This paper introduces the Digital Consciousness Model, a probabilistic framework for systematically assessing whether AI systems are conscious by incorporating multiple leading theories of consciousness.

This paper addresses novel questions about computational functionalism raised by technical details of modern AI systems, exploring how to think about consciousness when computations for many chat responses are interwoven in a GPU cluster.

This paper explores some aspects of computer architecture that might cast doubt on the possibility of digital consciousness.

This paper argues that given the possibility of non-deterministic effects, most routine actions do not radically alter the shape of the future.

We argue that conservationist organizations should make institutional changes to better promote the welfare of individual animals.

This paper argues that interactionists about consciousness have difficulty explaining how minds influence neurons to produce behavior.

This paper argues that moral attitudes are distinguished by their functional properties in a restricted range of contexts.

This paper reflects on the significance of Savulescu's principle of procreative beneficence for the long-term future of the human species.

This paper argues for the possibility of having conscious experiences that they cannot introspect.

This paper discusses how noncognitivists might extend their accounts of moral judgments to other kinds of moral attitudes, such as moral hopes and moral intuitions.

This paper suggests that metaethical expressivists should be unbothered by a large number of brute norms and hence that they should be free to postulate their way out of Unwin's negation problem.

This review examines the book's claim that comparative and evolutionary approaches to animal behavior can settle foundational questions about the nature and distribution of consciousness.

This review critically assesses the book's account of phenomena that resist scientific explanation, with attention to implications for our understanding of consciousness and the limits of inquiry.

This post argues that recent experimental results purporting to show genuine introspective ability in LLMs warrant careful scrutiny before drawing strong conclusions.

This post argues that while LLMs are trained with next-token prediction loss, their training actually optimizes for providing good context for future predictions rather than mere short-term accuracy.

This post argues that LLMs differ from human conversationalists in fundamental ways that should caution us against treating their outputs as genuine expressions of thought or feeling.

This post explores the extent to which LLMs can anticipate future information or outcomes based on patterns in their training data.

This post argues that studies on decorticate rats suggest consciousness is either restricted to primates or distributed widely across non-mammals, rather than being common to all mammals.

This post argues that expected value maximization commits fanatical EAs to supporting bizarre projects, and uses this to argue that fanaticism is unreasonable.

This post argues that without expert consensus on consciousness established before AI becomes widespread, we risk inadvertently creating vast numbers of AI systems that appear conscious but lack genuine subjective experience.