This article examines two questions about scientists’ search for knowledge. First, which search strategies generate discoveries effectively? Second, is it advantageous to diversify search strategies? We argue pace Weisberg and Muldoon, “Epistemic Landscapes and the Division of Cognitive Labor”, that, on the first question, a search strategy that deliberately seeks novel research approaches need not be optimal. On the second question, we argue they have not shown epistemic reasons exist for the division of cognitive labor, identifying the errors that led (...) to their conclusions. Furthermore, we generalize the epistemic landscape model, showing that one should be skeptical about the benefits of social learning in epistemically complex environments. (shrink)

Legg and Hutter, as well as subsequent authors, considered intelligent agents through the lens of interaction with reward-giving environments, attempting to assign numeric intelligence measures to such agents, with the guiding principle that a more intelligent agent should gain higher rewards from environments in some aggregate sense. In this paper, we consider a related question: rather than measure numeric intelligence of one Legg- Hutter agent, how can we compare the relative intelligence of two Legg-Hutter agents? We propose an elegant answer (...) based on the following insight: we can view Legg-Hutter agents as candidates in an election, whose voters are environments, letting each environment vote (via its rewards) which agent (if either) is more intelligent. This leads to an abstract family of comparators simple enough that we can prove some structural theorems about them. It is an open question whether these structural theorems apply to more practical intelligence measures. (shrink)

After generalizing the Archimedean property of real numbers in such a way as to make it adaptable to non-numeric structures, we demonstrate that the real numbers cannot be used to accurately measure non-Archimedean structures. We argue that, since an agent with Artificial General Intelligence (AGI) should have no problem engaging in tasks that inherently involve non-Archimedean rewards, and since traditional reinforcement learning rewards are real numbers, therefore traditional reinforcement learning cannot lead to AGI. We indicate two possible ways traditional reinforcement (...) learning could be altered to remove this roadblock. (shrink)

Do we live in a computer simulation? I will present an argument that the results of a certain experiment constitute empirical evidence that we do not live in, at least, one type of simulation. The type of simulation ruled out is very specific. Perhaps that is the price one must pay to make any kind of Popperian progress.

Can an AGI create a more intelligent AGI? Under idealized assumptions, for a certain theoretical type of intelligence, our answer is: “Not without outside help”. This is a paper on the mathematical structure of AGI populations when parent AGIs create child AGIs. We argue that such populations satisfy a certain biological law. Motivated by observations of sexual reproduction in seemingly-asexual species, the Knight-Darwin Law states that it is impossible for one organism to asexually produce another, which asexually produces another, and (...) so on forever: that any sequence of organisms (each one a child of the previous) must contain occasional multi-parent organisms, or must terminate. By proving that a certain measure (arguably an intelligence measure) decreases when an idealized parent AGI single-handedly creates a child AGI, we argue that a similar Law holds for AGIs. (shrink)

I present an argument that for any computer-simulated civilization we design, the mathematical knowledge recorded by that civilization has one of two limitations. It is untrustworthy, or it is weaker than our own mathematical knowledge. This is paradoxical because it seems that nothing prevents us from building in all sorts of advantages for the inhabitants of said simulation.

In this review essay, I review in detail Abram de Swann's fine new book, The Killing Compartments. The book is a theoretical analysis of the varieties and causes of genocides and other mass asymmetrical killing campaigns. I then suggest several criticisms of his analysis.

In their thought-provoking paper, Legg and Hutter consider a certain abstrac- tion of an intelligent agent, and define a universal intelligence measure, which assigns every such agent a numerical intelligence rating. We will briefly summarize Legg and Hutter’s paper, and then give a tongue-in-cheek argument that if one’s goal is to become more intelligent by cultivating music appreciation, then it is bet- ter to use classical music (such as Bach, Mozart, and Beethoven) than to use more recent pop music. The (...) same argument could be adapted to other media: books beat films, card games beat first-person shooters, parables beat dissertations, etc. We leave it to the reader to decide whether this argument tells us something about classical music, something about Legg-Hutter intelligence, or something about both. (shrink)

Jennifer Nagel (2010) has recently proposed a fascinating account of the decreased tendency to attribute knowledge in conversational contexts in which unrealized possibilities of error have been mentioned. Her account appeals to epistemic egocentrism, or what is sometimes called the curse of knowledge, an egocentric bias to attribute our own mental states to other people (and sometimes our own future and past selves). Our aim in this paper is to investigate the empirical merits of Nagel’s hypothesis about the psychology involved (...) in knowledge attribution. (shrink)

We define a notion of the intelligence level of an idealized mechanical knowing agent. This is motivated by efforts within artificial intelligence research to define real-number intelligence levels of compli- cated intelligent systems. Our agents are more idealized, which allows us to define a much simpler measure of intelligence level for them. In short, we define the intelligence level of a mechanical knowing agent to be the supremum of the computable ordinals that have codes the agent knows to be codes (...) of computable ordinals. We prove that if one agent knows certain things about another agent, then the former necessarily has a higher intelligence level than the latter. This allows our intelligence no- tion to serve as a stepping stone to obtain results which, by themselves, are not stated in terms of our intelligence notion (results of potential in- terest even to readers totally skeptical that our notion correctly captures intelligence). As an application, we argue that these results comprise evidence against the possibility of intelligence explosion (that is, the no- tion that sufficiently intelligent machines will eventually be capable of designing even more intelligent machines, which can then design even more intelligent machines, and so on). (shrink)

Can you find an xy-equation that, when graphed, writes itself on the plane? This idea became internet-famous when a Wikipedia article on Tupper’s self-referential formula went viral in 2012. Under scrutiny, the question has two flaws: it is meaningless (it depends on fonts) and it is trivial. We fix these flaws by formalizing the problem.

This paper argues that higher-order doubt generates an epistemic dilemma. One has a higher-order doubt with regards to P insofar as one justifiably withholds belief as to what attitude towards P is justified. That is, one justifiably withholds belief as to whether one is justified in believing, disbelieving, or withholding belief in P. Using the resources provided by Richard Feldman’s recent discussion of how to respect one’s evidence, I argue that if one has a higher-order doubt with regards to P, (...) then one is not justified in having any attitude towards P. Otherwise put: No attitude towards the doubted proposition respects one’s higher-order doubt. I argue that the most promising response to this problem is to hold that when one has a higher-order doubt about P, the best one can do to respect such a doubt is to simply have no attitude towards P. Higher-order doubt is thus much more rationally corrosive than non-higher-order doubt, as it undermines the possibility of justifiably having any attitude towards the doubted proposition. (shrink)

We construct a machine that knows its own code, at the price of not knowing its own factivity.

If we define tradition too hastily we leave to one side the question of what the relevance of tradition is for us. Here the concept of tradition is opened up by considering the different views of it taken by Hannah Arendt, Michael Oakeshott and Alasdair MacIntyre. We see that each has put tradition into a fully developed picture of what our predicament is in modernity; and that each has differed in their assessment of what our relation to tradition is or (...) should be. Arendt sees tradition as something which no longer conditions action, Oakeshott sees tradition as something which conditions all action, and MacIntyre sees tradition as something which should condition right action. In each case, the view of tradition is clearly one element in an attempt to see how the most important constituent elements of human existence – variously called the human condition, human conduct, or human virtue – should be understood in a modernity which is ours because it has put the traditional concept of tradition into question. (shrink)

A variation of Fitch’s paradox is given, where no special rules of inference are assumed, only axioms. These axioms follow from the familiar assumptions which involve rules of inference. We show (by constructing a model) that by allowing that possibly the knower doesn’t know his own soundness (while still requiring he be sound), Fitch’s paradox is avoided. Provided one is willing to admit that sound knowers may be ignorant of their own soundness, this might offer a way out of the (...) paradox. (shrink)

This is an appendix containing the stimulus materials for the experiments reported in the paper ‘Salient Alternatives in Perspective.’.

Background: Individuals with psychopathic traits demonstrate an attenuated emotional response to aversive stimuli. However, recent evidence suggests heterogeneity in emotional reactivity among individuals with psychopathic or callous-unemotional (CU) traits, the emotional detachment dimension of psychopathy. We hypothesize that primary variants of psychopathy will respond with blunted affect to negatively valenced stimuli, whereas individuals marked with histories of childhood trauma/maltreatment exposure, known as secondary variants, will display heightened emotional reactivity. To test this hypothesis, the present study examined fear-potentiated startle between psychopathy (...) variants while viewing aversive, pleasant, and neutral scenes. Method: 238 incarcerated adolescent (M age = 16.8, SD = 1.11 years) boys completed a picture-startle paradigm and self-report questionnaires assessing CU traits, antisocial-aggressive behavior, and maltreatment. Results: Latent profile analyses identified four classes; primary variants (high CU traits, high aggression, low maltreatment; n = 46), secondary variants (high CU traits, high aggression, high maltreatment; n = 42), and two nonpsychopathic groups differentiated on maltreatment experience (n = 148). Findings from an ANOVA comparing identified groups on startle amplitude difference scores (i.e., aversive-neutral) suggested a main effect for group, F(3,196)=8.91, p<.001, η2 = .12. Primary variants of juvenile psychopathy displayed reduced startle potentiation to aversive images (threat and victim scenes), whereas secondary variants distinguished by high levels of childhood maltreatment did not. Conclusions: Findings add to a rapidly growing body of literature supporting the possibility of multiple developmental pathways to psychopathy (i.e., equifinality), and extend it by finding support for divergent potential biomarkers between primary and secondary psychopathy variants. (shrink)

SummaryIn this paper, two different theoretical problems of induction are delineated. The first problem is addressed; the second problem is deferred to the sequel to this paper. The first problem of induction is taken to be the seemingly unformalizable nature of traditional inductive arguments. It is shown that the problem does not arise out of some particularly dubious argument form , but rather from the presupposition that inductive “logic” is, like deductive logic, assertoric. Rather , inductive logic is dialectical in (...) nature.RésuméDans cet article, deux problèmes théoriques de l'induction sont formulés. Le premier est traité, le second le sera dans une suite à cet article. Le premier problème de l'induction est ici la nature apparemment non formalisable des arguments inductifs traditionnels. L'auteur montre que le problème ne provient pas d'une forme d'argumentation particulièrement douteuse , mais au contraire de la présupposition que la «logique» inductive est, comme la logique déductive, assertorique. Il soutient au contraire qu'elle est par nature dialectique.ZusammenfassungIn der vorliegenden Arbeit werden zwei theoretische Probleme der Induktion formuliert, wovon das zweite in einem anderen Artikel behandelt werden soll. Das erste Problem betrifft die scheinbar nicht formalisierbare Natur der üblichen induktiven Argumente. Der Verfasser zeigt dass das Problem nicht von einer besonders zweifelhaften Argumentationsform herrührt , sondern von der Voraussetzung, dass die induktive «Logik» — wie die deduktive Logik — assertorisch sei. Es wird dagegen gezeigt, dass die induktive Logik dialektischer Natur ist. (shrink)

Reinhardt’s conjecture, a formalization of the statement that a truthful knowing machine can know its own truthfulness and mechanicalness, was proved by Carlson using sophisticated structural results about the ordinals and transfinite induction just beyond the first epsilon number. We prove a weaker version of the conjecture, by elementary methods and transfinite induction up to a smaller ordinal.

This is a review of the biopic of the great mathematician Ramanujan, 'The Man Who Knew Infinity'(2016).

A suprise may occur if we use a similar strategy to the Liar's paradox to mathematically formalize "This sentence does not contain the symbol X".

Christopher Johnson has put forward in this journal the view that ad hominem reasoning may be more generally reasonable than is allowed by writers such as myself, basing his view on virtue epistemology. I review his account, as well as the standard account, of ad hominem reasoning, and show how the standard account would handle the cases he sketches in defense of his own view. I then give four criticisms of his view generally: the problems of virtue conflict, vagueness, conflation (...) of speech acts, and self-defeating counsel. I then discuss four reasons why the standard account is superior: it better fits legal reality, the account of other fallacies, psychological science, and political reality. (shrink)

In his dissertation, Wadge defined a notion of guessability on subsets of the Baire space and gave two characterizations of guessable sets. A set is guessable if and only if it is in the second ambiguous class, if and only if it is eventually annihilated by a certain remainder. We simplify this remainder and give a new proof of the latter equivalence. We then introduce a notion of guessing with an ordinal limit on how often one can change one’s mind. (...) We show that for every ordinal $\alpha$, a guessable set is annihilated by $\alpha$ applications of the simplified remainder if and only if it is guessable with fewer than $\alpha$ mind changes. We use guessability with fewer than $\alpha$ mind changes to give a semi-characterization of the Hausdorff difference hierarchy, and indicate how Wadge’s notion of guessability can be generalized to higher-order guessability, providing characterizations of ${\mathbf{\Delta}}^{0}_{\alpha}$ for all successor ordinals $\alpha\gt 1$. (shrink)

A biologically unavoidable sequence is an infinite gender sequence which occurs in every gendered, infinite genealogical network satisfying certain tame conditions. We show that every eventually periodic sequence is biologically unavoidable (this generalizes König's Lemma), and we exhibit some biologically avoidable sequences. Finally we give an application of unavoidable sequences to cellular automata.