Desafíos actuales de la Inteligencia Artificial

202 Desafíos actuales de la Inteligencia Artificial ment. For instance, AI-driven algorithms can optimize supply chain logistics, tailor marketing strategies, and streamline customer service operations. In public administration, AI holds prom- ise for improving service delivery, public safety, and policy implementation (Henman 2020). Predictive policing, automated administrative tasks, and smart city initiatives are examples of how AI can augment governmental functions and enhance citizen engagement. Despite these benefits, the deployment of AI systems introduces several challenges. The opaque nature of AI decision-making processes, often referred to as the “black box” problem, complicates the understanding of how decisions are made. This lack of transparency can lead to issues such as biased outcomes, accountability concerns, and ethical dilemmas. As AI systems become more integral to decision-making processes, it becomes crucial to develop mechanisms that can elucidate their workings and impacts. AI and other emerging technologies must adhere to principles of sustainability to ensure their benefits do not come at the cost of long-term environmental and societal health. The concept of sustainability, as defined by the Oxford English Dictionary (Simpson et al. 1989), is “the capacity to endure.” The Brundtland Commission further elaborates on sustainable development as “meeting the needs of the present without compromising the ability of future generations to meet their needs” (Development 1987). In the context of AI, sustainability involves ensuring that technological advancements contribute positively across various di- mensions, including individual, social, economic, technical, and environmental. Given the complex nature of AI systems and their potential impacts, there is a pressing need for robust visualization tools to map out and manage these effects. Effective visualiza- tion tools are crucial for simplifying the intricate networks of cause and effect inherent in AI deployments. They help stakeholders—ranging from policymakers to technical experts and the general public—understand the potential outcomes and implications of integrating AI into various sectors. 2.2.The Need for Visualization Given the complexity and potential risks associated with AI technologies, there is a press- ing need for effective visualization tools. Visualization serves as a powerful means to represent complex data and processes in an accessible and interpretable manner. For AI systems, visual- ization tools can bridge the gap between technical intricacies and stakeholder understanding, facilitating better decision-making and communication (Beauxis-Aussalet et al. 2021). 2.2.1. Enhancing Understanding AI systems often operate with intricate algorithms and large datasets that can be chal- lenging to grasp in their raw form. Visualization tools help in simplifying these complexities