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)

The Oldest Living Things in the World is an epic journey through time and space. Over the past decade, artist Rachel Sussman has researched, worked with biologists, and traveled the world to photograph continuously living organisms that are 2,000 years old and older. Spanning from Antarctica to Greenland, the Mojave Desert to the Australian Outback, the result is a stunning and unique visual collection of ancient organisms unlike anything that has been created in the arts or sciences before, insightfully (...) and accessibly narrated by Sussman along the way. Her work is both timeless and timely, and spans disciplines, continents, and millennia. It is underscored by an innate environmentalism and driven by Sussman’s relentless curiosity. She begins at “year zero,” and looks back from there, photographing the past in the present. These ancient individuals live on every continent and range from Greenlandic lichens that grow only one centimeter a century, to unique desert shrubs in Africa and South America, a predatory fungus in Oregon, Caribbean brain coral, to an 80,000-year-old colony of aspen in Utah. Sussman journeyed to Antarctica to photograph 5,500-year-old moss; Australia for stromatolites, primeval organisms tied to the oxygenation of the planet and the beginnings of life on Earth; and to Tasmania to capture a 43,600-year-old self-propagating shrub that’s the last individual of its kind. Her portraits reveal the living history of our planet—and what we stand to lose in the future. These ancient survivors have weathered millennia in some of the world’s most extreme environments, yet climate change and human encroachment have put many of them in danger. Two of her subjects have already met with untimely deaths by human hands. Alongside the photographs, Sussman relays fascinating – and sometimes harrowing – tales of her global adventures tracking down her subjects and shares insights from the scientists who research them. The oldest living things in the world are a record and celebration of the past, a call to action in the present, and a barometer of our future. (shrink)

Pt. I. The apparent good. Evaluative cognition -- Perceiving the good -- Phantasia and the apparent good -- pt. II. The apparent good and non-rational motivation. Passions and the apparent good -- Akrasia and the apparent good -- pt. III. The apparent good and rational motivation. Phantasia and deliberation -- Happiness, virtue, and the apparent good -- Practical induction -- Conclusion : Aristotle's practical empiricism.

Moral philosophy is the study of what morality is and what it requires of us. As Socrates said, it's about "how we ought to live"-and why. It would be helpful if we could begin with a simple, uncontroversial definition of what morality is. Unfortunately, we cannot. There are many rival theories, each expounding a different conception of what it means to live morally, and any definition that goes beyond Socrates's simple formula-tion is bound to offend at least one of them. (...) This should make us cautious, but it need not paralyze us. In this chapter, I will describe the "minimum conception" of morality. (shrink)

Plato's Epistemology presents an original interpretation of one of the central topics in Plato's work: epistemology. Moss argues, against the grain of much modern scholarship, that Plato's epistemology is radically different from our own.

did plato and aristotle have anything to say about belief? The answer to this question might seem blindingly obvious: of course they did. Plato distinguishes belief from knowledge in the Meno, Republic, and Theaetetus, and Aristotle does so in the Posterior Analytics. Plato distinguishes belief from perception in the Theaetetus, and Aristotle does so in the De anima. They talk about the distinction between true and false beliefs, and the ways in which belief can mislead and the ways in which (...) it can steer us aright. Indeed, they make belief a central component of their epistemologies.The view underlying these claims—one so widespread these days as to remain largely unquestioned—is that when Plato and Aristotle talk... (shrink)

Many tasks in statistical and causal inference can be construed as problems of entailment in a suitable formal language. We ask whether those problems are more difficult, from a computational perspective, for causal probabilistic languages than for pure probabilistic (or “associational”) languages. Despite several senses in which causal reasoning is indeed more complex—both expressively and inferentially—we show that causal entailment (or satisfiability) problems can be systematically and robustly reduced to purely probabilistic problems. Thus there is no jump in computational complexity. (...) Along the way we answer several open problems concerning the complexity of well-known probability logics, in particular demonstrating the ${\exists \mathbb {R}}$ -completeness of a polynomial probability calculus, as well as a seemingly much simpler system, the logic of comparative conditional probability. (shrink)

This paper argues that several leading theories of subjunctive conditionals are incompatible with ordinary intuitions about what credences we ought to have in subjunctive conditionals. In short, our theory of subjunctives should intuitively display semantic humility, i.e. our semantic theory should deliver the truth conditions of sentences without pronouncing on whether those conditions actually obtain. In addition to describing intuitions about subjunctive conditionals, I argue that we can derive these ordinary intuitions from justified premises, and I answer a possible worry (...) for my derivation by refuting a subjunctive triviality result modeled on (Lewis 1976). I conclude that the debate over the correct theory of subjunctive conditionals requires settling meta-philosophical questions about the relative value of various virtues of first-order theories of subjunctive conditionals. (shrink)

Has the renewal of interest in the ostensible agency of living beings signaled an advance from a merely heuristic Kantian sense of purposiveness to an unequivocally, empirically grounded research program or are there as yet hidden tensions or contradictions in, for example, the organizational autonomy approach to natural agency? Can normativity be found to be immanent in nature but only beginning with the living cell or must a thoroughgoing naturalism find the seeds of normativity immanent throughout abiotic as well as (...) biotic nature? Beginning with a brief exposition of Kant´s influential treatment and recommendation for how to methodologically combine what he took to be the inevitable epistemological limit to explaining the origins of ostensible biotic purposefulness with the legitimate intentions of scientific research and explanation, this essay will critically engage with three recent essays that attempt to grapple with the preceding questions. Having putatively raised questions about the consistency and adequacy of each of the individual positions, the essay will attempt to move synthetically, drawing upon aspects of all three contributions, in the direction of a “cooperativity theoretic” approach to incipient natural normativity and agency. (shrink)

Traditional procedures for rational updating fail when it comes to self-locating opinions, such as your credences about where you are and what time it is. This paper develops an updating procedure for rational agents with self-locating beliefs. In short, I argue that rational updating can be factored into two steps. The first step uses information you recall from your previous self to form a hypothetical credence distribution, and the second step changes this hypothetical distribution to reflect information you have genuinely (...) learned as time has passed. While the second step resembles traditional procedures of updating by conditionalization, the first is best understood by analogy to traditional models of how agents transmit self-locating opinions through ordinary interpersonal communication. (shrink)

de Anima III.10 characterizes akrasia as a conflict between phantasia (“imagination”) on one side and rational cognition on the other: the akratic agent is torn between an appetite for what appears good to her phantasia and a rational desire for what her intellect believes good. This entails that akrasia is parallel to certain cases of perceptual illusion. Drawing on Aristotle's discussion of such cases in the de Anima and de Insomniis , I use this parallel to illuminate the difficult discussion (...) of akrasia in Nicomachean Ethics VII.3, arguing that its account of akrasia as involving ignorance is compatible with, and in fact crucially supplements, the more straightforward account we find elsewhere in the corpus of akrasia as a struggle between desires. (shrink)

A joke is amusing if and only if it’s fitting to be amused by it; an act is regrettable if and only if it’s fitting to regret it. Many philosophers accept these biconditionals and hold that analogous ones obtain between a wide range of additional evaluative properties and the fittingness of corresponding responses. Call these the _fit–value biconditionals_. The biconditionals give us a systematic way of recognizing the role of fit in our ethical practices; they also serve as the bedrock (...) of various metaethical projects, such as fitting-attitude analysis of value and the ‘fittingness first’ approach. Yet despite the importance of the biconditionals, there is very little discussion of their proper interpretation. This paper argues that any plausible interpretation of the fit_–_value biconditionals must disarm several kinds of apparent counterexample. For instance, that an achievement is pride-worthy doesn’t imply it is fitting for me to take pride in it because the achievement might not be mine or that of anyone close to me; that a joke is amusing doesn’t imply it is fitting for me to be amused by it for six straight months; and that a person is loveable doesn’t imply it is fitting for me to love him romantically because that person might be my sibling. We consider possible responses to such counterexamples and develop what we consider the most promising interpretation of the biconditionals. The upshot is that certain widespread assumptions about fit and its relation to value and reasons should be reconsidered. (shrink)

Lucidly and persuasively written, this book will be an invaluable resource for students and scholars seeking to understand Irigaray's original contribution to philosophical and feminist thought.

Stoics and Sceptics distinguish belief (doxa) from a representationally and functionally similar but sub-doxastic state: passive yielding to appearance. Belief requires active assent to appearances, that is, affirmation of the appearances as true. I trace the roots of this view to Plato's accounts of doxa in the Republic and Theaetetus. In the Republic, eikasia and pistis (imaging and conviction) are distinguished by their objects, appearances versus ordinary objects; in the Theaetetus, perception and doxa are distinguished by their objects, proper perceptibles (...) versus ‘commons’, including being. But underlying these ontological distinctions is a psychological one: the lower mental states are confined to their lower objects because they are passive; the higher mental states have access to higher objects because they result from questioning appearances and making active affirmations about how things are. This doctrine of doxa anticipates both the Hellenistic one and modern accounts of belief as ‘aiming at truth’; it also shows Plato's views of doxa to have more in their favour philosophically and to cohere better with one another than generally thought. (shrink)

ABSTRACT “Meister Eckhart’s Mysticism in Comparison with Zen Buddhism” originally appeared as the concluding section of Ueda Shizuteru’s first book, Die Gottesgeburt in der Seele und der Durchbruch zur Gottheit: Die mystische Anthropologie Meister Eckharts und ihre Konfrontation mit der Mystik des Zen-Buddhismus. It was first published in 1965 as an expanded version of Ueda’s doctoral dissertation, which was written under the supervision of Ernst Benz at the University of Marburg. Ueda’s careful analysis not only illuminates important points of affinity (...) between Eckhart and Zen, but also explains why Eckhart’s mysticism cannot be simply interchanged with Zen Buddhism. To allow the reader to read for themselves, I kept all interpretative notes to a minimum. The text is translated from the most recent edition of the German text, published in 2018 by Verlag Karl Alber, and edited by Wolf Burbat. The German pagination is interpolated in bold throughout the text. (shrink)

This chapter contains sections titled: Notes.

Aristotle's De Anima discusses the psychological causes of what he calls locomotion – i.e, roughly, purpose-driven behavior. One cause is desire. The other is cognition, which falls into two kinds: thought (nous) and imagination (phantasia). Aristotle’s discussion is dense and confusing, but I argue that we can extract from it an account that is coherent, compelling, and that in many ways closely anticipates modern psychological theories, in particular Dual Processing theory. Animals and humans are driven to pursue objects that attract (...) us. Objects take on that power when we cognize them as valuable. If we rely on imagistic, automatic, uncontrolled processing mechanisms – Aristotle’s phantasia, which closely anticipates the modern notion of Type 1 processing – our resulting desires and actions will be impulsive. If we rely instead on rational, critical, deliberative capacities – Aristotle’s thought, which closely anticipates the modern notion of Type 2 processing – our resulting desires and actions will be reflective. Animals are capable only of the first kind of behavior; the human psyche is constituted of an animal psyche united with an intellectual one, so we are capable of both. (shrink)

This paper provides a background to, detailed exploration, and then critique of, the influential notion of economic performativity. It begins with a broad sweep of the theoretical developments in economic sociology in the years before the advance of the performativity program. In doing so it outlines the theoretical quandary that performativity sought to move beyond. Having set the scene, it then looks at the performativity thesis in detail, explaining how it seeks to do away with modern ontological binaries like the (...) object/subject dichotomy. Once the performativity thesis has been unpacked, it is critiqued. Leading to the conclusion, that while there have been some benefits to the performative approach, especially in research looking at finance, the performativity thesis ultimately fails in its attempt to move beyond the ontology of “the moderns”, and does little to shift the debates in economic sociology that predated its popularity. (shrink)

The Dialectics of Absolute Nothingness examines the influence of German philosophical traditions on the development of the Kyoto School. Contributors explore the Kyoto School's engagement with Western thought, highlighting the centrality of German philosophy while also showing the many ways the Kyoto School critiques the philosophical traditions it incorporates.

Anthology of readings in moral philosophy.

This article draws on the philosophical work on dialogic rationality offered by Charles Taylor as well as qualitative studies of dialogues between politically opposed college students to argue that these conversations succeed as tools of democracy precisely because they fail as interventions. That is, the democratic strength of such dialogue is the way in which it is unreliable as a means of producing particular outcomes. Students whose political views eventually shifted partly in response to dialogue understood this not as a (...) process of changing their commitments, but rather of finding a better expression for the commitments they already held. But the interviews show, too, just how rare it is for such shifts to occur at all, at least for these students in a period of political polarization and mutual distrust. The interviews suggest as well that the appeal to already shared ground is not the whole story of what prompts people to change their minds in these kinds of conversations. Rather the asking of direct but respectful questions was a crucial ingredient. Finally, dialogue did not independently cause the most profound changes in political views. Instead, it was the direction of students' lives that over time shifted their outlook. Their dialogue experience then gained meaning in retrospect, as the beginning of a process of self-questioning that was brought to fruition only later, as a first time that they had heard alternative possibilities for how value commitments can be expressed through politics. (shrink)

The postwar welfare state provides social insurance against economic, health, and related risks in an uncertain world. Because everyone can envision themselves to be among the unfortunate, social insurance fuses self-interest and solidarism in a normative principle Friedman (2020) calls probabilistic justice. But there is a competing principle of status defense, where the aim is to erect boundaries between socioeconomic strata and discourage cross-class mobility. We argue that this principle dominates when inequality is high and uncertainty low. The current moment (...) is one of high inequality and high uncertainty, which results in intense status anxiety, yet does not rule out solidaristic solutions. Our contributions are to diagnose the causes of our current malaise, and to theorize the normative bases for the political choice facing contemporary western democracies. (shrink)

In Talking about, Unnsteinsson defends an intentionalist theory of reference by arguing that confused referential intentions degrade reference. Central to this project is a “belief model” of both identity confusion and unconfused thought. By appealing to a well‐known argument from Campbell, I argue that this belief model falls short, because it fails to explain the inferential behavior it promises to explain. Campbell's argument has been central in the contemporary literature on Frege's puzzle, but Unnsteinsson's account of confusion provides an opportunity (...) for more clarity about how the argument is best interpreted, and what it shows. (shrink)

Trois parties: la démocratie athénienne; les institutions oligarchiques; les fédérations de cités. [SDM].

The so-called ‘mental files theory’ in the philosophy of mind stems from an analogy comparing object-concepts to ‘files’, and the mind to a ‘filing system’. Though this analogy appears in philosophy of mind and language from the 1970s onward, it remains unclear to many how it should be interpreted. The central commitments of the mental files theory therefore also remain unclear. Based on influential uses of the file analogy within philosophy, I elaborate three central explanatory roles for mental files. Next, (...) I outline several common criticisms of the file picture, which have been a source of resistance to the view. Finally, I outline several interpretations of the theory, thus highlighting that the best interpretation of the file-theory's central analogy remains a live issue. (shrink)