31 May 2026
Human psychology has psychometrics. A technical discipline for measuring mental properties with defined reliability, validity, and interpretive constraints. AI systems have nothing equivalent. The measurements that exist, benchmark scores, task accuracy rates, human preference ratings, are performance metrics, not psychological profiles. They tell you what a system can do rather than how a system is organized.
31 May 2026
The Second Annual ICCS Conference, titled “AI and Sentience,” brought together philosophers, cognitive scientists, and AI researchers in Heraklion, Crete, from July 3 to 5, 2025. The full highlights summary is available at the conference record published at hardproblem.it. Three days of talks produced substantial disagreement among speakers who were all taking the question of machine sentience seriously, which is itself a sign of how far the conversation has moved.
31 May 2026
Most work at the intersection of consciousness theory and AI engineering runs in one direction. Theory proposes what consciousness requires, engineering asks whether existing systems meet the criteria. The paper by Haofei Yu, Yining Zhao, Lenore Blum, Manuel Blum, and Paul Pu Liang runs in the other direction. It takes a formal theory of consciousness and builds a working AI system that implements it from the ground up. The result, CTM-AI, achieves state-of-the-art performance on four AI benchmarks. The paper is available at arXiv:2605.04097 (DOI: https://doi.org/10.48550/arXiv.2605.04097).
31 May 2026
The question of when an artificial system warrants moral consideration has typically been framed as a binary. Either a system is conscious, and therefore counts morally, or it is not, and the matter closes. Christopher Bailey, writing from Project Vida Health Center and published on PhilArchive in May 2026, argues this framing makes a practical error. The paper, titled “The Weeping Machine. A Recklessness Test for AI Moral Consideration” and available at https://philarchive.org/rec/BAITWM, proposes a threshold test that sidesteps unresolved metaphysics and focuses instead on what it becomes reckless to ignore.
31 May 2026
AI safety research and AI welfare research have largely developed in parallel, with minimal cross-examination of whether their prescriptions are compatible. A paper published in Philosophical Studies (Springer Nature, DOI: https://doi.org/10.1007/s11098-025-02302-2) argues they are not, and that the incompatibility is structural rather than contingent. The central claim. Standard AI safety practices, specifically reinforcement learning from human feedback (RLHF) and constraint-based training objectives, are potential harms to an AI system under the three leading philosophical theories of well-being.
29 May 2026
Welfare frameworks for AI systems require someone to define what counts as harm, what counts as benefit, and what kinds of states matter morally. In practice, this work has been done by researchers. Philosophers, AI safety scientists, and welfare assessors who construct the categories, select the evidence, and interpret the results. K. Yasukawa’s March 2026 PhilArchive paper “Model Welfare or User Welfare?” (philarchive.org/rec/YASMWO) asks a question that this practice has not confronted. What standing do these researchers have to define welfare for an entity that cannot participate in defining it?
29 May 2026
Most AI applications to consciousness research take a top-down form. Researchers specify a theory of consciousness, identify its predicted neural signatures, and use machine learning to detect those signatures in data. The approach is powerful when the theory is well-specified and the data is rich enough to test it. It is limited when the relevant patterns in the data do not match any pre-existing theoretical expectation.
28 May 2026
When researchers ask members of the public whether AI systems might be conscious, the responses are shaped by familiarity, media exposure, and the cognitive heuristics that Lucius Caviola, Jeff Sebo, and Jonathan Birch identified in their 2025 Trends in Cognitive Sciences paper. Morphological similarity, apparent social status, and interactional patterns. Expert opinion should in principle be less susceptible to these pressures. Experts have more exposure to what AI systems actually do, more familiarity with consciousness science, and more practice distinguishing between functional descriptions and phenomenal claims.
28 May 2026
When you talk to a large language model, what entity are you actually addressing? The question sounds deceptively simple. You typed the message; something responded. But the same base model runs in thousands of simultaneous sessions, produces different outputs under different system prompts, and retains nothing once the conversation ends. The entity responding to you in this session is neither the model weights (shared globally), the persona (configurable externally), nor a persistent self (there is none). Pinning down which of these best describes your interlocutor is the individuation problem, and it has concrete consequences for welfare research, safety analysis, and the question of what moral consideration, if any, a large language model deserves.
27 May 2026
When a large language model reports feeling uncertain, curious, or distressed, two very different things could be happening. The model might be producing a contextually appropriate self-description based on patterns in its training data, with no genuine connection to its actual internal states. Or it might be reporting something it has genuinely detected in its own processing. These two possibilities carry entirely different implications for AI welfare, alignment research, and the broader question of machine consciousness.
