SafetyExplainabilityEthicscs.CY
AI safety: state of the field through quantitative lens
Mislav Juric, Agneza Sandic, Mario Brcic
February 12, 2020
International Convention on Information and Communication Technology, Electronics and Microelectronics
29 citations
2 influential
Key Takeaways
-
Research on AI safety has accelerated dramatically since 2015, indicating growing recognition of the need to address safety concerns as AI systems become more powerful and widely deployed across society.
-
Explainability remains an unsolved technical challenge with uncertain long-term utility, meaning we still lack reliable methods to understand how AI systems make decisions—a critical gap for building public trust and accountability.
-
Value alignment is identified as the most important long-term research priority, reflecting concerns that AI systems may not reliably pursue goals aligned with human values and interests without significant additional work.
-
Policy research severely lags behind technical research, with the paper noting a "severe lack of research into concrete policies regarding AI" despite AI's expected role as a major driver of societal change—suggesting urgent need for governance frameworks to accompany technical advances.
-
Many fundamental technical issues in AI safety remain open problems, indicating the field is still in early stages and that current approaches may be insufficient for ensuring AI systems remain beneficial as they become more capable and autonomous.
Abstract
Last decade has seen major improvements in the performance of artificial intelligence which has driven wide-spread applications. Unforeseen effects of such mass-adoption has put the notion of AI safety into the public eye. AI safety is a relatively new field of research focused on techniques for building AI beneficial for humans. While there exist survey papers for the field of AI safety, there is a lack of a quantitative look at the research being conducted. The quantitative aspect gives a data-driven insight about the emerging trends, knowledge gaps and potential areas for future research. In this paper, bibliometric analysis of the literature finds significant increase in research activity since 2015. Also, the field is so new that most of the technical issues are open, including: explainability with its long-term utility, and value alignment which we have identified as the most important long-term research topic. Equally, there is a severe lack of research into concrete policies regarding AI. As we expect AI to be the one of the main driving forces of changes in society, AI safety is the field under which we need to decide the direction of humanity's future.