•  1329
    In Defence of the Hivemind Society
    Neuroethics 14 (2): 253-267. 2020.
    The idea that humans should abandon their individuality and use technology to bind themselves together into hivemind societies seems both farfetched and frightening – something that is redolent of the worst dystopias from science fiction. In this article, we argue that these common reactions to the ideal of a hivemind society are mistaken. The idea that humans could form hiveminds is sufficiently plausible for its axiological consequences to be taken seriously. Furthermore, far from being a dyst…Read more
  •  234
    Machines learning values
    In S. Matthew Liao (ed.), Ethics of Artificial Intelligence, Oxford University Press. forthcoming.
    Whether it would take one decade or several centuries, many agree that it is possible to create a *superintelligence*---an artificial intelligence with a godlike ability to achieve its goals. And many who have reflected carefully on this fact agree that our best hope for a "friendly" superintelligence is to design it to *learn* values like ours, since our values are too complex to program or hardwire explicitly. But the value learning approach to AI safety faces three particularly philosophical …Read more
  •  276
    There are writers in both metaphysics and algorithmic information theory (AIT) who seem to think that the latter could provide a formal theory of the former. This paper is intended as a step in that direction. It demonstrates how AIT might be used to define basic metaphysical notions such as *object* and *property* for a simple, idealized world. The extent to which these definitions capture intuitions about the metaphysics of the simple world, times the extent to which we think the simple world …Read more
  •  319
    A Normative Yet Coherent Naturalism
    Philo 17 (1): 77-91. 2014.
    Naturalism is normally taken to be an ideology, censuring non-naturalistic alternatives. But as many critics have pointed out, this ideological stance looks internally incoherent, since it is not obviously endorsed by naturalistic methods. Naturalists who have addressed this problem universally foreswear the normative component of naturalism by, in effect, giving up science’s exclusive claim to legitimacy. This option makes naturalism into an empty expression of personal preference that can carr…Read more
  •  853
    Is it good for them too? Ethical concern for the sexbots
    In John Danaher & Neil McArthur (eds.), Robot Sex: Social Implications and Ethical, Mit Press. pp. 155-171. 2017.
    In this chapter I'd like to focus on a small corner of sexbot ethics that is rarely considered elsewhere: the question of whether and when being a sexbot might be good---or bad---*for the sexbot*. You might think this means you are in for a dry sermon about the evils of robot slavery. If so, you'd be wrong; the ethics of robot servitude are far more complicated than that. In fact, if the arguments here are right, designing a robot to serve humans sexually may be very good for the robots themselv…Read more
  •  798
    Superintelligence as superethical
    In Patrick Lin, Keith Abney & Ryan Jenkins (eds.), Robot Ethics 2.0, Oxford University Press. pp. 322-337. 2017.
    Nick Bostrom's book *Superintelligence* outlines a frightening but realistic scenario for human extinction: true artificial intelligence is likely to bootstrap itself into superintelligence, and thereby become ideally effective at achieving its goals. Human-friendly goals seem too abstract to be pre-programmed with any confidence, and if those goals are *not* explicitly favorable toward humans, the superintelligence will extinguish us---not through any malice, but simply because it will want our…Read more
  •  549
    Composition as pattern
    Philosophical Studies 176 (5): 1119-1139. 2019.
    I argue for patternism, a new answer to the question of when some objects compose a whole. None of the standard principles of composition comfortably capture our natural judgments, such as that my cat exists and my table exists, but there is nothing wholly composed of them. Patternism holds, very roughly, that some things compose a whole whenever together they form a “real pattern”. Plausibly we are inclined to acknowledge the existence of my cat and my table but not of their fusion, because the…Read more
  •  62
    Methodological naturalism states (roughly speaking) that only science can be a route to knowledge. This purported piece of knowledge looks self-condemning, however; after all, it was formulated in the armchair, and not in the laboratory. I argue that on a popular (if largely unarticulated) construal of naturalism as inference to the best explanation, methodological naturalism escapes this charge of internal incoherence, and in fact is self-endorsing rather than self-condemning.
  •  32
    given at the 2007 Formal Epistemology Workshop at Carnegie Mellon June 2nd. Good compression must track higher vs lower probability of inputs, and this is one way to approach how simplicity tracks truth.
  •  76
    Jeffrey conditioning allows updating in Bayesian style when the evidence is uncertain. A weighted average, essentially, over classically updating on the alternatives. Unlike classical Bayesian conditioning, this allows learning to be unlearned.
  •  380
    The ethics of robot servitude
    Journal of Experimental and Theoretical Artificial Intelligence 19 (1): 43-54. 2007.
    Assume we could someday create artificial creatures with intelligence comparable to our own. Could it be ethical use them as unpaid labor? There is very little philosophical literature on this topic, but the consensus so far has been that such robot servitude would merely be a new form of slavery. Against this consensus I defend the permissibility of robot servitude, and in particular the controversial case of designing robots so that they want to serve human ends. A typical objection to this ca…Read more
  •  44
    The simplicity of a theory seems closely related to how well the theory summarizes individual data points. Think, for example, of classic curve-fitting. It is easy to get perfect data-fit with a ‘‘theory’’ that simply lists each point of data, but such a theory is maximally unsimple (for the data-fit). The simple theory suggests instead that there is one underlying curve that summarizes this data, and we usually prefer such a theory even at some expense in data-fit. In general, it seems, theoriz…Read more
  •  263
    Designing People to Serve
    In Patrick Lin, George Bekey & Keith Abney (eds.), Robot Ethics, Mit Press. 2011.
    I argue that, contrary to intuition, it would be both possible and permissible to design people - whether artificial or organic - who by their nature desire to do tasks we find unpleasant.
  •  186
    Analysis, schmanalysis
    Canadian Journal of Philosophy 38 (2). 2008.
    In Naming and Necessity, Saul Kripke employs a handy philosophical trick: he invents the term ‘schmidentity’ to argue indirectly for his favored account of identity. Kripke says in a footnote that he wishes someday “to elaborate on the utility of this device”. In this paper, I first take up a general elaboration on his behalf. I then apply the trick to support an attractive but somewhat unorthodox picture of conceptual analysis—one according to which it is a process of forming intentions for word…Read more
  •  64
    What is “property”? Property Roughly, thing x is the (private) property of agent A if and only if A has exclusive and extensive legal rights of access and / or use for x.
  •  601
    Utilitarian epistemology
    Synthese 190 (6): 1173-1184. 2013.
    Standard epistemology takes it for granted that there is a special kind of value: epistemic value. This claim does not seem to sit well with act utilitarianism, however, since it holds that only welfare is of real value. I first develop a particularly utilitarian sense of “epistemic value”, according to which it is closely analogous to the nature of financial value. I then demonstrate the promise this approach has for two current puzzles in the intersection of epistemology and value theory: firs…Read more
  •  40
    Functions, creatures, learning, emotion
    Hudlicka and Canamero. 2004.
    I propose a conceptual framework for emotions according to which they are best understood as the feedback mechanism a creature possesses in virtue of its function to learn. More specifically, emotions can be neatly modeled as a measure of harmony in a certain kind of constraint satisfaction problem. This measure can be used as error for weight adjustment (learning) in an unsupervised connectionist network.
  •  113
    Arthur Falk has proposed a new construal of faith according to which it is not a mere species of belief, but has essential components in action. This twist on faith promises to resurrect Pascal’s Wager, making faith compatible with reason by believing as the scientist but acting as the theist. I argue that Falk’s proposal leaves religious faith in no better shape; in particular, it merely reframes the question in terms of rational desires rather than rational beliefs.
  •  193
    Belief-desire coherence
    Dissertation, University of Michigan. 2003.
    Tradition compels me to write dissertation acknowledgements that are long, effusive, and unprofessional. Fortunately for me, I heartily endorse that tradition.