Research Interests

Cognition and the Web

My main area of research interest relates to the cognitive and epistemic impacts of Web technology. I am particularly interested in issues that lie at the intersection of Web Science, cognitive science and the philosophy of mind. One of the main issues I have sought to address relates to whether the features of current and near-future Web-enabled devices are able to support the emergence of Web-based forms of cognitive extension. This is an issue that emerges from recent debates in the philosophy of mind concerning the notion of active externalism. I have argued that we should adopt an ecological approach to the Web, recognizing the Web as an important aspect of the cognitive ecology in which a variety of cognitive systems may emerge. In contrast to the usual focus in cognitive science and the philosophy of mind, I have argued that issues of cognitive extension enable us to see how the Web provides opportunities for the realization of novel forms of machine intelligence, especially ones in which human agents serve as part of the physical fabric underlying machine-based cognitive processes.

Key Concepts
  • Web-Extended Minds: The idea that the informational and technological elements of the Web can, on occasion, serve as part of the physical machinery that realizes human mental states and processes.
  • Web-Extended Knowers: The idea that our modes of interaction with the Web serve as the basis for extended cognitive systems that generate extended knowledge, i.e., true beliefs that result from extended cognitive processes that involve elements of the Web.
  • Human-Extended Machine Cognition: The idea that the Web supports the emergence of new forms of machine intelligence by enabling human agents to form part of the physical fabric that realizes machine-based cognitive processes.
  • Conjoined Minds: The idea that mutual access to online resources gives rise to the possibility of cognitive systems in which the physical machinery of one person's mind also forms part of the physical machinery of another person's mind. The notion of conjoined minds is essentially the idea that the Web supports forms of cognitive inter-penetration and cognitive inter-dependence that are just as intimate as those proposed in the context of research into brain-to-brain interfaces.

Predictive Processing

An emerging consensus in cognitive science views the biological brain as a hierarchically-organized predictive processing system. This is a system in which higher-order regions are continuously attempting to predict the activity of lower-order regions at a variety of (increasingly abstract) spatial and temporal scales. The brain is thus revealed as a hierarchical prediction machine that is constantly engaged in the attempt to predict the flow of information originating from the sensory surfaces. Such a view seems to afford a great deal of explanatory leverage when it comes to a broad swathe of seemingly disparate psychological phenomena (e.g., learning, memory, perception, action, emotion, planning, reasoning, imagination, and conscious experience). In the most positive case, the predictive processing story seems to provide our first glimpse as to what a unified (computationally-tractable and neurobiologically plausible) account of human psychology might look like. This obviously marks out one reason why predictive processing models deserve to be the focus of current empirical and theoretical attention. Another reason, however, is rooted in the potential of such models to advance the current state-of-the-art in machine intelligence and machine learning. Interestingly, the vision of the brain as a hierarchical prediction machine is one that establishes contact with work that goes under the heading of 'deep learning'. Deep learning systems thus often attempt to make use of predictive processing schemes and (increasingly abstract) generative models as a means of supporting the analysis of large data sets. But are such computational systems sufficient (by themselves) to provide a route to general human-level intelligence? I suggest that they are not and that closer attention to a broader range of forces and factors (many of which are not confined to the neural realm) may be required to understand what it is that gives human cognition its distinct (and largely unique) flavour. The vision that emerges is one of 'homomimetic deep learning systems', systems that situate a hierarchically-organized predictive processing core within a larger nexus of developmental, behavioural, symbolic, technological and social influences. Relative to that vision, I suggest that predictive processing accounts provide a conceptual springboard for a range of novel ideas relating to digital immortality, machine consciousness, and the 'minding' of social activities and social processes.

Key Concepts
  • Digital Immortality: The idea that synthetic predictive processing systems provide the basis for a form of digital immortality or digital resurrection. This idea is grounded in the fact that emerging digital technologies are able to yield ever-more intimate and detailed records of our actions, experiences, and physiological responses. When such data are presented to a predictive processing system, we can ask whether the human biological body is serving as a platform for the creation of digital doppelgangers, i.e., synthetic systems that capture psychologically-significant traces of ourselves. Such claims have a range of important implications for our notions of mind uploading, digital resurrection, data cryogenics, digital immortality, and digital preservation. They also, it should be clear, introduce an additional dimension to current debates concerning digital surveillance, privacy and the quantified self.
  • Homomimetic Deep Learning Systems: The idea that extant deep learning systems overlook a number of crucial features associated with human (neuro)cognition. By attending to these factors, we can, I suggest, envisage state-of-the-art improvements in our current arsenal of deep learning systems.
  • The Neocortex 2.0: The idea that the functional neuroanatomy of the mammalian neocortex should be revised to accommodate the insights of the predictive processing framework. Such revisions include abandoning the conventional idea of primary and secondary motor cortices as the source of motoric commands. It also involves a shift in our views regarding the functional role of the prefrontal cortices, especially as they relate to the realization of higher cognitive functions.
  • Mind of Society: The idea that predictive processing systems could be applied to bodies of social data (such as those gathered by major technology vendors such as Google, Facebook and Twitter) in order to yield a predictively-relevant generative model of the human social environment. Predictive processing accounts thus emerge as relevant not just to human psychology but also (as a result of the advent of the Social Web) as applicable to the explanatory and predictive interests of the social science community. It should also be clear, of course, that inasmuch as predictive processing architectures serve as the mechanistic substrate of experientially-potent forms of cognition, then the notion of socially-oriented predictive machines encourages us to think about the possibility of new forms of synthetic forms of conscious experience that are grounded in the dynamics of the (human) social world.

Collective Cognition

What are the factors that determine the collective performance of groups engaged in collective tasks? This question lies at the heart of a broad array of research endeavours that straddle the disciplines of social psychology, network science, and cognitive systems engineering. My approach to this question is grounded in the use of cognitive architectures and multi-agent computer simulations. I rely on the use of cognitive modelling techniques to develop computational models of individual agent cognition. I then study the relationship between agent interaction and collective performance outcomes. This approach to simulating the performance of collective cognitive systems sometimes goes under the heading of cognitive social simulation.

Key Concepts
  • Mandevillian Intelligence: Mandevillian intelligence is a form of collective intelligence in which cognitive shortcomings, limitations and biases at the level of individual agents play a positive functional role in supporting collective intelligence. The hypothesis of mandevillian intelligence is essentially the idea that individual cognitive vice can, on occasion, give rise to collective cognitive virtue.
  • Adaptive Coupling: The idea that adaptive changes in the effective connectivity of a networked ensemble can work to the cognitive and epistemic good of the larger systemic organization. The biological brain incorporates mechanisms (e.g., control neurons, gating mechanisms, and convergence zones) that support dynamic forms of information flow and routing between disparate anatomical regions. Functionally similar processes, operating in the context of socio-technical systems, may help to improve the quality of socio-cognitive and socio-epistemic outcomes.

Cognition in the Virtual World

This area of research subsumes two major topics of interest. Firstly, I am interested in the use of virtual environments for the purposes of performing computer simulations of embodied, extended, enactive and embedded cognitive processes. By making use of contemporary game engines and 3D modelling techniques, I suggest that we can use computer simulation techniques to study the way in which intelligent behaviour emerges from processing loops that extend across the brain, the body and the world. A second focus of interest relates to the cognitive impact of virtual and augmented reality technology. I have suggested that augmented reality technologies enable us to rapidly design 'effective' environments that shape, scaffold and support our cognitive profiles. I have also sought to understand the implications of holographic computing devices for notions of cognitive extension.

Key Concepts
  • Hologrammatically-Extended Minds: The idea that holographic projections may form part of the supervenience base for human mental states and processes.
  • Virtual Designer Environments: The idea that virtual technologies enable us to engineer our environments in ways that support our cognitive processing capabilities. Unlike the physical environment, virtual environments can be tailored to individual agents. They also tend to be more flexible than physical environments. 
  • Computational Embodied Cognition: The idea that computer simulation techniques can be used to improve our understanding of embodied, embedded, extended and enactive (4E) cognition. Contemporary game engines and 3D modelling techniques provide novel opportunities to test ideas concerning the cognitive significance of information processing loops that subtend the brain, the body and the world.

Computational Epistemology

Computational epistemology is a branch of contemporary epistemology that relies on the use of computational techniques to answer questions of major epistemological significance. One area of computational epistemology focuses on the use of knowledge engineering techniques to develop machine-readable representations of epistemological concepts. These machine-readable resources can be subjected to forms of automated analysis and reasoning that seek to further our understanding of the nature, function and value of epistemological concepts. A second branch of computational epistemology focuses on the veritistic value of social practices, structures and organizations with respect to the generation of collective knowledge. Computer simulation techniques are often used to evaluate different forms of socio-epistemic system, although empirical observation using digital cognitive ethnographic and socioscopic techniques are also important. A third branch of computational epistemology focuses on the nature of our interaction with technological devices and systems. The aim, here, is to understand the extent to which technological systems can work to enhance the epistemic standing of both human and machine agents.

Key Concepts
  • Virtue Relativism: The idea that notions of intellectual virtue in contemporary epistemology (most notably in virtue epistemology) depend on whether our attention is focused at the individual or collective (social) level.
  • Second-Order Process Reliabilism: The idea that metacognitive abilities play an important role in accounting for why an agent believes the truth in cases of extended knowledge (i.e., knowledge resulting from extended cognitive processes). This potentially undermines the need for a weakening of virtue reliabilist positions in order to accommodate cases of cognitive extension. 
  • Extended Philosophy: The idea that we can make intellectual progress in philosophy by creating and manipulating machine-readable external representations. A key aim of extended philosophy is to understand how to exploit computational systems as part of the processing loops that make up philosophical thought and reasoning.

Research

Click here to view a larger version of my research interest wordle.
Research Interests