How should we make choices when we know so little about our futures? L. A. Paul argues that we must view life decisions as choices to make discoveries about the nature of experience. Her account of transformative experience holds that part of the value of living authentically is to experience our lives and preferences in whatever ways they evolve.

If updating with E has the same result across all epistemically possible worlds, then the agent has no uncertainty as to the behavior of the update, and we may call it a transparent update. If an agent is uncertain about the behavior of an update, we may call it opaque. In order to model the uncertainty an agent has about the result of an update, the same update must behave differently across different possible worlds. In this paper, I study opaque (...) updates using a simple system of dynamic epistemic logic suitably modified for that purpose. The paper highlights the connection between opaque updates and the dynamic-epistemic principles Perfect-Recall and No-Miracles. I argue that opaque updates are central to contemporary discussions in epistemology, in particular to externalist theories of knowledge and to the related problem of epistemic bootstrapping, or easy knowledge. Opaque updates allow us to explicitly investigate a dynamic form of uncertainty, using simple and precise logical tools. (shrink)

Conditionalization is one of the central norms of Bayesian epistemology. But there are a number of competing formulations, and a number of arguments that purport to establish it. In this paper, I explore which formulations of the norm are supported by which arguments. In their standard formulations, each of the arguments I consider here depends on the same assumption, which I call Deterministic Updating. I will investigate whether it is possible to amend these arguments so that they no longer depend (...) on it. As I show, whether this is possible depends on the formulation of the norm under consideration. (shrink)

In this article, I cast doubt on an apparent truism, namely, that if evidence is available for gathering and use at a negligible cost, then it’s always instrumentally rational for us to gather that evidence and use it for making decisions. Call this ‘value of information’ (VOI). I show that VOI conflicts with two other plausible theses. The first is the view that an agent’s evidence can entail non-trivial propositions about the external world. The second is the view that epistemic (...) rationality requires us to update our credences by conditionalization. These two theses, given some plausible assumptions, make room for rationally biased inquiries where VOI fails. I go on to argue that this is bad news for defenders of VOI. (shrink)

Accuracy-first epistemology aims to show that the norms of epistemic rationality can be derived from the effective pursuit of accuracy. This paper explores the prospects within accuracy-first epistemology for vindicating “modesty”: the thesis that ideal rationality permits uncertainty about one’s own rationality. I argue that accuracy-first epistemology faces serious challenges in accommodating three forms of modesty: uncertainty about what priors are rational, uncertainty about whether one’s update policy is rational, and uncertainty about what one’s evidence is. I argue that the (...) problem stems from the representation of epistemic decision problems. The appropriate representation of decision problems, and corresponding decision rules, for (diachronic) update policies should be a generalization of decision problems and decision rules for (synchronic) coherence. I argue that extant accounts build in conflicting assumptions about which kinds of information about the believer should be used to structure epistemic decision problems. In particular, extant accounts of update build in a form of epistemic consequentialism. Related forms of epistemic consequentialism have been shown to generate problems for accuracy-first epistemology’s purported justifications of probabilism, conditionalization, and the principal principle. These results are vindicated only with nonconsequentialist epistemic decision theories. I close with suggestive examples of how, with a fully nonconsequentialist representation of epistemic decision problems, accuracy-first epistemology can allow for rational modesty. (shrink)

It is natural to think that there is something epistemically objectionable about avoiding evidence, at least in ideal cases. We argue that this thought is inconsistent with a kind of risk-avoidance...

This paper addresses the question of whether gathering additional evidence is always rationally required, both from the point of view of instrumental rationality and of epistemic rationality. It is shown that in certain situations, it is not instrumentally rational to look for more evidence before making a decision. These are situations in which the risk of “misleading” evidence – a concept that has both instrumental and epistemic senses – is not offset by the gains from the possibility of non-misleading evidence. (...) These situations also undermine a particular argument for the claim that it is epistemically rational to look for more evidence, though it is argued that the relationship between instrumental and epistemic rationality makes such arguments flawed to begin with. Furthermore, these situations show us that our epistemic and our practical goals sometimes point us in different directions, not merely because of our limited resources or because our desires are sometimes best served by being in particular epistemic states, but because of the nature of rational action. (shrink)

Diachronic Dutch book arguments seem to support both conditionalization and Bas van Fraassen's Reflection principle. But the Reflection principle is vulnerable to numerous counterexamples. This essay addresses two questions: first, under what circumstances should an agent obey Reflection, and second, should the counterexamples to Reflection make us doubt the Dutch book for conditionalization? In response to the first question, this essay formulates a new "Qualified Reflection" principle, which states that an agent should obey Reflection only if he or she is (...) certain that he or she will conditionalize on veridical evidence in the future. Qualified Reflection follows from the probability calculus together with a few idealizing assumptions. The essay then formulates a "Distorted Reflection" principle that approximates Reflection even in cases where the agent is not quite certain that he or she will conditionalize on veridical evidence. In response to the second question, the essay argues that contrary to a common misconception, not all Dutch books dramatize incoherence—some dramatize a less blameworthy sort of epistemic frailty that the essay calls "self-doubt." The distinction between Dutch books that dramatize incoherence and those that dramatize self-doubt cross-cuts the distinction between synchronic and diachronic Dutch books. The essay explains why the Dutch book for conditionalization reveals true incoherence, whereas the Dutch book for Reflection reveals only self-doubt. (shrink)

Risk-weighted expected utility theory permits preferences which violate the Sure-Thing Principle. But preferences that violate the STP can lead to bad decisions in sequential choice problems. In particular, they can lead decision-makers to adopt a strategy that is dominated – i.e. a strategy such that some available alternative leads to a better outcome in every possible state of the world.

This paper considers a puzzling conflict between two positions that are each compelling: it is irrational for an agent to pay to avoid `free' evidence before making a decision, and rational agents may have imprecise beliefs and/or desires. Indeed, we show that Good's theorem concerning the invariable choice-worthiness of free evidence does not generalise to the imprecise realm, given the plausible existing decision theories for handling imprecision. A key ingredient in the analysis, and a potential source of controversy, is the (...) general approach taken for resolving sequential decision problems { we make explicit what the key alternatives are and defend our own approach. Furthermore, we endorse a resolution of the aforementioned puzzle { we privilege decision theories that merely permit avoiding free evidence over decision theories for which avoiding free evidence is uniquely admissible. Finally, we situate this particular result about free evidence within the broader `dynamic-coherence' literature. (shrink)

How should we respond to evidence when our evidence indicates that we are rationally impaired? I will defend a novel answer based on the analogy between self-doubt and memory loss. To believe that one is now impaired and previously was not is to believe that one’s epistemic position has deteriorated. Memory loss is also a form of epistemic deterioration. I argue that agents who suffer from epistemic deterioration should return to the priors they had at an earlier time. I develop (...) this argument regarding memory loss then extend it to cases of self-doubt. (shrink)

Good’s theorem is the apparent platitude that it is always rational to ‘look before you leap’: to gather information before making a decision when doing so is free. We argue that Good’s theorem is not platitudinous and may be false. And we argue that the correct advice is rather to ‘make your act depend on the answer to a question’. Looking before you leap is rational when, but only when, it is a way to do this.

Richard Pettigrew offers an extended investigation into a particular way of justifying the rational principles that govern our credences. The main principles that he justifies are the central tenets of Bayesian epistemology, though many other related principles are discussed along the way. Pettigrew looks to decision theory in order to ground his argument. He treats an agent's credences as if they were a choice she makes between different options, gives an account of the purely epistemic utility enjoyed by different sets (...) of credences, and then appeals to the principles of decision theory to show that, when epistemic utility is measured in this way, the credences that violate the principles listed above are ruled out as irrational. The account of epistemic utility set out here is the veritist's: the sole fundamental source of epistemic utility for credences is their accuracy. Thus, Pettigrew conducts an investigation in the version of epistemic utility theory known as accuracy-first epistemology. (shrink)

Selected by the Economist as one of the best books of 2020. -/- Privacy Is Power argues that people should protect their personal data because privacy is a kind of power. If we give too much of our data to corporations, the wealthy will rule. If we give too much personal data to governments, we risk sliding into authoritarianism. For democracy to be strong, the bulk of power needs to be with the citizenry, and whoever has the data will have (...) the power. Privacy, I argue, is not a personal preference; it is a political concern. I also argue that personal data is a toxic asset, and should be regulated as if it were a toxic substance, similar to asbestos. I call for a complete ban to the trade in personal data. The book is at once philosophical and political, and extremely practical and accessible. It discusses liberal democracy, equality, justice, and autonomy. It includes a chapter for policymakers, and one that addresses what ordinary citizens can do to protect privacy. Finally, the book covers the tension between privacy and public health in the context of covid19. -/- . (shrink)

Artificial Intelligence: A Modern Approach, 3e offers the most comprehensive, up-to-date introduction to the theory and practice of artificial intelligence. Number one in its field, this textbook is ideal for one or two-semester, undergraduate or graduate-level courses in Artificial Intelligence. Dr. Peter Norvig, contributing Artificial Intelligence author and Professor Sebastian Thrun, a Pearson author are offering a free online course at Stanford University on artificial intelligence. According to an article in The New York Times, the course on artificial intelligence is (...) “one of three being offered experimentally by the Stanford computer science department to extend technology knowledge and skills beyond this elite campus to the entire world.” One of the other two courses, an introduction to database software, is being taught by Pearson author Dr. Jennifer Widom. Artificial Intelligence: A Modern Approach, 3e is available to purchase as an eText for your Kindle™, NOOK™, and the iPhone®/iPad®. (shrink)

There has been a movement aiming to teach agents about their privilege by making the information about their privilege as costless as possible. However, some argue that in risk-sensitive frameworks, such as Lara Buchak’s (2013), it can be rational for privileged agents to shield themselves from learning about their privilege, even if the information is costless and relevant. This threatens the efficacy of these information-access efforts in alleviating the problem of elite-group ignorance. In response, I show that even within the (...) same framework, in this case David Kinney and Liam Kofi Bright’s (2021), the rationality of this information avoidance rests on shaky ground in practice. In this framework, whether an agent should avoid information depends on the precise details of (1) how relevant they expect the information to be, (2) their priors about the value of various options, and (3) their risk attitudes. The model suggests that the rationality of elite-group ignorance is a function of structural factors that are pervasive but nonetheless not insurmountable, thus offering a way out of pessimism about elite group education. (shrink)

Are predictions about how one will freely and intentionally behave in the future ever relevant to how one ought to behave? There is good reason to think they are. As imperfect agents, we have responsibilities of self-management, which seem to require that we take account of the predictable ways we're liable to go wrong. I defend this conclusion against certain objections to the effect that incorporating predictions concerning one's voluntary conduct into one's practical reasoning amounts to evading responsibility for that (...) conduct. There is, however, some truth to this sort of objection. To understand the legitimate role of self-prediction in practical reasoning, we need to distinguish instances of coping responsibly with an anticipated failure to behave as one ought, on the one hand, from mere acquiescence in one's flaws, on the other. I argue that, to draw this distinction, we must recognize certain limits on the use of self-prediction as a ground of choice. (shrink)

The paper deals with credible and relevantinformation flow in dialogs: How useful is it for areceiver to get some information, how useful is it fora sender to give this information, and how much credibleinformation can we expect to flow between sender andreceiver? What is the relation between semantics andpragmatics? These Gricean questions will be addressedfrom a decision and game-theoretical point of view.

The paper deals with credible and relevantinformation flow in dialogs: How useful is it for areceiver to get some information, how useful is it fora sender to give this information, and how much credibleinformation can we expect to flow between sender andreceiver? What is the relation between semantics andpragmatics? These Gricean questions will be addressedfrom a decision and game-theoretical point of view.

Summary of Transformative Experience by L.A. Paul and replies to symposiasts. Discussion of undefined values, preference change, authenticity, experiential value, collective minds, mind control.

A "symptomatic act" is an act that is evidence for a state that it has no tendency to cause. In this paper I show that when the evidential value of a symptomatic act might influence subsequent choices, causal decision theory may initially recommend against its own use for those subsequent choices. And if one knows that one will nevertheless use causal decision theory to make those subsequent choices, causal decision theory may favor the one-box solution in Newcomb's problem, and may (...) recommend against making cost-free observations. But if one can control one's future choices, then causal decision theory never recommends against cost-free observation. (shrink)

This paper considers the question of whether predictions of wrongdoing are relevant to our moral obligations. After giving an analysis of ‘won’t’ claims (i.e., claims that an agent won’t Φ), the question is separated into two different issues: firstly, whether predictions of wrongdoing affect our objective moral obligations, and secondly, whether self-prediction of wrongdoing can be legitimately used in moral deliberation. I argue for an affirmative answer to both questions, although there are conditions that must be met for self-prediction to (...) be appropriate in deliberation. The discussion illuminates an interesting and significant tension between agency and prediction. (shrink)

This paper considers the question of whether predictions of wrongdoing are relevant to our moral obligations. After giving an analysis of 'won't' claims, the question is separated into two different issues: firstly, whether predictions of wrongdoing affect our objective moral obligations, and secondly, whether self-prediction of wrongdoing can be legitimately used in moral deliberation. I argue for an affirmative answer to both questions, although there are conditions that must be met for self-prediction to be appropriate in deliberation. The discussion illuminates (...) an interesting and significant tension between agency and prediction. (shrink)

Robert Aumann presents his Agreement Theorem as the key conditional: “if two people have the same priors and their posteriors for an event A are common knowledge, then these posteriors are equal” (Aumann, 1976, p. 1236). This paper focuses on four assumptions which are used in Aumann’s proof but are not explicit in the key conditional: (1) that agents commonly know, of some prior μ, that it is the common prior; (2) that agents commonly know that each of them updates (...) on the prior by conditionalization; (3) that agents commonly know that if an agent knows a proposition, she knows that she knows that proposition (the “K K” principle); (4) that agents commonly know that they each update only on true propositions. It is shown that natural weakenings of any one of these strong assumptions can lead to countermodels to Aumann’s key conditional. Examples are given in which agents who have a common prior and commonly know what probability they each assign to a proposition nevertheless assign that proposition unequal probabilities. To alter Aumann’s famous slogan: people can “agree to disagree”, even if they share a common prior. The epistemological significance of these examples is presented in terms of their role in a defense of the Uniqueness Thesis: If an agent whose total evidence is E is fully rational in taking doxastic attitude D to P, then necessarily, any subject with total evidence E who takes a different attitude to P is less than fully rational. (shrink)

Critical race theorists and standpoint epistemologists argue that agents who are members of dominant social groups are often in a state of ignorance about the extent of their social dominance, where this ignorance is explained by these agents' membership in a socially dominant group (e.g., Mills 2007). To illustrate this claim bluntly, it is argued: 1) that many white men do not know the extent of their social dominance, 2) that they remain ignorant as to the extent of their dominant (...) social position even where this information is freely attainable, and 3) that this ignorance is due in part to the fact that they are white men. We argue that on Buchak's (2010, 2013) model of risk averse instrumental rationality, ignorance of one's privileges can be rational. This argument yields a new account of elite-group ignorance, why it may occur, and how it might be alleviated. (shrink)

Considers that intuitive predictions follow a judgmental heuristic-representativeness. By this heuristic, people predict the outcome that appears most representative of the evidence. Consequently, intuitive predictions are insensitive to the reliability of the evidence or to the prior probability of the outcome, in violation of the logic of statistical prediction. The hypothesis that people predict by representativeness was supported in a series of studies with both naive and sophisticated university students. The ranking of outcomes by likelihood coincided with the ranking by (...) representativeness, and Ss erroneously predicted rare events and extreme values if these happened to be representative. The experience of unjustified confidence in predictions and the prevalence of fallacious intuitions concerning statistical regression are traced to the representativeness heuristic. (shrink)

The pragmatic character of the Dutch book argument makes it unsuitable as an "epistemic" justification for the fundamental probabilist dogma that rational partial beliefs must conform to the axioms of probability. To secure an appropriately epistemic justification for this conclusion, one must explain what it means for a system of partial beliefs to accurately represent the state of the world, and then show that partial beliefs that violate the laws of probability are invariably less accurate than they could be otherwise. (...) The first task can be accomplished once we realize that the accuracy of systems of partial beliefs can be measured on a gradational scale that satisfies a small set of formal constraints, each of which has a sound epistemic motivation. When accuracy is measured in this way it can be shown that any system of degrees of belief that violates the axioms of probability can be replaced by an alternative system that obeys the axioms and yet is more accurate in every possible world. Since epistemically rational agents must strive to hold accurate beliefs, this establishes conformity with the axioms of probability as a norm of epistemic rationality whatever its prudential merits or defects might be. (shrink)

We present a theoretical model of health beliefs and behaviors that explicitly takes into account the emotional impact of possible bad news, ex-ante in the form of anxiety and ex-post in the form of disappointment. Our model makes it possible to explain a number of anomalies such as ’low’ testing rates, heterogeneous perceptions of risk levels, underestimation of health risk, ostriches and hypochondriacs, over-use and under-use of health services, patient preference for information when relatively certain of not being ill, yet (...) avoiding information when relatively certain of being ill, etc. Our model matches observed patterns both in health beliefs and health behaviors and irrational health beliefs and behaviors can be characterized as the optimal response under a given structure of emotions and preferences. (shrink)

When does it make sense to act randomly? A persuasive argument from Bayesian decision theory legitimizes randomization essentially only in tie-breaking situations. Rational behaviour in humans, non-human animals, and artificial agents, however, often seems indeterminate, even random. Moreover, rationales for randomized acts have been offered in a number of disciplines, including game theory, experimental design, and machine learning. A common way of accommodating some of these observations is by appeal to a decision-maker’s bounded computational resources. Making this suggestion both precise (...) and compelling is surprisingly difficult. Toward this end, I propose two fundamental rationales for randomization, drawing upon diverse ideas and results from the wider theory of computation. The first unifies common intuitions in favour of randomization from the aforementioned disciplines. The second introduces a deep connection between randomization and memory: access to a randomizing device is provably helpful for an agent burdened with a finite memory. Aside from fit with ordinary intuitions about rational action, the two rationales also make sense of empirical observations in the biological world. Indeed, random behaviour emerges more or less where it should, according to the proposal. (shrink)

Generalized probabilistic learning takes place in a black-box where present probabilities lead to future probabilities by way of a hidden learning process. The idea that generalized learning can be partially characterized by saying that it doesn’t foreseeably lead to harmful decisions is explored. It is shown that a martingale principle follows for finite probability spaces.

Behaviour norms are considered for decision trees which allow both objective probabilities and uncertain states of the world with unknown probabilities. Terminal nodes have consequences in a given domain. Behaviour is required to be consistent in subtrees. Consequentialist behaviour, by definition, reveals a consequence choice function independent of the structure of the decision tree. It implies that behaviour reveals a revealed preference ordering satisfying both the independence axiom and a novel form of sure-thing principle. Continuous consequentialist behaviour must be expected (...) utility maximizing. Other plausible assumptions then imply additive utilities, subjective probabilities, and Bayes' rule. (shrink)

A person required to risk money on a remote digit of π would, in order to comply fully with the theory [of personal probability] have to compute that digit, though this would really be wasteful if the cost of computation were more than the prize involved. For the postulates of the theory imply that you should behave in accordance with the logical implications of all that you know. Is it possible to improve the theory in this respect, making allowance within (...) it for the cost of thinking, or would that entail paradox? (shrink)

We investigate the general properties of general Bayesian learning, where “general Bayesian learning” means inferring a state from another that is regarded as evidence, and where the inference is conditionalizing the evidence using the conditional expectation determined by a reference probability measure representing the background subjective degrees of belief of a Bayesian Agent performing the inference. States are linear functionals that encode probability measures by assigning expectation values to random variables via integrating them with respect to the probability measure. If (...) a state can be learned from another this way, then it is said to be Bayes accessible from the evidence. It is shown that the Bayes accessibility relation is reflexive, antisymmetric and non-transitive. If every state is Bayes accessible from some other defined on the same set of random variables, then the set of states is called weakly Bayes connected. It is shown that the set of states is not weakly Bayes connected if the probability space is standard. The set of states is called weakly Bayes connectable if, given any state, the probability space can be extended in such a way that the given state becomes Bayes accessible from some other state in the extended space. It is shown that probability spaces are weakly Bayes connectable. Since conditioning using the theory of conditional expectations includes both Bayes’ rule and Jeffrey conditionalization as special cases, the results presented generalize substantially some results obtained earlier for Jeffrey conditionalization. (shrink)

According to Bayesian epistemology, the epistemically rational agent updates her beliefs by conditionalization: that is, her posterior subjective probability after taking account of evidence X, pnew, is to be set equal to her prior conditional probability pold(·|X). Bayesians can be challenged to provide a justification for their claim that conditionalization is recommended by rationality—whence the normative force of the injunction to conditionalize? There are several existing justifications for conditionalization, but none directly addresses the idea that conditionalization will be epistemically rational (...) if and only if it can reasonably be expected to lead to epistemically good outcomes. We apply the approach of cognitive decision theory to provide a justification for conditionalization using precisely that idea. We assign epistemic utility functions to epistemically rational agents; an agent’s epistemic utility is to depend both upon the actual state of the world and on the agent’s credence distribution over possible states. We prove that, under independently motivated conditions, conditionalization is the unique updating rule that maximizes expected epistemic utility. (shrink)