Under what conditions would an artificially intelligent system have wellbeing? Despite its clear bearing on the ethics of human interactions with artificial systems, this question has received little direct attention. Because all major theories of wellbeing hold that an individual’s welfare level is partially determined by their mental life, we begin by considering whether artificial systems have mental states. We show that a wide range of theories of mental states, when combined with leading theories of wellbeing, predict that certain existing (...) artificial systems have wellbeing. Along the way, we argue that there are good reasons to believe that artificial systems can have wellbeing even if they are not phenomenally conscious. While we do not claim to demonstrate conclusively that AI systems have wellbeing, we argue that there is a significant probability that some AI systems have or will soon have wellbeing, and that this should lead us to reassess our relationship with the intelligent systems we create. (shrink)

Many think that a specific aspect of phenomenal consciousness—valenced or affective experience—is essential to consciousness’s moral significance (valence sentientism). They hold that valenced experience is necessary for well-being, or moral status, or psychological intrinsic value (or all three). Some think that phenomenal consciousness generally is necessary for non-derivative moral significance (broad sentientism). Few think that consciousness is unnecessary for moral significance (non-necessitarianism). In this paper, I consider the prospects for these views. I first consider the prospects for valence sentientism in (...) light of Vulcans, beings who are conscious but without affect or valence of any sort. I think Vulcans pressure us to accept broad sentientism. But I argue that a consideration of explanations for broad sentientism opens up possible explanations for non-necessitarianism about the moral significance of consciousness. That is, once one leans away from valence sentientism because of Vulcans, one should feel pressure to accept a view on which consciousness is not necessary for well-being, moral status, or psychological intrinsic value. (shrink)

Which beings have moral status? I argue that moral status requires some capacity for hedonic feelings of pleasure or displeasure. David Chalmers rejects this view on the grounds that it denies moral status to Vulcans, which are defined as conscious creatures with no capacity for hedonic feelings. On his more inclusive view, all conscious beings have moral status. We agree that only conscious beings have moral status, but we disagree about how to explain this. I argue that we cannot explain (...) why consciousness is necessary for moral status without denying that it is sufficient for moral status. Hence, we cannot exclude unconscious zombies without excluding conscious Vulcans too. (shrink)

If moral status depends on the capacity for consciousness, what kind of consciousness matters exactly? Two popular answers are that any kind of consciousness matters (Broad Sentientism), and that what matters is the capacity for pleasure and suffering (Narrow Sentientism). I argue that the broad answer is too broad, while the narrow answer is likely too narrow, as Chalmers has recently argued by appeal to ‘philosophical Vulcans’. I defend a middle position, Motivational Sentientism, on which what matters is motivating consciousness: (...) any kind of consciousness which presents its subject with reasons for action. (shrink)

There is a strong intuition that a zombie’s life is never good or bad for the zombie. What explains this? In this paper, I consider five possible explanations of the intuition that a zombie’s life is never worth living, plus the option of rejecting the intuition. I point out the considerable costs of each option, though making clear which option strikes me as least problematic.

The aesthetic tradition has identified as paradigmatically sublime such objects as imposing mountains and intense storms, as well as monumental art. But the tradition also acknowledges less paradigmatic cases, including sometimes mathematical structures or abstract concepts. In this paper, we argue that there is also a case for considering phenomenal consciousness—the experiential quality of subjective awareness—as a sublime phenomenon. One appreciates this, we argue, when one is struck by (fitting) awe upon contemplating (a) the perplexing existence of something like phenomenal (...) consciousness in an otherwise completely material universe and (b) the role of consciousness in injecting meaning and value in an otherwise brutally factual reality. (shrink)

Most philosophers think that phenomenal consciousness underlies, or at any rate makes a large contribution, to moral considerability. This paper argues that many such accounts invoke question-begging arguments. Moreover, they’re unable to explain apparent differences in moral status across and within different species. In the light of these problems, I argue that we ought to take very seriously a view according to which moral considerability is grounded in functional properties. Phenomenal consciousness may be sufficient for having a moral value, but (...) it may not be necessary, and it may contribute relatively little to our overall considerability. (shrink)

I show that there are good reasons to think that some individuals without any capacity for consciousness should be counted as welfare subjects, assuming that desire-fulfilment is a welfare good and that any individuals who can accrue welfare goods are welfare subjects. While other philosophers have argued for similar conclusions, I show that they have done so by relying on a simplistic understanding of the desire-fulfilment theory. My argument is intended to be sensitive to the complexities and nuances of contemporary (...) developments of the theory, while avoiding highly counter-intuitive implications of previous arguments for the same conclusion. (shrink)

Prudential value is the kind of value that something has when it is good for someone, in the sense that is conceptually tied to welfare or well‐being. Impersonal value is the kind of value that something has when it is good simply, absolutely, or “from the point of view of the universe.” According to the Moorean position on prudential value, the concept of prudential value can be analyzed in terms of that of impersonal value and is unintelligible if it is (...) not so analyzed. I answer some recent arguments for the Moorean position due to Kris McDaniel and Thomas Hurka, and I make a more comprehensive case than has thus far been made for rejecting that position. (shrink)

In this critical discussion, I evaluate David Chalmers’ position on the moral grounding question from his (2022) Reality +. The moral grounding question asks: in virtue of what does an entity x have moral standing? Chalmers argues for the claim that phenomenal consciousness is a necessary condition for moral standing. After a brief introduction to his book, I evaluate his position on the moral grounding question from the perspective of access consciousness as opposed to phenomenal consciousness, as well as the (...) Jain doctrine of non-violence, and the differentiation of creatures in terms of their sense capacities. (shrink)

We normally assume that there’s just one conscious individual per animal. Some question this, suggesting that there may be nonhuman taxonomic groups whose normal, adult members house more than one conscious subject. Call this the multitudes view (“MV)”. Our aim is methodological: we hope to understand how we might assess whether MV is true. To that end, we distinguish two strategies for counting conscious subjects: the duplication strategy and the mind-first strategy. We use human split-brain patients and octopuses to illustrate (...) the prospects and limitations of each approach, as some have argued that both sets of individuals contain multitudes. We’ll contend that while the duplication strategy is applicable to split-brain patients, it isn’t apt for octopuses. The mind-first strategy fairs better here, but there’s a catch: it requires, in a way the duplication strategy does not, numerous contentious assumptions about the nature of mentality, the individuation of minds, and consciousness to simply get off the ground. These difficulties clarify just what it would take to establish MV. (shrink)

Intuitively, there is a close link between moral status and phenomenal consciousness. Taking the link seriously can serve as the basis of a proposal that appears to have a surprising number of theoretical benefits. This proposal is the moral option, according to which moral status is partly determinative of phenomenal consciousness, and phenomenal consciousness is sufficient for possession of a moral property I refer to as “moral status.” I argue for this view on the basis of its ability to shed (...) light on the distribution problem of determining which systems are phenomenally conscious. I explicate it by drawing on a theory of attributions of phenomenal consciousness, and considering its relationship to positions on the metaphysics of consciousness. I defend it against a series of objections, including three based on previous theories of the link between phenomenal consciousness and moral status. The moral option should, in my view, be a serious contender among theories of phenomenal consciousness. (shrink)

Birch (2022a) proposes a theory-light methodology for studying whether invertebrates have the capacity for (phenomenal) consciousness. The success of any such methodology turns on the positive markers it proposes, and whether they are genuinely ecumenical. After providing an account of what it is for a marker to be ecumenical, it is argued that one of the more influential set of markers offered – unlimited associative learning – clearly counts as positive evidence for consciousness on only a small handful of theories, (...) rendering Birch???s theory-light approach tacitly theory-heavy. What emerges is a broader lesson concerning what exactly is required for the theory-light framework to be successful. (shrink)

Welfare subjectivists face a dilemma. On the one hand, traditional subjectivist theories—such as the desire-fulfillment theory—are too permissive to account for the well-being of typical mature human beings. On the other hand, more “refined” theories—such as the life-satisfaction theory—are too restrictive to account for the well-being of various welfare subjects, including newborns, those with profound cognitive impairments, or non-human animals. This paper examines a class of welfare subjectivism that addresses this dilemma with sensitivity to the diversity in welfare subjects. First, (...) the most-sophisticated-attitude view (MSA) is introduced. MSA holds that an object, $$x$$ x, is good for a subject, $$S$$ S, in proportion to the strength of $$S$$ S ’s pro-attitude towards $$x$$ x if and only if the pro-attitude at issue is $$S$$ S ’s most sophisticated type. Typically, the well-being of typical mature human beings is assessed in terms of one’s authentic whole-life satisfaction, whereas that of human newborns is assessed in terms of something less sophisticated such as pleasure. MSA offers the rationale for this difference based on an underexplored version of perfectionism: procedural perfectionism. Next, provided that MSA may involve an implausibly strong claim, this paper examines two moderate variations of MSA that accept the partial relevance of less sophisticated types of valenced attitude. Finally, it is illustrated how MSA and its variations have plausible implications regarding the well-being of enhanced or dis-enhanced people. (shrink)

A central question in AI ethics concerns the moral status of robots. This article argues against the idea that they have moral status. It proceeds by defending the assumption that consciousness is necessary for welfare subjectivity. Since robots most likely lack consciousness, and welfare subjectivity is necessary for moral status, it follows that robots lack moral status. The assumption that consciousness is necessary for welfare subjectivity appears to be in tension with certain widely accepted theories of wellbeing, especially versions of (...) Desire Satisfaction Theory and Objective List Theory. However, instead of elevating non-conscious robots to welfare subjects, this tension should lead us to reject versions of these theories that have this implausible implication. (shrink)