2020/11/26 by Nadisha-Marie Aliman, Aliman, Nadisha-Marie, Leon Kester +3
Neuroscience · Social Sciences · #Artificial Intelligence (cs.AI) #Computers and Society (cs.CY) #Ethics and Social Impacts of AI #FOS: Computer and information sciences #Innovation, Sustainability, Human-Machine Systems #Psychology of Moral and Emotional Judgment
paper · pdf · doi:10.48550/arxiv.2012.02592
openalex publication_date 2020/11/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In the last years, AI safety gained international recognition in the light of\nheterogeneous safety-critical and ethical issues that risk overshadowing the\nbroad beneficial impacts of AI. In this context, the implementation of AI\nobservatory endeavors represents one key research direction. This paper\nmotivates the need for an inherently transdisciplinary AI observatory approach\nintegrating diverse retrospective and counterfactual views. We delineate aims\nand limitations while providing hands-on-advice utilizing concrete practical\nexamples. Distinguishing between unintentionally and intentionally triggered AI\nrisks with diverse socio-psycho-technological impacts, we exemplify a\nretrospective descriptive analysis followed by a retrospective counterfactual\nrisk analysis. Building on these AI observatory tools, we present near-term\ntransdisciplinary guidelines for AI safety. As further contribution, we discuss\ndifferentiated and tailored long-term directions through the lens of two\ndisparate modern AI safety paradigms. For simplicity, we refer to these two\ndifferent paradigms with the terms artificial stupidity (AS) and eternal\ncreativity (EC) respectively. While both AS and EC acknowledge the need for a\nhybrid cognitive-affective approach to AI safety and overlap with regard to\nmany short-term considerations, they differ fundamentally in the nature of\nmultiple envisaged long-term solution patterns. By compiling relevant\nunderlying contradistinctions, we aim to provide future-oriented incentives for\nconstructive dialectics in practical and theoretical AI safety research.\n