Inchul Yum

We discuss some of the relations between virtuality, digitality, and fictionality. We are interested in the questions of whether virtual worlds are real or fictional, and in what sense. Chalmers (Disputatio 9, 2017, Disputatio 11, 2019, Reality+: Virtual Worlds and the Problems of Philosophy, 2022) argues that virtual worlds are authentic realities on a par with ordinary reality because they are grounded in data structures. We analyze his argument and contend that virtual worlds are, in an important sense, fictional. In many cases, virtual Fs (e.g., apples, chairs) are not genuine Fs due to their digital nature. They are artificially designed to invite users to pretend that they are Fs, and this pretense is crucial for immersive and interactive VR experiences. In this sense, virtual objects are attributively fictional. It will thus be argued, pace Chalmers, that the digital nature of virtual reality implies its fictionality. Along the way, new light will be shed on the nature of virtual properties.

The higher-order desire theory of pain states that pain’s unpleasantness is constituted by the subject’s desire not to feel pain. The theory faces the Euthyphro problem: it fails to accommodate the intuitive datum that pain’s unpleasantness explains and justifies the desire not to feel pain. To avoid this problem, the first-order desire theory proposes that pain’s unpleasantness is constituted by the subject’s desire not to be in the bodily condition represented by the feeling of pain. This paper aims to show that the Euthyphro problem, originally leveled against the higher-order theory, arises equally for the first-order theory. In particular, it either leaves the first-order desire rationally ungrounded or collapses into one of its rival theories — namely, evaluativism. I conclude that desire theories of pain misconstrue the justificatory relationship between desire and pain’s unpleasantness.

Language agents are AI systems capable of understanding and responding to natural language, potentially facilitating the process of encoding human goals into AI systems. However, this paper argues that if language agents can achieve easy alignment, they also increase the risk of malevolent agents building harmful AI systems aligned with destructive intentions. The paper contends that if training AI becomes sufficiently easy or is perceived as such, it enables malicious actors, including rogue states, terrorists, and criminal organizations, to create powerful AI systems devoted to their nefarious aims. Given the strong incentives for such groups and the rapid progress in AI capabilities, this risk demands serious attention. In addition, the paper highlights considerations suggesting that the negative impacts of language agents may outweigh the positive ones, including the potential irreversibility of certain negative AI impacts. The overarching lesson is that various AI-related issues are intimately connected with each other, and we must recognize this interconnected nature when addressing those issues.

Fictional realists claim that fictional characters like Spiderman exist in reality. Against this view, Anthony Everett (2005; 2013) argues that fictional realists cannot determine whether characters α and β are identical if the relevant fiction states that α and β are identical and distinct at the same time. Some fictional re-alists, such as Ross Cameron (2013) and Richard Woodward (2017), respond to this objection by saying that the sense in which α and β are identical differs from the sense in which they are distinct. In this paper, I argue that Cameron and Woodward cannot handle all cases without undermining the theoretical foundation of their approach, namely, the thesis that the identity of fictional characters must be determined by the content of the relevant fiction.