Philosophy Discussion Club: Trystan Goetze, Sue G. and Harry E. Bovay Program
Friday, November 10, 2023 3pm to 5pm
About this Event
232 East Ave, Central Campus
Trystan Goetze, Director of the Sue G. and Harry E. Bovay Program in the History and Ethics of Professional Engineering
AI Art is Theft
In this paper, I engage with and formalize a set of arguments developed in a protest movement against generative artificial intelligence, focusing on the case of visual art. Since the explosion of interest over the last year in AI technologies that can produce novel texts, images, and music from user-provided text prompts, there has been much public discussion of the potential benefits and dangers of these technologies. At the forefront of this discourse are professional artists and other creatives, who (correctly) predicted that generative AI may be used by corporations to replace their labour. In response to this threat, artists have engaged in various protest actions, from the circulation of anti-AI arguments online, to pressuring art sharing platforms to ban or disincentivize AI-generated images, to legal cases that are now pending before the courts. The specific protest argument that I concentrate on in this paper, central to the anti-AI movement, is that the use of pre-existing artworks to train AI algorithms constitutes a form of theft. I start by analyzing what “art theft” may mean, determining that the primary sense in these arguments is theft of labour. I then formalize an ethical argument produced in the anti-AI protest community, using a Lockean framework for understanding the ethics of intellectual property and creative labour. I then engage with an objection, which claims that the way in which AI algorithms draw upon earlier works is relevantly similar to how human artists learn and practise their own craft. My response to this objection employs three mutually supportive parts: First, Locke’s proviso that anyone who appropriates material from the commons must leave “enough and as good” for others. Second, an argument that the speed and scale at which generative AI draws upon existing works makes it ethically distinct from similar human practices. And third, that the pattern of data extraction employed by generative AI developers follows a pattern nearly identical to resource extraction in the historical colonial era, which points to injustices in the development of these applications and the distribution of benefits from their use. I close with some implications of the conclusion that AI art is theft for other uses of AI.
Discussion Club is a lecture series hosted by the Sage School of Philosophy
Event Details
See Who Is Interested
0 people are interested in this event
User Activity
No recent activity