Edward A. Parson, UCLA School of Law, Alona Fyshe, University of Alberta, and Dan Lizotte, University of Western Ontario, have published Artificial Intelligence's Societal Impacts, Governance, and Ethics as UCLA School of Law PUblic Law Research Paper No. 19-44. Here is the abstract.
AI’s societal impacts and governance challenges carry high stakes but are deeply uncertain, due to uncertainties in both the characteristics and capabilities of the technologies and in how they will be developed, deployed, and used in social and political context. We provide an introduction and synthesis of discussions at the 2019 Summer Institute on AI and Society, an intense interdisciplinary exploration of these issues. The summer institute used an experimental model to promote informed discussions among a diverse international group of researchers, professionals, and students: one day of focused briefings, followed by two days of collaborative group work on projects proposed and selected by participants via a real-time decision process. The workgroups that emerged from this process examined a wide range of questions: embedded values in automated mobility systems; the concept of meaningful human control in complex human-machine decision systems; the prospect of AI systems replacing policy analysis in government decision-making; how AI challenges and changes conventional understandings of agency; the feasibility and requirements for AI to drive large-scale societal progress; and two distinct perspectives on the multiple sets of proposed AI ethical principles and attempts to make these operational: one a practical attempt to develop and specify such an operationalization, and one a critique of an existing operationalization recently proposed by an EU expert group. Workgroup discussions offered new insights on the risks posed to corollary or implicit values in present institutions or systems that may be inherent in automating or codifying decisions; the frequency with which seemingly prosaic matters of AI system design directly implicate deep political and moral issues; the value of a dialectical or adversarial approach to assessing impacts of AI systems or governance efforts; and the acute need for methods to assess and manage medium-term AI developments and impacts.
Download the article from SSRN at the link.
Comments
You can follow this conversation by subscribing to the comment feed for this post.