Calendar

Abstract. The Noble Lie proposed by Plato for the Just City in Republic III has been much misunderstood. Its agenda is twofold: to get the citizens of the City to see their society as a natural entity, with themselves as all ‘family’ and akin; and to get the Guardians in particular to make class mobility, on which the justice of the City depends, a top priority. Since the second is taken to depend on the first, the Lie passage amounts to an argument (1) that the survival of a just community depends on the existence of social solidarity between elite and mass, which allows for full class mobility and genuine meritocracy; (2) that this solidarity in turn depends on an ideology of natural unity; and (3) that such ideologies are always false. So the Lie really is a lie, but a necessary one; as such it poses an awkward ethical problem for Plato and, if he is right, for our own societies as well.

Presented by SWIP-NYC

Generative art made with algorithms has existed since the early days of computing in the 1960s. In recent years, a new strand of generative art has emerged: AI-generated art, which leverages the recent progress of artificial intelligence to create artworks. Unlike old-fashioned generative art, AI-generated art is not produced with an explicit set of programming instructions provided by human artists; instead, it involves training an algorithm on a dataset so that it can later produce artworks (images, music, or video clips) using its own internal parameters that have not been explicitly defined by a human. This process raises fascinating questions at the intersection of computer science, art history, and the philosophy of art. At a superficial level of analysis, AI-generated art seems to offload much of the creative impetus of art production to the machine, requiring minimal intervention from the artist. On closer inspection, however, it involves a novel process of curation at two key stages: upstream in the selection of the dataset on which the algorithm is trained, and downstream in the selection of the outputs that should qualify as artworks. Instead of replacing human artists with computers, AI-generated art can be understood as a new kind of collaboration between mind and machine, both of which contribute to the aesthetic value of the final artwork.

This seminar will bring together AI artists and philosophers to explore the significance of this new mode of art production. It will discuss the implications of AI-generated art for the definition of art, the nature of the relationship between artists and tools, the process of digital curation, and whether AI systems can be as creative as humans.

Event Speakers

• Sougwen Chung, artist and researcher
• Helena Sarin, visual artist
• Anne Spalter, digital mixed-media artist
• Katherine Thomson-Jones, Professor of Philosophy at Oberlin College
• Moderated by Raphaël Millière, Presidential Scholar in Society and Neuroscience at Columbia University

Event Information

Free and open to the public. Registration is required via Eventbrite. Registered attendees will receive an event link shortly before the seminar begins.

This event is hosted by the Presidential Scholars in Society and Neuroscience as part of the Seminars in Society and Neuroscience series.

The Center for Science and Society makes every reasonable effort to accommodate individuals with disabilities. If you require disability accommodations to attend a Center for Science and Society event, please contact us at scienceandsociety@columbia.edu or (212) 853-1612 at least 10 days in advance of the event. For more information, please visit the campus accessibility webpage.

What makes right acts right? A Stoic answer to Ross’s question.

When W.D. Ross poses the question, “what makes right acts right?” (The Right and the Good ch. 2), he is asking a question that is prior to the deliberative question, “how do I determine the right thing to do?” The Stoics recognize this: in De Officiis 1.7, Cicero says that every inquiry about duty has two parts: (1) a theoretical part concerned with the end of goods and evils, which addresses such matters as whether all duties are perfect, whether some are more important than others, and what are the kinds of duties, and (2) a practical part which sets out rules (praecepta) by which our conduct can be made to conform with the end.  This paper focuses on (1) and in particular asks Ross’s question about Stoic right actions (kathêkonta).

The endpoint of Stoic deliberation is determining what token action is the right action.  The paper begins with the Stoic distinction between a thing’s choiceworthiness, its intrinsic disposition to elicit a choice response in a suitable subject, and its possession being to-be-chosen. The determination of what is to-be-done is made by weighing against each other all the values of the relevant action types specified by their content (the so-called ‘intermediate actions’) that are in accordance with nature, as Stoic value theory says that according with nature is an objective reason to do an action.  What constitutes the rightness of the token right action, and is given in its reasonable defense, is the same as what constitutes the rightness of a perfect (katorthôma) action.   The Stoic distinction between right and perfect action depends on the action’s moral goodness—not rightness—which is due to its causal origin.

Presented by Professor Rachana Kamtekar (Cornell University)

In Sanskrit epistemology, philosophers are preoccupied with the notion of pramā. A pramā, roughly, is a mental event of learning or knowledge-acquisition. Call any such mental event a knowledge-event. In A Confection of Refutation (Khaṇḍanakhaṇḍakhādya), the 12th century philosopher and poet Śrīharṣa argued that knowledge-events are indefinable. Any satisfactory (and therefore non-circular) definition of knowledge-events will have to include an anti-luck condition that doesn’t appeal back to the notion of learning or knowledge-acquisition itself. But there is no such anti-luck condition. What is novel about Śrīharṣa’s argument is that it is motivated by his commitment to a certain “knowledge first” approach to epistemology: the view that knowledge-events are epistemically prior to other non-factive mental states and events. On this view, when we are trying to determine whether an agent has undergone a knowledge-event, we don’t initially ascribe to them some other non-factive mental event, and then check if that event meets some further conditions (like truth or reliability) necessary for it to count as a knowledge-event; rather, we treat certain mental events by default as knowledge-events until a defeater comes along.  Surprisingly, Śrīharṣa argues that this kind of “knowledge first” epistemology should give us reason to doubt whether our ordinary attributions of knowledge-events are reliably tracking any sui generis psychological kind. In this talk, I reconstruct Śrīharṣa’s position.

With responses from Rosanna Picascia (Swarthmore College)

RSVP is required for dinner. Dinner will take place at a nearby restaurant. Please contact Lucilla at lm3335@columbia.edu for further information.