In 1885, during initial discussions of J. C. Maxwell's celebrated thermodynamic demon, Whiting (1) observed that the demon-like velocity selection of molecules can occur in a gravitationally bound gas. Recently, a gravitational Maxwell demon has been proposed which makes use of this observation [D. P. Sheehan, J. Glick, and J. D. Means, Found. Phys. 30, 1227 (2000)]. Here we report on numerical simulations that detail its microscopic phase space structure. Results verify the previously hypothesized mechanism of its paradoxical behavior. This (...) system appears to be the only example of a fully classical mechanical Maxwell demon that has not been resolved in favor of the second law of thermodynamics. (shrink)

In a recent article in this journal our unit was accused of a number of errors of judgment in applying covert video surveillance (CVS) to infants and children suspected of life-threatening abuse. The article implied, that on moving from the Royal Brompton Hospital in London to North Staffordshire Hospital, we failed to present our work to the Research Ethics Committee (REC). We did send our protocol to the REC though we did not consider that, after a total of 16 patients (...) had been documented as being the subject of life-threatening abuse, this was research. The REC in Staffordshire agreed with us. We were also accused of undertaking work that should be pursued by the Police. We agree with this. However, unlike the Metropolitan Police the Staffordshire Police would not undertake CVS. We fail to agree that 'working together' with parents is necessarily practical or safe when trying to protect children from life-threatening abuse of this kind. (shrink)

There is a history of thinking that photographs provide a special kind of access to the objects depicted in them, beyond the access that would be provided by a painting or drawing. What is included in the photograph does not depend on the photographer’s beliefs about what is in front of the camera. This feature leads Kendall Walton to argue that photographs literally allow us to see the objects which appear in them. Current generative algorithms produce images in response to (...) users’ text prompts. Depending on the parameters, the output can resemble specific people or things which are named in the prompt. This resemblance does not depend on the user’s beliefs, so generated images are in this sense like photographs. Given this parallel, how should we think about AI-generated images? (shrink)

Given the fact that many people use Wikipedia, we should ask: Can we trust it? The empirical evidence suggests that Wikipedia articles are sometimes quite good but that they vary a great deal. As such, it is wrong to ask for a monolithic verdict on Wikipedia. Interacting with Wikipedia involves assessing where it is likely to be reliable and where not. I identify five strategies that we use to assess claims from other sources and argue that, to a greater of (...) lesser degree, Wikipedia frustrates all of them. Interacting responsibly with something like Wikipedia requires new epistemic methods and strategies. (shrink)

First published in 1966. Professor Dryer has furnished a highly illuminating account of Kant’s _Critique of Pure Reason _by unfolding its central argument. _Kant’s Solution for Verification in Metaphysics _brings out the light which Kant has to throw on central topics of philosophy. It takes its place as an indispensable guide to every student of the _Critique of Pure Reason. _.

HANDBOOK OF DEFEASIBLE REASONING AND UNCERTAINTY MANAGEMENT SYSTEMS EDITORS: DOV M. ... and A. Hunter Volume 3: Belief Change Edited by D. Dubois and H. Prade HANDBOOK OF DEFEASIBLE REASONING AND ...

We argue that artificial networks are explainable and offer a novel theory of interpretability. Two sets of conceptual questions are prominent in theoretical engagements with artificial neural networks, especially in the context of medical artificial intelligence: Are networks explainable, and if so, what does it mean to explain the output of a network? And what does it mean for a network to be interpretable? We argue that accounts of “explanation” tailored specifically to neural networks have ineffectively reinvented the wheel. In (...) response to, we show how four familiar accounts of explanation apply to neural networks as they would to any scientific phenomenon. We diagnose the confusion about explaining neural networks within the machine learning literature as an equivocation on “explainability,” “understandability” and “interpretability.” To remedy this, we distinguish between these notions, and answer by offering a theory and typology of interpretation in machine learning. Interpretation is something one does to an explanation with the aim of producing another, more understandable, explanation. As with explanation, there are various concepts and methods involved in interpretation: Total or Partial, Global or Local, and Approximative or Isomorphic. Our account of “interpretability” is consistent with uses in the machine learning literature, in keeping with the philosophy of explanation and understanding, and pays special attention to medical artificial intelligence systems. (shrink)