I examine different strategies involved in stating anti-theistic arguments from natural evil, and consider some theistic replies. There are, traditionally, two main types of arguments from natural evil: those that purport to deduce a contradiction between the existence of natural evil and the existence of God, and those that claim that the existence of certain types or quantities of natural evil significantly lowers the probability that theism is true. After considering peripheral replies, I state four prominent theistic rebutting strategies: skeptical (...) theism; Richard Swinburne's view that moral knowledge entails natural evil; the soul-making theodicy; and the natural law theodicy. (shrink)

In its original form, Nozick’s experience machine serves as a potent counterexample to a simplistic form of hedonism. The pleasurable life offered by the experience machine, its seems safe to say, lacks the requisite depth that many of us find necessary to lead a genuinely worthwhile life. Among other things, the experience machine offers no opportunities to establish meaningful relationships, or to engage in long-term artistic, intellectual, or political projects that survive one’s death. This intuitive objection finds some support in (...) recent research regarding the psychological effects of phenomena such as video games or social media use. After a brief discussion of these problems, I will consider a variation of the experience machine in which many of these deficits are remedied. In particular, I’ll explore the consequences of a creating a virtual world populated with strongly intelligent AIs with whom users could interact, and that could be engineered to survive the user’s death. The presence of these agents would allow for the cultivation of morally significant relationships, and the world’s long-term persistence would help ground possibilities for a meaningful, purposeful life in a way that Nozick’s original experience machine could not. While the creation of such a world is obviously beyond the scope of current technology, it represents a natural extension of the existing virtual worlds provided by current video games, and it provides a plausible “ideal case” toward which future virtual worlds will move. While this improved experience machine would seem to represent progress over Nozick’s original, I will argue that it raises a number of new problems stemming from the fact that that the world was created to provide a maximally satisfying and meaningful life for the intended user. This, in turn, raises problems analogous in some ways to the problem(s) of evil faced by theists. In particular, I will suggest that it is precisely those features that would make a world most attractive to potential users—the fact that the AIs are genuinely moral agents whose well-being the user can significantly impact—that render its creation morally problematic, since they require that the AIs inhabiting the world be subject to unnecessary suffering. I will survey the main lines of response to the traditional problem of evil, and will argue that they are irrelevant to this modified case. I will close by considering by consider what constraints on the future creation of virtual worlds, if any, might serve to allay the concerns identified in the previous discussion. I will argue that, insofar as the creation of such worlds would allow us to meet morally valuable purposes that could not be easily met otherwise, we would be unwise to prohibit it altogether. However, if our processes of creation are to be justified, they must take account of the interests of the moral agents that would come to exist as the result of our world creation. (shrink)

Abstract. Since Darwin, scholars have contemplated what our growing understanding of natural selection, combined with the fact that great suffering occurs, allows us to infer about the possibility that a benevolent God created the universe. Building on this long line of thought, I develop a model that illustrates how undesirable characteristics of the world (stylized “evils”) can influence long-run outcomes. More specifically, the model considers an evolutionary process in which each generation faces a risk from a “natural evil” (e.g., predation, (...) disease, or a natural disaster) subsequent to a basic resource allocation game. This allows both resource allocation and the natural evil to influence the number of surviving offspring. As the model shows, when the risk from the natural evil can be mitigated through the benevolent behavior of neighbors, the population may have increasing benevolence as a result of (1) greater risk from the natural evil and (2) a greater degree to which selfish individuals transfer resources to themselves in the resource allocation game. The main implication is that a world with evolutionary processes (in contrast to a world of static design) can allow two factors that have traditionally been considered “evils”—namely, the indiscriminate cruelty of the natural world and the capacity for humans to harm each other—to promote desirable long-run outcomes. (shrink)