This paper engages with two compelling challenges to physicalism, each designed to show that the nature of experience is elusive from the standpoint of physical science. It is argued that the physicalist is ultimately well placed to meet both challenges.

David Lewis has recently deployed a contextualist strategy for defending ordinary claims to know.1 In this paper, I wish to extend that strategy to ordinary claims about freedom.2 The result is a species of compatibilism that, while foreign to current debates, has a good deal going for it.

This paper presents an argument that certain AI safety measures, rather than mitigating existential risk, may instead exacerbate it. Under certain key assumptions - the inevitability of AI failure, the expected correlation between an AI system's power at the point of failure and the severity of the resulting harm, and the tendency of safety measures to enable AI systems to become more powerful before failing - safety efforts have negative expected utility. The paper examines three response strategies: Optimism, Mitigation, and (...) Holism. Each faces challenges stemming from intrinsic features of the AI safety landscape that we term Bottlenecking, the Perfection Barrier, and Equilibrium Fluctuation. The surprising robustness of the argument forces a re-examination of core assumptions around AI safety and points to several avenues for further research. (shrink)