In addition, AI introduces another layer of complexity as even many experts lack a clear understanding of the workings of many systems that are in use. This problem is further complicated by the increasingly abstract nature of the digital world where there are even more layers of complexity that can be used to hide what is going on. This is especially true for systems that interact with or interface with the public at some level as it is not always obvious to the public how the process works or they might not have the expertise needed to engage with a service. As governance becomes more complex, which is the case with most developed societies, it is often hard in democracies to engage the public even in the basics of governing processes such as voting. rapid advance: how the system performance is controlled.Īccording to this list of challenges it seems that the most demanding goal is to understand the system and to gain the knowledge to control its performance.Ī human-centric AI transformation in the governance of public services requires participation from the public to ensure that the governance works as intended. ( 2016) list the challenges in building AI systems as follows:Ĭost of adoption vs. That is, to help individuals prepare, understand, accept and embrace AI, and maybe even to refuse the use AI. This is something that the governments need to deal with: to see that the potential of AI can flourish. One of the key questions is whether we accept the benefits of AI if we do not totally understand the impacts of the technology on society and citizens. Holton and Boyd ( 2021) argue that “while human consciousness retains distinctive features, these do not support an anthropocentric perspective on human–machine interactions.”
The answer is not straightforward, but a great fear is that the human viewpoint remains distant in AI applications. Holton and Boyd ask where the people are in these elements of AI. They collect and process both structured and unstructured data and make inferences based on this data. To achieve a given complex goal, AI systems observe the environment, acquire data, and make inferences and decisions based on the data and information. Central areas to AI include machine learning, natural language processing, computer vision, speech recognition, planning and scheduling, optimization, robotics, and expert systems (Holton and Boyd, 2021 Pietikäinen and Silvén, 2021). AI systems are built on a variety of methods. We often hear talk of weak and strong AI, the latter of which has not even been implemented in practice. In this development, the human focus should be on values that emphasize humanity and the understanding of the social impact of AI.ĪI is difficult to define, and no single universally accepted definition has been established within the scientific community. It will ultimately change the structures of society and everyday lives of people in a profound way. Introduction: AI and Sociotechnical ChangeĪI has a great promise to offer solutions to the problems of humankind. We draw particular attention to the efforts made by the AuroraAI Ethics Board in deliberating the AuroraAI solution options and working toward a sustainable and inclusive AI society. With the help of this case study, we investigate the challenges posed by the development and use of AI in the service of public administration. To concretize this discussion, we study the co-development of a Finnish national AI program AuroraAI, which aims to provide citizens with tailored and timely services for different life situations, utilizing AI. We examine the ethical issues and the role of the public in the debate on developing public sector governance of socially and democratically sustainable and technology-intensive societies.
#DIGIME CREATE PDF HOW TO#
This paper contributes to the debate on how to develop persuasive government approaches for steering the development and use of AI. How can the public sector use AI ethically and responsibly for the benefit of people? The sustainable development and deployment of artificial intelligence (AI) in the public sector requires dialogue and deliberation between developers, decision makers, deployers, end users, and the public.
0 kommentar(er)
