The course examines the continuity between classical Platonic metaphysics and contemporary questions about the nature of artificial intelligence. Plato laid the groundwork for many of the fundamental ways in which later traditions inquired into the nature of what is and of being. Beyond the classic distinction between being and becoming, Forms and spatiotemporal entities, he also established an understanding of nature through geometric structures (Timaeus), an account of the world and its paradigm in the categories of living beings (Timaeus, Sophist), and he prompted debates about the status of artifacts (Republic, Cratylus). Artificial intelligence introduces a new kind of entity that on the one hand resembles tools and technologies, yet on the other shares certain characteristics of living beings, such as the capacity for adaptation and learning. AI displays, in a different physical substrate, capacities traditionally (exclusively) attributed to humans: it writes, reads, and speaks; composes poems and songs; and uncovers new knowledge. How should we understand artifacts that come to life? How are we to understand logos, which in the case of humans depends on an evolved biological structure, but in the case of AI "speaks" by means of artificial neural networks? How should we understand culture, science, and social structures that now also operate in a silicon body? Do the successes of deep neural networks vindicate Plato's geometrization of the cosmos? We will also investigate the linkage between Platonic metaphysics and AI through the concept of intelligence, whichin the form of nous and logosshapes the basic contours of Plato's ontological scheme, and whichin the form of instrumental rationality and reinforcement learningconstitutes the backbone of contemporary artificial neural networks. To develop this connection, we will also draw on the work of biologist Michael Levin and his reflections on the ideal underpinnings of both biological and artificial intelligent systems.
|