Foundation models such as GPT-4 are fine-tuned to avoid unsafe or otherwise problematic behavior, so that, for example, they refuse to comply with requests for help with committing crimes or with producing racist text. One approach to fine-tuning, called reinforcement learning from human feedback, learns from humans' expressed preferences over multiple outputs. Another approach is constitutional AI, in which the input from humans is a list of high-level principles. But how do we deal with potentially diverging input from humans? How (...) can we aggregate the input into consistent data about "collective" preferences or otherwise use it to make collective choices about model behavior? In this paper, we argue that the field of social choice is well positioned to address these questions, and we discuss ways forward for this agenda, drawing on discussions in a recent workshop on Social Choice for AI Ethics and Safety held in Berkeley, CA, USA in December 2023. (shrink)

A-theorists and B-theorists debate whether the "Now" is metaphysically distinguished from other time slices. Analogously, one may ask whether the "I" is metaphysically distinguished from other perspectives. Few philosophers would answer the second question in the affirmative. An exception is Caspar Hare, who has devoted two papers and a book to arguing for such a positive answer. In this paper, I argue that those who answer the first question in the affirmative---A-theorists---should also answer the second question in the affirmative. This (...) is because key arguments in favor of the A-theory are more effective asarguments in favor of the resulting combined position, and key arguments against the A-theory are ineffective against the combined position. (shrink)

In the context of the Sleeping Beauty problem, it has been argued that so-called “halfers” can avoid Dutch book arguments by adopting evidential decision theory. I introduce a Dutch book for a variant of the Sleeping Beauty problem and argue that evidential decision theorists fall prey to it, whether they are halfers or thirders. The argument crucially requires that an action can provide evidence for what the agent would do not only at other decision points where she has exactly the (...) same information, but also at decision points where she has different but “symmetric” information. (shrink)

In metaphysics, there are a number of distinct but related questions about the existence of “further facts”—facts that are contingent relative to the physical structure of the universe. These include further facts about qualia, personal identity, and time. In this article I provide a sequence of examples involving computer simulations, ranging from one in which the protagonist can clearly conclude such further facts exist to one that describes our own condition. This raises the question of where along the sequence the (...) protagonist stops being able to soundly conclude that further facts exist. (shrink)

In metaphysics, there are a number of distinct but related questions about the existence of “further facts”—facts that are contingent relative to the physical structure of the universe. These include further facts about qualia, personal identity, and time. In this article I provide a sequence of examples involving computer simulations, ranging from one in which the protagonist can clearly conclude such further facts exist to one that describes our own condition. This raises the question of where along the sequence the (...) protagonist stops being able to soundly conclude that further facts exist. (shrink)

Newcomb’s problem has spawned a debate about which variant of expected utility maximisation should guide rational choice. In this paper, we provide a new argument against what is probably the most popular variant: causal decision theory. In particular, we provide two scenarios in which CDT voluntarily loses money. In the first, an agent faces a single choice and following CDT’s recommendation yields a loss of money in expectation. The second scenario extends the first to a diachronic Dutch book against CDT.

How should we update de dicto beliefs in the face of de se evidence? The Sleeping Beauty problem divides philosophers into two camps, halfers and thirders. But there is some disagreement among halfers about how their position should generalize to other examples. A full generalization is not always given; one notable exception is the Halfer Rule, under which the agent updates her uncentered beliefs based on only the uncentered part of her evidence. In this brief article, I provide a simple (...) example for which the Halfer Rule prescribes credences that, I argue, cannot be reasonably held by anyone. In particular, these credences constitute an egregious violation of the Reflection Principle. I then discuss the consequences for halfing in general. (shrink)

Significant controversy remains about what constitute correct self-locating beliefs in scenarios such as the Sleeping Beauty problem, with proponents on both the “halfer” and “thirder” sides. To attempt to settle the issue, one natural approach consists in creating decision variants of the problem, determining what actions the various candidate beliefs prescribe, and assessing whether these actions are reasonable when we step back. Dutch book arguments are a special case of this approach, but other Sleeping Beauty games have also been constructed (...) to make similar points. Building on a recent article :491–508, 2013), I show that in general we should be wary of such arguments, because unintuitive actions may result for reasons that are unrelated to the beliefs. On the other hand, I show that, when we restrict our attention to additive games, then a thirder will necessarily maximize her ex ante expected payout, but a halfer in some cases will not . I conclude that this does not necessarily settle the issue and speculate about what might. (shrink)

It is sometimes the case that one solution concept in game theory is equivalent to applying another solution concept to a modified version of the game. In such cases, does it make sense to study the former separately, or should we entirely subordinate it to the latter? The answer probably depends on the particular circumstances, and indeed the literature takes different approaches in different cases. In this article, I consider the specific example of Stackelberg mixed strategies. I argue that, even (...) though a Stackelberg mixed strategy can also be seen as a subgame perfect Nash equilibrium of a corresponding extensive-form game, there remains significant value in studying it separately. The analysis of this special case may have implications for other solution concepts. (shrink)

BackgroundIn the early stages of the COVID-19 pandemic, many health systems, including those in the UK, developed triage guidelines to manage severe shortages of ventilators. At present, there is an insufficient understanding of how the public views these guidelines, and little evidence on which features of a patient the public believe should and should not be considered in ventilator triage.MethodsTwo surveys were conducted with representative UK samples. In the first survey, 525 participants were asked in an open-ended format to provide (...) features they thought should and should not be considered in allocating ventilators for COVID-19 patients when not enough ventilators are available. In the second survey, 505 participants were presented with 30 features identified from the first study, and were asked if these features should count in favour of a patient with the feature getting a ventilator, count against the patient, or neither. Statistical tests were conducted to determine if a feature was generally considered by participants as morally relevant and whether its mean was non-neutral.ResultsIn Survey 1, the features of a patient most frequently cited as being morally relevant to determining who would receive access to ventilators were age, general health, prospect of recovery, having dependents, and the severity of COVID symptoms. The features most frequently cited as being morally irrelevant to determining who would receive access to ventilators are race, gender, economic status, religion, social status, age, sexual orientation, and career. In Survey 2, the top three features that participants thought should count in favour of receiving a ventilator were pregnancy, having a chance of dying soon, and having waited for a long time. The top three features that participants thought should count against a patient receiving a ventilator were having committed violent crimes in the past, having unnecessarily engaged in activities with a high risk of COVID-19 infection, and a low chance of survival.ConclusionsThe public generally agreed with existing UK guidelines that allocate ventilators according to medical benefits and that aim to avoid discrimination based on demographic features such as race and gender. However, many participants expressed potentially non-utilitarian concerns, such as inclining to deprioritise ventilator allocation to those who had a criminal history or who contracted the virus by needlessly engaging in high-risk activities. (shrink)

Commentary about the role of AI in bioethics for the 10th anniversary issue of AJOB: Empirical Bioethics.

This paper presents the main topics, arguments, and positions in the philosophy of AI at present (excluding ethics). Apart from the basic concepts of intelligence and computation, the main topics of ar-tificial cognition are perception, action, meaning, rational choice, free will, consciousness, and normativity. Through a better understanding of these topics, the philosophy of AI contributes to our understand-ing of the nature, prospects, and value of AI. Furthermore, these topics can be understood more deeply through the discussion of AI; so (...) we suggest that “AI Philosophy” provides a new method for philosophy. (shrink)

Theories of quantum gravity generically presuppose or predict that the reality underlying relativistic spacetimes they are describing is significantly non-spatiotemporal. On pain of empirical incoherence, approaches to quantum gravity must establish how relativistic spacetime emerges from their non-spatiotemporal structures. We argue that in order to secure this emergence, it is sufficient to establish that only those features of relativistic spacetimes functionally relevant in producing empirical evidence must be recovered. In order to complete this task, an account must be given of (...) how the more fundamental structures instantiate these functional roles. We illustrate the general idea in the context of causal set theory and loop quantum gravity, two prominent approaches to quantum gravity. (shrink)

Holism maintains that a phenomenon is more than the sum of its parts. Yet analysis--a mental process crucial to comprehension--involves dismantling the whole to grasp it piecemeal and relationally. Wading through such quandaries, Vincent Descombes guides readers to a deepened appreciation of the entity that enables understanding: the human mind.

While it is often said that robotics should aspire to reproducible and measurable results that allow benchmarking, I argue that a focus on benchmarking can be a hindrance for progress in robotics. The reason is what I call the ‘measure-target confusion’, the confusion between a measure of progress and the target of progress. Progress on a benchmark (the measure) is not identical to scientific or technological progress (the target). In the past, several academic disciplines have been led into pursuing only (...) reproducible and measurable ‘scientific’ results – robotics should be careful to follow that line because results that can be benchmarked must be specific and context-dependent, but robotics targets whole complex systems for a broad variety of contexts. While it is extremely valuable to improve benchmarks to reduce the distance be- tween measure and target, the general problem to measure progress towards more intelligent machines (the target) will not be solved by benchmarks alone; we need a balanced approach with sophisticated benchmarks, plus real-life testing, plus qualitative judgment. (shrink)

Criminal law as public law -- Criminal law as public law -- Criminal law as public law -- Mass incarceration and the theory of punishment -- Reasons to criminalize -- Formalism and pragmatism in criminal procedure -- Responsibility without resentment.

This open access monograph offers a detailed study and a systematic defense of a key intuition we typically have, as human beings, with respect to the nature of time: the intuition that the future is open, whereas the past is fixed. For example, whereas it seems unsettled whether there will be a fourth world war, it is settled that there was a first world war. -/- The book contributes, in particular, three major and original insights. First, it provides a coherent, (...) non-metaphorical, and metaphysically illuminating elucidation of the intuition. Second, it determines which model of the temporal structure of the world is most appropriate to accommodate the intuition, and settles on a specific version of the Growing Block Theory of time (GBT). Third, it puts forward a naturalistic foundation for GBT, by exploiting recent results of our best physics (viz. General Relativity, Quantum Mechanics, and Quantum Gravity). Three main challenges are addressed: the dismissal of temporal asymmetries as non-fundamental phenomena only (e.g., thermodynamic or causal phenomena), the epistemic objection against GBT, and the apparent tension between GBT and relativistic physics. It is argued that the asymmetry between the open future and the fixed past must be grounded in the temporal structure of the world, and that this is neither precluded by our epistemic device, nor by the latest approaches to Quantum Gravity (​e.g., the Causal Set Theory). -/- Aiming at reconciling time as we find it in ordinary experience and time as physics describes it, this ​innovative book ​will raise the interest of both academic researchers and ​graduate students working on the philosophy of time. More generally, it ​presents contents of interest for all metaphysicians and non-dogmatic philosophers of physics. (shrink)

In this paper, I first introduce one of the most prominent objections against the Growing Block Theory of time (GBT), the so-called ‘epistemic objection’, according to which GBT provides no way of knowing that our time is the objective present and, therefore, leads at best to absolute skepticism about our temporal location, at worst to the quasi-certainty that we are located in the objective past. Secondly, I express my dissatisfaction regarding the various traditional attempts to address this objection, especially Merricks (...) (2006), Forrest (2004) and Correia and Rosenkranz (2018). Thirdly, I show that the passage of time leads to an anti-essentialist picture of natural kinds. Finally, I develop my own solution to the epistemic objection, based on the continued existence of bare particulars. (shrink)

This chapter contains sections titled: Pluralistic Ignorance Modal Operator Epistemology Agents and Inquiry Methods Multimodal Systems Knowledge Transmissibility Knowledge Transmissibility and Public Announcement Knowledge Transmissibility and Pluralistic Ignorance References.

Western democracies experience profound conflicts that induce concerns about polarization and social cohesion. Yet although conflicts are a core feature of democracies, the forms, functions, and dynamics of democratic conflicts have rarely been subject of political theory. This paper aims at furthering our understanding of democratic conflicts. It analyzes the theory of conflict in Mouffe's agonistic pluralism, confronts it with sociological conflict theory, and presents concrete points of departure for a more comprehensive theory of democratic conflicts. The paper, thus, contributes (...) to two lines of research: (1) Regarding agonistic theories, the paper shows that agonistic pluralism fails to provide a convincing theory of conflict since it underestimates the mechanisms and effects of conflict dynamics (e.g. intergroup cohesion, intragroup conflict, domination, and escalation) and fails to account for the variety of conflict interactions. Proponents of agonistic pluralism should therefore invest more into clarifying their core concept. (2) For a general account of democratic conflicts, the paper proposes to pursue interdisciplinary research on the cognitive concepts shaping conflict interactions, the linked practices of conflict regulation, and the processual dynamics of conflicts. (shrink)

L'identité, dans les acceptions que ce terme revêt aujourd'hui, est une véritable énigme lexicale : elle désigne tout autant l'objet de contrôles sécuritaires policiers, un retour à la religion de ses parents, que, dans un guide touristique, la spécificité en voie de disparition d'un quartier. Reprenons. «Qui suis-je?», «Qui sommes-nous?», ce sont là ce qu'on appelle précisément des «questions d'identité». Nous comprenons spontanément de quoi il retourne parce que nous disposons d'un modèle : connaître l'identité de quelqu'un, c'est savoir comment (...) il s'appelle. Toutefois, lorsque la question de l'identité est posée à la première personne, mon intention n'est pas d'apprendre quels sont mes nom, prénoms et qualité, comme si je devais passer un «contrôle d'identité». Que signifie le mot dès lors qu'il est utilisé avec le possessif («mon identité», «notre identité») et qu'il ne désigne pas l'énoncé d'un état civil? Jadis le mot voulait dire exclusivement qu'il n'y a qu'une seule et même chose là où on aurait pu penser qu'il y en avait deux. Or, depuis quelques dizaines d'années, le mot a revêtu une signification autre, à savoir qu'il y a une chose ou un être qui possèdent la vertu d'être singulièrement eux-mêmes. Ainsi, que des guerres puissent éclater pour des questions qui ne relèvent pas strictement des intérêts matériels bien compris des antagonistes, nul ne saurait s'en étonner, sinon ceux qui nourrissent une conception utilitariste étriquée de l'être humain. En revanche, pourquoi est-ce le mot «identité» qui se trouve désormais chargé de signifier l'enjeu et l'objet de tels conflits? Tel est donc le point précis soulevé par Vincent Descombes : dans tout cela, que vient faire le mot «identit黠? Et que reste-t-il du concept d'identité? (shrink)

There is, in some quarters, concern about high–level machine intelligence and superintelligent AI coming up in a few decades, bringing with it significant risks for humanity. In other quarters, these issues are ignored or considered science fiction. We wanted to clarify what the distribution of opinions actually is, what probability the best experts currently assign to high–level machine intelligence coming up within a particular time–frame, which risks they see with that development, and how fast they see these developing. We thus (...) designed a brief questionnaire and distributed it to four groups of experts in 2012/2013. The median estimate of respondents was for a one in two chance that high-level machine intelligence will be developed around 2040-2050, rising to a nine in ten chance by 2075. Experts expect that systems will move on to superintelligence in less than 30 years thereafter. They estimate the chance is about one in three that this development turns out to be ‘bad’ or ‘extremely bad’ for humanity. (shrink)

Lauréat du Grand Prix de Philosophie de l'Académie française, Vincent Descombes est aujourd'hui Directeur d'études à l'EHESS et membre du Centre de recherches politiques Raymond Aron. Il est l'un des rares philosophes français à s'inscrire dans le courant de la philosophie analytique, mouvement qui place la question du langage au centre de sa réflexion. Que savoir? Comment dire? Telles sont les questions que n'aura de cesse de poser le philosophe analytique. Ces entretiens avec Philippe de Lara, maître de conférences (...) à l'université Panthéon Assas, permettront aux non-spécialistes de découvrir l'essentiel de sa pensée. Vincent Descombes revient sur son parcours aux côtés de Paul Ricoeur, sa rencontre intellectuelle avec Russell, et dégage sa méthode conceptuelle, avant d'exposer et d'approfondir les questions qu'il aborde dans ses oeuvres, celle de l'identité, les rapports entre politique et langage... (shrink)

Nous partageons, au sujet de la nature du temps, l’intuition fondamentale selon laquelle le futur est ouvert tandis que le passé est fermé. Par exemple, alors que nous pensons pouvoir influencer le cours du futur, nous savons qu’aucune de nos actions ne peut influencer le cours du passé. Cependant, bien que cette intuition soit largement partagée, identifier la nature de l’asymétrie qu’elle reflète n’est pas chose aisée. Dans cet article, j’explore différentes manières de caractériser l’asymétrie entre le ‘futur ouvert’ et (...) le ‘passé fermé’ dans le but de rendre justice à notre intuition. En particulier, je discute la question de savoir si cette asymétrie doit être caractérisée de façon sémantique, épistémique, métaphysique ou ontologique. Je conclus que, bien que plusieurs de ces caractérisations contribuent à une compréhension globale du phénomène, une caractérisation ontologique de l’asymétrie doit être préférée, car elle s’avère supérieure en termes de pouvoir explicatif, d’intelligibilité, ainsi que dans sa capacité à inclure différents sens que l’on peut prêter au concept d’ouverture. (shrink)

À partir d’une lecture intégrale de l’oeuvre d’Augustin d’Hippone (354-430), Vincent Giraud met au jour les aspects de ce qu’il appelle, dès l’introduction, une « condition herméneutique ». Il s’agit de démontrer que le « il y a », le « es gibt » propre à toute « donation » phénoménale, se situe, pour Augustin, entièrement sous condition du signe : la compréhension de l’existence humaine peut alors faire l’objet d’une reprise radicale à partir du concept de signe. Être (...) au monde, à soi, à autrui, à Dieu, c’est donc d’abord faire face à des signes. Car c’est le signe (signum) qui prévaut dès lors que la chose même (res ipsa) fait défaut. Cette tension fondamentale entre signification et manifestation peut alors être considérée comme la trame profonde qui régit toute la pensée augustinienne de l’humain. L’étude s’ordonne en deux grandes parties, qui permettent de déployer l’ensemble du spectre couvert par l’existence herméneutique telle que l’a conçue Augustin. (shrink)

Cet ouvrage propose de retracer le parcours intellectuel et spirituel de l’écrivain roumain d’expression française Emil Cioran (1911-1995). Le texte s’appuie sur une approche biographique pour saisir les enjeux philosophiques et identitaires de la vie et de l’oeuvre de Cioran. Tout d’abord son enfance imprégnée de culture allemande où liberté et insouciance cèdent vite la place à l’ennui ; sa fascination pour les thèses ultra-nationalistes pendant sa jeunesse studieuse et angoissée ; la volonté d’aveuglement que lui inspire ensuite son mépris (...) des affaires humaines ; la fuite de la Roumanie pendant la Guerre et l’affirmation de son incertitude identitaire et intellectuelle ; l’abandon finalement de son pays et de sa langue comme refus de sa propre histoire. L’ouvrage tente d’apporter un regard nouveau sur l’homme et son oeuvre pour résoudre les contradictions inhérentes à son renversement identitaire: bien que la mémoire soit au centre de sa pensée, Cioran se dirigea toujours plus vers la nécessité de l’oubli. La démarche de Vincent Piednoir démontre que cette volonté d’oubli a inscrit le philosophe dans la mémoire collective. De nombreux livres ont commenté la pensée de Cioran. L’originalité de celui-ci est de s’étendre, sans esprit de polémique mais sans tabou, sur sa jeunesse roumaine et profasciste. (shrink)

Rivers stopped or flowing backward -- Harmony, number, and others --Twentieth-Century measures.

A l'instantanéité doit succéder la réflexion, à la réaction incessante le recul stratégique. La stratégie naît il y a 2500 ans, à Athènes et en Chine avec Sun Tzu et son célèbre Art de la guerre. Art militaire au départ, méthode de pensée pour l'action, elle apparaît de plus en plus indispensable dans notre monde moderne. Pourtant, le sens en a été oublié. Aujourd'hui, les décideurs - entrepreneurs, politiques, militaires - sont accaparés par le court terme et la tactique. Atteints (...) de myopie décisionnelle, ils éprouvent des difficultés grandissantes à prendre du recul. Ils ont perdu l'habitude et le goût de la stratégie, qu'ils pratiquent de moins en moins alors qu'elle seule peut apporter des réponses à la complexité du monde et à l'accélération du temps. Il faut entrer en stratégie, utiliser celle-ci pour modeler le présent et bâtir l'avenir. Le but de cet ouvrage est de faciliter cette démarche, de l'éclairer et de la démythifier. Entrer en stratégie est un guide de survie et de succès. (shrink)