, it may remain unpleasant, and we may sometimes refer to it colloquially as suffering, but it does not carry the same ethical significance as an experience that a sentient being would try to avoid.

There have been other proposals for how to define suffering, such as Jennifer Corns’s significantly disrupted agency.Footnote ⁶ Lack of agency clearly matters to suffering: as I just suggested, unpleasant states that are experienced as entirely voluntary do not have an essential quality to properly qualify as suffering. It is, in fact, the perception of agency or lack of it that determines how the unpleasantness is interpreted, that is, as involuntary or not, rather than actual capacity to act in the world. And a general feeling of lack of agency can be a cause of depression,Footnote ⁷ a form of psychological suffering. Increasing a sense of agency or control is one pragmatic means of alleviating this and other forms of suffering.

However, adhering to a definition of suffering as a phenomenological state—distinct from any actual or perceived threat to physical or mental well-being or integrity, or any other objective aspect of a situation that may correlate with the phenomenological state—more closely captures the essence of why it matters. Without the unpleasantness, there is an essential quality missing, a necessary aspect of what we call suffering. The excruciating pain of a cluster headache attack causes suffering not only because the person is unable to abort or otherwise mitigate it, but because of the raw sensation of intense pain itself.Footnote ⁸ The same applies to other sources of physical pain or discomfort. Conversely, a lack of actual or perceived agency in a situation can also be accepted with equanimity, through a release of any craving for change, and not necessarily be accompanied by an unpleasant experience, and so it appears not to be a sufficient condition for what we usually call suffering.

Relatedly, even when there is no perceived agency to change a situation, intensities of suffering—including of a very different character, such as suffering caused by physical pain and by psychological distress—can, in principle, be compared by how strongly the subject wishes to be rid of the experience, which is more directly a function of the experience itself than of the degree of perceived agency. Furthermore, the phenomenological definition I am using also ensures that we do not prematurely exclude simpler biological beings or potential artificial sentient beings, lacking some of the relationships to the world or physical properties of complex organisms, but still having some comparable experiences. A lack of perceived agency as a contributing factor might nonetheless guide our understanding of the neurological pathways that lead to suffering,Footnote ⁹ and also help avoid its inadvertent creation in artificial beings.

The intrinsic badness of non-trivial suffering as an unpleasant and unwanted subjective state is irreducible.Footnote ¹⁰ There may be many different kinds of experiences that qualify as suffering. They may be caused by many different factors, including the actual or feared loss of goods of some kind that are valued by the subject of the suffering. There may be different neurological pathways that can ultimately lead to the unpleasant experience. But they all have in common distress and aversiveness, with an inherent urgency of some degree to change things. Even if we justify the suffering,Footnote ¹¹ as observers or retrospectively as subjects, the badness of the experience itself remains undiminished as something that would be better not to exist.

The asymmetry in urgency of happiness and suffering

While happiness and suffering, especially when both are understood as subjective states, are conceptually polar opposites and are often treated numerically as negatives of one another in the utilitarian branch of ethics, they actually have entirely different statuses from the perspective of urgency of intervention. Suffering has an inherent need to be alleviated. Happiness does not have a comparable need to be created—out of nothing, or as a replacement for a neutral hedonic state.Footnote ^{12 ,} Footnote ¹³

This is not to suggest that happiness and suffering are functionally unrelated within an individual. Happiness can suppress suffering experienced in the same moment,Footnote ¹⁴ and when suffering is relieved, overall happiness is increased. Even when experienced separately from it, happiness can make occasional suffering easier to bear over longer periods of time. A lack of high-intensity blissful moments in one’s life, especially when others are experiencing them, can lead to frustration—a form of suffering. And we may choose some unpleasant experiences that we expect to lead to greater happiness in the future. Although mitigated by their nominally voluntary nature, these unpleasant experiences can still represent a form of suffering, as they may not be perceived as entirely voluntary, just the best choices available given the alternatives.

However, the creation of bliss in a content individual who is not suffering, including from unsatisfied cravings, is never inherently necessary. There is no urgency to change the situation, no implicit call to action. And the creation of future happy beings may represent the fulfilled wishes of individuals in the present who desire such scenarios, but it also has no urgency.

While this perspective, which recognizes a fundamental ethical asymmetry between happiness and suffering, may be at odds with conventional views on the relationship between the two, it actually provides the basis for a more objective ethical framework—one that does not give excessive weight to the hard-wired intuitions of Darwinian minds, programmed to reproduce and expand, despite the suffering this may entail. Humans’ tendency to do almost anything to prolong our own existence has been termed “existence bias” by Thomas Metzinger, who refers to it as our deepest cognitive bias.Footnote ¹⁵ The perspective that I am arguing for resists this bias by denying any inherent urgency to extend existence, except to the extent that not doing so creates suffering.

This perspective also does not negate the relevance of these and related intuitions, such as the desire to thrive, and moral intuitions against causing direct harm or concentrating suffering. A “holistic” ethical framework, which aims to reconcile rationality and intuitions, can deny the ability of sufficient happiness experienced by some to formally outweigh the extreme suffering of others, but still accommodate intuitions that feel essential to what it is to be human.

The flaws of aggregation

If happiness cannot simply balance out suffering, then aggregating happiness and suffering across a population to derive a single measure of value lacks a rational basis. It means trying to reduce a complex mixture of phenomenological states to a number that lacks the ethically relevant information content of the whole and, in particular, fails to accurately capture the degree of urgency of suffering within that population. This is an obvious flaw of still-influential classical utilitarian reasoning, which aims for impact but neglects the distinction between experiences that there is an urgency to relieve, and experiences that, as observers, we might simply wish to bring about or claim to value.Footnote ¹⁶

Just adding together different intensities of suffering experienced by different numbers of individuals to yield a single measure of suffering is also problematic. When intensities of suffering are different enough, they can no longer be meaningfully categorized as the same phenomenon. Most importantly, extreme suffering that is literally unbearable—so extreme that it would cause the subject to prefer to end their life, in that moment overwhelming their existence bias—has a unique quality and cannot be equaled by many people experiencing a significantly lower intensity of suffering. As an analogy, one can ascend many small hills without developing altitude sickness as one can on a single Himalayan mountain trek.Footnote ¹⁷ Just as gaining altitude can only have deleterious consequences if carried out in one continuous climb, neurons signaling some degree of distress can only create a desperate state of agony if firing together simultaneously in one brain.

In practice, our intuitions may strongly sway us toward using our limited resources to help larger numbers of individuals suffering moderately instead of much smaller numbers suffering more intensely. But this is not an inherently rational decision—nor necessarily an irrational one, either—but a search for balance given the realistic need to employ arbitrary solutions that most people would find intuitively reasonable.

The significance and intolerability of extreme suffering

It is worth pausing for a moment to reflect on extreme suffering, not as an abstraction but as something real and horrific. It can be caused by excruciating physical pain, severe mental illness, or other distress, as a result of disease or an accident, or inflicted by humans through armed conflict, torture, or other forms of violence and cruelty, toward other humans or toward animals. The opacity of suffering makes it inherently difficult for those not experiencing it to fully grasp. But the reality of the experience is essential to take into account in our ethical and practical deliberations, which is why it is worth exposing ourselves to relevant descriptions and videos, even if these make for uncomfortable reading and viewing.

What is literally intolerable for those experiencing it may, for all practical purposes, be tolerated by those making policy, who have the privilege of separation from the experiences of those suffering but who would never be able to voluntarily undergo such experiences themselves. In fact, the strongly intuitive belief in the separateness and continuity of our personal identities, while essential for making sense of the world and relating to others on a daily basis, can be seen as an illusion that hides a deeper metaphysical truth, as long recognized by Buddhist traditions and more recently by analytical philosophers such as Derek Parfit.Footnote ^{18 ,} Footnote ^{19 ,} Footnote ²⁰ From this broader, more existential perspective, extreme suffering is intolerable per se, and not just specifically for the physical being experiencing it.Footnote ²¹ Overlooking it and allowing it to persist as we each cede to our existence bias can be seen as irrational behavior from this perspective. On the other hand, its prevention as an overriding goal is a reflection of rationality in a deeper sense, and not merely a function of the vagaries of compassion for others, as important as that also is in practice as a driver of behavior.Footnote ²²

Prioritizing extreme suffering while accommodating intuitions

I have proposed the shorthand designation “xNU+” for an ethical framework that puts the core focus on impact in preventing suffering.Footnote ²³ “NU” represents negative utilitarianism—a form of utilitarianism that focuses mainly or exclusively on suffering, and that has often been treated with skepticism or even disdain because it seems to deny the value of happiness and, if radically interpreted and myopically applied, could theoretically lead to drastic consequences.Footnote ²⁴ The “x” represents a particular concern for extreme suffering as a top priority, and also reflects the rational limits to the use of aggregation to measure impact, as just discussed, denying that any number of instances of trivial suffering could ever have the same importance. And the “+” explicitly recognizes the need to respect human intuitions to thrive and have happy, meaningful lives for the framework to be viable and acceptable—in a sense, applying a deontological safeguard against causing harm or demanding asceticism—without placing the phenomenon of happiness on the same ethical level as that of suffering and its inherent urgency.

Much of our efforts as we seek to improve the world will, indeed, still require that we accommodate people’s many strong intuitions and take sufficiently into concern their own needs, even if others elsewhere are in much greater distress. But this framework reminds us that some suffering is inherently intolerable and always demands attention.

The focus on suffering as the prime concern provides an ethical lens through which to see the world and identify situations that require addressing. As a form of conscious experience, suffering matters for its own sake, independently of the physical form that generates it and the species, substrate, location, or time in which it exists. In deciding on our priorities, the three core aspects of suffering that matter are intensity, duration, and number of instances, while other factors may matter indirectly due to their alignment or misalignment with our intuitions.

All things being equal, a situation with less overall suffering is objectively better than one with more overall suffering. By extension, and again under the hypothetical assumption of all things being equal, any action that has a net impact in preventing suffering from happening—that results in a future with less suffering than would otherwise occur—improves the world. Such actions are rational because they treat all suffering of the same intensity as equivalent and deserving of the same measures to alleviate or prevent it. These actions are also compassionate by definition when driven by a desire to alleviate suffering. This combination of rationality and compassion—subject to the practical constraints of our human intuitions—could be viewed as the essence of ethical thinking and practice.

In reality, such determinations can only be clearly carried out in a small fraction of situations. One limitation is theoretical: as mentioned above with respect to aggregation, there is often no rational, non-arbitrary way of comparing situations that involve different numbers of individuals and intensities of suffering. The other main limitation is practical: it is usually impossible to know all the possible consequences of an action, including longer-term knock-on effects, even if the immediate result appears to be a reduction in suffering. The clearest such determinations are usually in small-scale situations where most factors are constant and there are distinct differences in possible outcomes—for example, deciding whether to provide pain relief to one person here or to several people there—and in much larger-scale, existential reflections on possible future scenarios, such as whether to expand sentient life to other planets if this would entail significant amounts of extreme suffering.

With humans, decision-making also involves the practical need to negotiate with a range of individuals and groups with their own preferences and interests, even if they are not suffering significantly. And as discussed, humans have very strong intuitions about life—including the desires to preserve existence, have blissful moments, create meaning, respect people’s autonomy, and not cause harm or neglect visible suffering—that need to be respected in any viable decision-making system, even if the consequence is a greater amount of extreme suffering. With non-human animals, we are less subject to such constraints. This means that ethical decision-making concerning them can, in principle, apply compassion in a more focused way, aiming specifically to prevent their suffering and prevent future beings from coming into existence who would suffer intensely. Similarly, in considering the potential experiences of an organism or being whose sentience is uncertain and the ethical implications for our actions, we do not need to consider as strongly whether they also experience happiness, but rather, whether they suffer.

Who suffers? Practical and existential implications

This then brings us to a key question: who suffers? As for consciousness more generally, the only suffering we know with absolute certainty is our own. The ability of other humans to suffer is self-evidently true, and we can conclude with high probability that other beings with similar anatomy and physiology have a similar capacity to suffer. The lower the degree of similarity, the greater the uncertainty. We still do not know the precise biological requirements for conscious experience or, more specifically, the ability to suffer, and when these capacities arose in our evolutionary past. Neither do we know the physical requirements more generally for a system to be conscious, although several distinct theories exist.Footnote ²⁵

Sentience in biological organisms is widely believed to require a nervous system with some degree of centralization.Footnote ²⁶ Unpleasant sensations, mediated by such a nervous system, indicate to the organism that it is exposed to some degree of danger. However, simple avoidance behaviors in response to harmful stimuli need not require unpleasant sensations: even in humans, we withdraw our hand from a hot burner more quickly than pain is registered. Even the sensation of pain itself may require higher-order processing to be interpreted as suffering—a feature that allows experienced human practitioners of mindfulness meditation to reduce their suffering in the presence of pain.Footnote ²⁷ It is unclear whether the unpleasantness of pain as something to be relieved—that is, suffering—was one of the earliest forms of sentience (again, in the sense of consciousness) to have appeared in our evolutionary past, closely associated with harm avoidance; or whether some degree of subjective experience in the form of conscious perception first arose as an inextricable property of organisms’ evolved complexity, with the capacity for actual suffering only emerging later with higher degrees of complexity, allowing the interpretation of pain or lack of agency as an involuntary unpleasant experience.

Our intuitions can provide hints, but we need to be cautious not to be misled by them either. Intelligence and sentience are often found together, but they are not equivalent. Responsiveness to the environment, mobility, and communication with other organisms may reflect chemical and biological processes and forms of information processing without anything actually being experienced.

These questions arise concretely, for example, if we examine the Mollusca phylum, which includes organisms ranging in sophisticatedness from the worm-like aplacophoran to the fascinating octopus. The latter is highly intelligent, and its behavior and physiology also provide strong signs of sentience.Footnote ²⁸ Perhaps sentience emerged more than once over the course of evolution, including within this phylum. Or perhaps it was already present over 500 million years ago in our worm-like common ancestor with the Mollusca phylum, which would suggest that even simple organisms like C. elegans are sentient.

Answers to such questions about sentience have practical implications, such as whether bivalves like oysters, also in the Mollusca phylum (though, unusually, with less sophisticated nervous systems than their ancestorsFootnote ²⁹), can be harvested and eaten as sources of nutrients without causing suffering to them or other organisms (an incidental harm of industrial plant agriculture), or whether experimentation on any worm-like animals, in this phylum or another, can be carried out painlessly. The likelihood that octopuses can suffer has direct implications for how we treat them, including in industrial farming operations for human consumption. Similar concerns arise for other sophisticated invertebrates, such as insects.Footnote ³⁰ The commercial farming of insects as a source of protein has also expanded, with large numbers of individual animals involved. If they are able to suffer significantly, their commercialization could represent a substantial increase in the suffering imposed by humans on members of the animal kingdom, beyond the massive and entirely unjustifiable horrors of factory farming carried out on vertebrates.

Reflections on the determinants of consciousness also have profound existential implications: how much intense suffering occurs on our planet independently of human activity or presence, due to factors such as disease, the elements, and predation? Such considerations might seem tangential, especially from a perspective that regards the non-human world as a separate realm with its own “circle of life” dynamics. But once we recognize suffering as an internal phenomenon that matters for its own sake, independently of the entity that experiences it, the specific cause becomes less ethically relevant, and the more important question becomes whether we can actually do anything to address it. To a limited extent, the answer is already yes—there are measures we can take to help some wild animals.Footnote ³¹ Obstacles to larger-scale interventions are not just technical but political, as is already the case for the abolition of widespread cruelty to farmed animals.

The determinants of sentience and suffering are also central to an entirely new concern, with the rapid development of powerful artificial intelligence (AI) and other technologies. There is little reason to believe that sentience could not be created using non-biological substrates. It is uncertain whether digital computer systems with their current architecture have the required properties. We may be led to believe that they do because of the pervasive use of analogies between brains and computers as information processors; our readiness to equate intelligence and sentience; and the development of large language models (LLMs) that closely mimic human communication. But the possibility of AI systems being conscious in the near future is already being taken seriously by some.Footnote ³² And whether as an emergent side effect of technology or through intentional design, artificial sentient systems would bring with them the real risk of suffering, possibly even of extreme severity.

Because of the intrinsic badness of intense suffering, it is reasonable to continuously take measures to progressively reduce its prevalence, and in cases of doubt, to apply the precautionary principle to avoid the possibility of allowing or increasing such suffering.Footnote ^{33 ,} Footnote ³⁴ A moratorium on the creation of any artificial system that might have conscious experiences, therefore, appears a sensible proposition.Footnote ³⁵ In practice, the evidential bar for invoking the precautionary principle remains high—even insects are not currently covered by UK animal welfare legislation.Footnote ^{36 ,} Footnote ³⁷

In some cases, causing harm to a few could actually prevent greater amounts of suffering, although this approach conflicts with our strong intuition against causing direct harm to others or concentrating suffering. This intuition is strongest when it concerns our fellow humans. For example, carrying out medical experiments on unwilling human subjects is among the most abhorrent ideas we can imagine. And yet, we continue to conduct such research on non-human animals, even in the many cases where there is no obvious net benefit in terms of suffering, and the experiments are carried out to enrich our knowledge and satisfy our intellectual curiosity. Applying an ethical framework such as xNU+ would lead us to refrain from such experiments, at least unless an objective analysis, without the influence of any self-interested concerns, demonstrated that the expected result was a clear decrease in extreme suffering. In practice, in the usual absence of such truly objective analyses, a rights-based approach that accords much greater respect to individual non-human lives—in a sense, attributing greater weight to the intuition-respecting “+” part of the framework—may actually provide the conditions for better outcomes (less extreme suffering) in the long run.

Suffering metrics

If the metrics we use to design policies and track progress are to align closely with what matters, then there is a need for specific suffering metrics—not as a replacement for existing metrics, but as a complement, to be used in parallel. The most widely used metrics in health economics and global health, the Quality-Adjusted Life-Year (QALY)Footnote ³⁸ and the Disability-Adjusted Life-Year (DALY),Footnote ³⁹ and newer metrics used for studying societal well-being, termed the Wellbeing-Adjusted Life Year (WELLBY or WALY, with slightly different approaches),Footnote ⁴⁰ only partly correlate with suffering.

Conceptually, the first two are based on the reference point of a life without disease or disability, and the last ones on maximal well-being. The QALY implicitly regards years of healthy life as a parameter to maximize with given resources, and in theory, even low-quality lives just above a threshold of worthiness as potentially meriting replication, regardless of the suffering they contain. As used in some contexts, QALY weights can have negative values for health states considered “worse than dead,” such as ones that involve extreme suffering. But this adaptation of an existing metric is still subject to its existing conceptual limitations, especially an adherence to the idea of aggregating happiness and suffering, the associated questionable notion that an overall life can be objectively determined as worth living or not—which is treated as a distinct question from whether literally unbearable suffering ever occurs during a life—and a failure to track suffering adequately. The DALY more explicitly tracks departure from perfect health, but in its current form, requires observers to compare conditions they do not have, again without an adequate assessment of the suffering caused by conditions. And the WELLBY measures increases in life satisfaction situated anywhere on a scale of 0–10, also without directly assessing suffering.

If our primary concern is the phenomenon of suffering itself, then we need to try to measure it directly, by what is arguably the most reliable means—asking people how much they are suffering, such as on a scale of 0–10.Footnote ⁴¹ The Suffering-Intensity-Adjusted Life-Year (SALY)Footnote ⁴² has been proposed, modeled on the DALY but replacing Years Lived with Disability (YLD) with Years Lived with Suffering (YLS), and with a weighting based on suffering intensity. As the Years of Life Lost (YLL) component of the DALY is not directly relevant to suffering, it could be removed, leaving just YLS.

In order to limit the problems associated with aggregation, we can increase the granularity of such measurements and maintain a distinction between different levels of suffering. I recently proposed the additional metrics Years Lived with Severe Suffering (YLSS), capturing suffering at an intensity of 7/10 and above, and Days Lived with Extreme Suffering (DLES), capturing suffering at an intensity of 9/10 and above.Footnote ⁴³ The choice of days rather than years for the last measure is intended to emphasize the extreme seriousness of such suffering, even when experienced on short timescales. These metrics have already been employed in estimating the global burden of pain from cluster headache,Footnote ⁴⁴ with the assumption that pain and suffering would have roughly similar ratings in such patients.

The main issue with such subjective scales is that people have different internal reference points for what constitutes 10/10 suffering. Even if it is defined as the worst suffering one can imagine, not everyone will have equivalent first-hand knowledge of what such suffering might be like. This was illustrated, with respect to pain, in a study on cluster headache patients, who have an acute understanding of just how intense pain can be, and rate other experienced sources of pain with a correspondingly lower intensity.Footnote ⁴⁵ One solution is to accumulate and analyze large amounts of data from people who have each experienced multiple sources of suffering in order to arrive at a calibrated scale that assigns 10/10 to the very worst suffering actually experienced. Alternatively, a more operational definition would assign 10/10 to a level of suffering at which someone would, in that moment, choose to end their life if they could, even if still higher intensities were possible.

In principle, we would want to be able to use a similar scale to track the suffering of all sentient beings. Since only humans can directly provide numerical estimates of their own suffering, the methodology for assessing non-human suffering would differ. For animals, it would largely rely on human estimates, based on behavioral observations and physiological measures.Footnote ⁴⁶ For potential non-biological entities, it would likely have to be based on calculations, using a hypothetical, validated model of consciousness and suffering, normalized to human experiences. These major differences notwithstanding, a reasonable goal is to use such metrics to inform policy, with the goal of preventing as much significant suffering as possible, with the highest priority to the most intense suffering.

Concluding thoughts

We are at a pivotal point in the history of humanity, with our world undergoing dramatic transformations through the deployment of powerful technologies. As we seek to influence our trajectory and the existence and well-being of our planet’s future inhabitants, it is essential that the ethical frameworks and metrics we employ, and the policies derived from them, be freed as much as possible from entrenched biases, whether rooted in our evolutionary past or more recent cultural norms. Just as our world has become increasingly interconnected and disciplinary boundaries less rigid, the ethical principles we apply also need to transcend specialized domains to become universally relevant across space and time.

Efforts to create bliss for some cannot reasonably come at the expense of extreme suffering for others. A rational and compassionate ethical framework calls on us to identify such suffering wherever it occurs, or could occur in the future, and to take effective steps to prevent as much of it as possible. Embedding such a framework in governance and policy-making may increase the likelihood of it retaining influence far into the future.Footnote ⁴⁷

Note: many of the arguments made in this paper are developed in more detail in The Tango of Ethics: Intuition, Rationality and the Prevention of Suffering (2023), listed in the references.