The role of experts in the public perception of risk of artificial intelligence
- PDF / 1,031,252 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 2 Downloads / 174 Views
OPEN FORUM
The role of experts in the public perception of risk of artificial intelligence Hugo Neri1 · Fabio Cozman1 Received: 22 May 2019 / Accepted: 31 October 2019 © Springer-Verlag London Ltd., part of Springer Nature 2019
Abstract The goal of this paper is to describe the mechanism of the public perception of risk of artificial intelligence. For that we apply the social amplification of risk framework to the public perception of artificial intelligence using data collected from Twitter from 2007 to 2018. We analyzed when and how there appeared a significant representation of the association between risk and artificial intelligence in the public awareness of artificial intelligence. A significant finding is that the image of the risk of AI is mostly associated with existential risks that became popular after the fourth quarter of 2014. The source of that was the public positioning of experts who happen to be the real movers of the risk perception of AI so far instead of actual disasters. We analyze here how this kind of risk was amplified, its secondary effects, what are the varieties of risk unrelated to existential risk, and what is the dynamics of the experts in addressing their concerns to the audience of lay people. Keywords Artificial intelligence · Social impacts of artificial intelligence · Risk · Risk perception · Experts
1 Introduction In this paper, we apply the social amplification of risk framework [SARF] (Renn et al. 1992; Kasperson et al. 1988, 2003; Pidgeon et al. 2003) to the public perception of artificial intelligence [AI] using data collected from Twitter. A major finding presented in this paper is that we identified that experts are the real movers of the risk perception of AI. Public awareness of Artificial Intelligence has been growing since the second half of the last decade. This is a new wave of popularization of the term AI. The first wave occurred at the beginnings of the discipline, during the 1960s and early 1970s. This new wave is relevant because the phrase “artificial intelligence” became popular (i.e., Twitter users tweet more about AI) after 2014, which coincides with experts’ public interventions. In this paper, we take the set of public messages available on Twitter that explicitly used the phrase “artificial intelligence” from 2007 to 2018 as an abstract representation * Hugo Neri [email protected] Fabio Cozman [email protected] 1
Center for Artificial Intelligence, University of São Paulo, São Paulo, Brazil
of the public awareness of AI. A portion of this public awareness refers to the anticipation of likely negative consequences related to the variety of applications of AI as a technology. We call that as the public perception of risk of AI or merely the risk perception of AI. Even though the risk perception is just a fraction of the public awareness of AI, according to our data, its rate of growth is higher than the public awareness excluding the risk perception (Fig. 2). Classical risk analysis takes the concept of risk as the statistical expectation of unwante
Data Loading...
