Pedersen Presents Critique of AI Alarmism at Annual Philosophy Meeting

Posted on 9/7/2017 1:28:28 PM

Hans Pedersen, Department of Philosophy, presented a paper titled “A Heideggerian Critique of AI Alarmism” at the annual meeting of the Society for Philosophy in the Contemporary World, held in Estes Park, Colorado, July 30–August 4, 2017.

Abstract

The past several years have seen an increase in what I will call “AI alarmism”—the idea that we are on the cusp of developing artificial intelligence that is so powerful that its emergence could potentially produce catastrophic consequences for the human race. There is, of course, a spectrum of how dire these consequences are predicted to be, but the direst predictions involve scenarios seen in science fiction movies like The Terminator and The Matrix, in which a highly advanced form of artificial intelligence sets out to destroy or enslave humanity.

It is this most alarmist version of AI alarmism that I will be questioning here. My criticism of this direst form of alarmism will be loosely based on issues found in Martin Heidegger’s work, though perhaps not in the traditional way that Heidegger has been used to cast doubt on the possibilities of AI. The aim of this paper will not be to freshen the Dreyfusian/Heideggerian critique against the possibility of developing AI that can perform certain tasks. Rather, the focus will be on the “motivation” that a sophisticated AI would have to perform certain actions, particularly, the destruction or enslavement of the human race. I will argue that when we think about the likely motivational structure of AI and compare it to the Heideggerian account of the human agency, there is little reason to think that a form of AI would intentionally try to wipe out humanity.