Empiricism without magic: transformational abstraction in deep convolutional neural networks

In artificial intelligence, recent research has demonstrated the remarkable potential of Deep Convolutional Neural Networks (DCNNs), which seem to exceed state-of-the-art performance in new domains weekly, especially on the sorts of very difficult perceptual discrimination tasks that skeptics thought would remain beyond the reach of artificial intelligence. However, it has proven difficult to explain why DCNNs perform so well. In philosophy of mind, empiricists have long suggested that complex cognition is based on information derived from sensory experience, often appealing to a faculty of abstraction. Rationalists have frequently complained, however, that empiricists never adequately explained how this faculty of abstraction actually works. In this paper, I tie these two questions together, to the mutual benefit of both disciplines. I argue that the architectural features that distinguish DCNNs from earlier neural networks allow them to implement a form of hierarchical processing that I call “transformational abstraction”. Transformational abstraction iteratively converts sensory-based representations of category exemplars into new formats that are increasingly tolerant to “nuisance variation” in input. Reflecting upon the way that DCNNs leverage a combination of linear and non-linear processing to efficiently accomplish this feat allows us to understand how the brain is capable of bi-directional travel between exemplars and abstractions, addressing longstanding problems in empiricist philosophy of mind. I end by considering the prospects for future research on DCNNs, arguing that rather than simply implementing 80s connectionism with more brute-force computation, transformational abstraction counts as a qualitatively distinct form of processing ripe with philosophical and psychological significance, because it is significantly better suited to depict the generic mechanism responsible for this important kind of psychological processing in the brain.     

Non-Naturalist Moral Realism,Autonomy and Entanglement

It was something of a dogma for much of the twentieth century that one cannot validly derive an ought from an is. More generally, it was held that non-normative propositions do not entail normative propositions. Call this thesis about the relation between the natural and the normative Natural-Normative Autonomy (or Autonomy for short). The denial of Autonomy involves the entanglement of the natural with the normative. Naturalism entails entanglement—in fact it entails the most extreme form of entanglement—but entanglement does not entail naturalism. In a ground-breaking paper “The autonomy of ethics” Arthur Prior constructed some intriguing counterexamples to Autonomy. While his counterexamples have convinced few, there is little agreement on what is wrong with them. I present a new analysis of Autonomy, one which is grounded in a general and independently plausible account of subject matters. While Prior’s arguments do establish shallow natural-normative entanglement, this is a consequence of simple logical relationships that hold between just about any two subject matters. It has nothing special to do with the logical structure of normativity or its relation to the natural. Prior’s arguments (along with several others) leave the fundamental idea behind natural-normative Autonomy intact. I offer a new argument for deep entanglement. I show that in any framework adequate for dealing with the natural and the normative spheres, a purely natural proposition entails a purely normative proposition, and vice-versa. But this is no threat to non-naturalist moral realism. In fact it helps ameliorate the excesses of an extreme non-naturalism, delivering a more palatable and plausible position.     

Nonlinear Predictive Models for Multiple Mediation Analysis: With an Application to Explore Ethnic Disparities in Anxiety and Depression Among Cancer Survivors

Mediation analysis allows the examination of effects of a third variable (mediator/confounder) in the causal pathway between an exposure and an outcome. The general multiple mediation analysis method (MMA), proposed by Yu et al., improves traditional methods (e.g., estimation of natural and controlled direct effects) to enable consideration of multiple mediators/confounders simultaneously and the use of linear and nonlinear predictive models for estimating mediation/confounding effects. Previous studies find that compared with non-Hispanic cancer survivors, Hispanic survivors are more likely to endure anxiety and depression after cancer diagnoses. In this paper, we applied MMA on MY-Health study to identify mediators/confounders and quantify the indirect effect of each identified mediator/confounder in explaining ethnic disparities in anxiety and depression among cancer survivors who enrolled in the study. We considered a number of socio-demographic variables, tumor characteristics, and treatment factors as potential mediators/confounders and found that most of the ethnic differences in anxiety or depression between Hispanic and non-Hispanic white cancer survivors were explained by younger diagnosis age, lower education level, lower proportions of employment, less likely of being born in the USA, less insurance, and less social support among Hispanic patients.     

Psychometric Models of Small Group Collaborations

The social combination theory of group problem solving is used to extend existing psychometric models to collaborative settings. A model for pairwise group work is proposed, the implications of the model for assessment design are considered, and its estimation is addressed. The results are illustrated with an empirical example in which dyads work together on a twelfth-grade level mathematics assessment. In conclusion, attention is given to avenues of research that seem most fruitful for advancing current initiatives concerning the assessment of collaboration, teamwork, and related constructs.     

Does ‘Ought’ Imply ‘Might’? How (not) to Resolve the Conflict between Act and Motive Utilitarianism

Utilitarianism has often been understood as a theory that concerns itself first and foremost with the rightness of actions; but many other things (e.g., moral rules, motives, laws, etc.) are also properly subject to moral evaluation, and utilitarians have long understood that the theory must be able to provide an account of these as well. In a landmark article from 1976, Robert Adams argues that traditional act utilitarianism faces a particular problem in this regard. He argues that a on a sensible utilitarian account of the rightness of an agent’s motives, right motives will sometimes conflict with right actions, leaving the theory internally incoherent. The puzzle Adams raises has received a good deal of attention but few proposed solutions. Fred Feldman, however, has offered a solution that seems to be gaining adherents. In this paper I argue that Feldman’s approach cannot succeed. At bottom, it relies on a version of the principle that ‘ought’ implies ‘can’—and subsequently an account of an agent’s alternatives—that is far too restrictive to be plausible. Despite the failure of this solution, however, I argue that the conflict Adams develops is not as theoretically troubling as he suggests. While traditional act utilitarianism may fail for other reasons, it will not fail due to the conflict between acts and motives.     

Moods,Colored Lenses,and Emotional Disconnection: a Comment on Gallegos

In “Moods Are Not Colored Lenses: Perceptualism and the Phenomenology of Moods” (2017) Francisco Gallegos presents a challenge to popular view about the phenomenology of being in a mood that he calls “perceptualism”. In this essay, I offer a partial defense of perceptualism about moods and argue that perceptualism and Gallegos’s preferred Heideggerian alternative need not be viewed as in opposition to one another.     

<Emphasis Type="Bold">W</Emphasis><Emphasis Type="Bold">ang</Emphasis><Emphasis Type="Bold">Fuzhi’s Criticism of Buddhism and Its Limitations</Emphasis>

Wang Fuzhi’s 王夫之 remarks on Buddhism have not been given sufficient attention despite increasing research on him. The few works on this topic either focus on just one aspect of his view of Buddhism or fail to disclose the purpose and uniqueness of his attack of it. This essay analyzes his view of Buddhism comprehensively, in particular his insight into the paradox of Buddhist universal love and his rejection of Buddhist retribution and reincarnation from Confucian righteousness and qi 氣-monism. In addition, it also explores the reason, context, and limitations of his criticism, that is, his reaction to the popular approach of “understanding Confucian classics through learning Buddhism” in the late Ming 明, his response to Zongmi’s 宗密 criticism of Confucian cosmology and human nature, and his misunderstanding of some Buddhist concepts. Yet his criticism is still illuminating to our understanding of the interaction of Confucianism with Buddhism and other religions.     

Maker Cultures and the Prospects for Technological Action

Supported by easier and cheaper access to tools and expanding communities, maker cultures are pointing towards the ideas of (almost) everyone designing, creating, producing and distributing renewed, new and improved products, machines, things or artefacts. A careful analysis of the assumptions and challenges of maker cultures emphasizes the relevance of what may be called technological action, that is, active and critical interventions regarding the purposes and applications of technologies within ordinary lives, thus countering the deterministic trends of current directions of technology. In such transformative potential, we will explore a set of elements what is and could be technological action through snapshots of maker cultures based on the empirical research conducted in three particular contexts: the Fab Lab Network, Maker Media core outputs and initiatives such as Maker Faires, and the Open Source Hardware Association (OSHWA). Elements such as control and empowerment through material engagement, openness and sharing, and social, cultural, political and ethical values of the common good in topics such as diversity, sustainability and transparency, are critically analysed.     

Condom Use Among South African University Students in the Province of KwaZulu-Natal

South Africa has one of the highest HIV infection rates in the world, with young people particularly affected. Condoms are an effective preventative method against HIV among sexually active adults. This study assessed the level of condom usage among university students in South Africa and their attitudes towards condom usage, negotiation efficacy and confidence in condom usage. It was a cross-sectional study conducted amongst South African university students living at the university residences. Students were randomly selected and required to complete a self-administered questionnaire. The mean age of participants (n?=?441) was 22.7?±?4.3 years. The mean age of sexual debut was 17.7?±?3.0 years with 1.4?±?1.47 (range 0–13) current partners. Less than half of the students used a condom at sexual debut and only 28.5% used condoms during recent sexual activity. Those students who used a condom at first sexual intercourse had marginally more positive attitudes about condom usage than those who did not do so (p?=?0.056). Students with a single current partner had more positive attitudes about condom usage than those with multiple current partners (p?=?0.021). Only 32.5% (n?=?127) of the students were very confident in using condoms. Similarly, only 33.0% (n?=?130) of the participants felt that they could definitely negotiate condom use with their partners. We conclude that condom use among South African students is low and that they lack the confidence to use condoms. We recommend that programmes to step up condom use must also incorporate educational interventions on usage.