Machine Reason

By: Lynn Holt   Clint Johnson   Ross Mc Cool   Jonathan Barlow  

Machine learning is on the verge of re-discovering Aristotle. This is doubly ironic: while pursuing the four hundred year old algorithmic ideal of reason, the new wave of deep learning not only takes a two millennium retro turn, it rejects the anti-Aristotelianism animating early modern thinkers. But the Aristotelian quality of deep learning should not give hope to Luddites and atavists. On the contrary, we hope that it spurs humanists to engage and contribute to the frontiers of artificial intelligence. Our four person project has aims at three distinct but interwoven levels: Theoria. To embed the history of machine learning in the wider history of the early modern calculative re-conception of reason, to articulate the distinctively apprehensive nature of Aristotelian reason, to show how calculative reason and machine learning requires apprehensive reason to accomplish its own aims, and to examine the transformation of objectivity which results. Techné. To explore innovative ways of constructing and informing artificial neural networks with the goal of instantiating machine analogues of Aristotelian intellectual “virtues”, and to break out of the replacement paradigm with circular modes of human-machine interaction. Praxis. To put theoria and techne to the test in a practical and open-ended setting in the domain of career counseling for teens and young adults. We aim to make an existing app smarter and make it interact with human counselors. In addition to brief presentations on theoria, techné, and praxis, our colloquium will include a propaedeutic on algorithmic machine learning with a focus on neural networks.

Digital humanities Philosophy
Critical Cultural Studies

Lynn Holt

Professor, Philosophy, MIssissippi State University, United States
MS, United States

I've been a full time academic in philosophy since 1987. I started at Centenary College of Louisiana in Shreveport, moved to the University of the South at Sewanee, and am now at Mississippi State University. Perhaps the hallmark of my research and teaching has been inter- and transdisciplinarity. I've published on ancient Greek philosophy, early modern astronomy, eighteenth century metaphor, and contemporary ethics, nonlinear dynamics, and social psychology. I've researched and taught with faculty in literature, religion, classics, archaeology, physics, computer science, and chemistry. My persistent focus in all of this is on the various notions of reason, rationality, and expertise embodied in these disciplines and sometimes professed by their practitioners. I've been around long enough to have had two separate stints as department head in two different departments for a total of 11 years in administration, though I have purposely stepped away both times to return to teaching and research, and that is what I do now. Most recently, my research has been focused on artificial intelligence, not as a model for human intelligence, but sui generis. I've been working with some colleagues who, though they have humanities PhDs, are earning their living in the land of AI and data analysis.

Clint Johnson

Senior Research Scientist, Algorithms and sensor optimization, Georgia Tech Research Institute, United States
United States

Ross Mc Cool

Graduate Student, Department of Psychology, Mississippi State University, United States
MS, United States

Jonathan Barlow

Associate Director, Software Design and Architecture, National Strategic Planning and Analysis Research Center, United States
United States