Skip to Content
Last modified by Michaël Grauwde on 2023/05/02 12:29
Show last authors
1 This project will be focused on the designing of conversational agents in the domain of public safety. The project takes place in a larger project entitled AI-MAPS (AI for Multi-Agency Public Safety) which is one of the ELSA (Ethical, Legal, Societal Aspects) labs in the Netherlands. This project [[linked here>>https://www.eur.nl/en/research/research-initiatives/artificial-intelligence-multi-agency-public-safety/about-ai-maps]] focuses on the development of trustworthy AI systems in a the domain of public safety where AI can assist in the understanding and mitigating of social unrest.
2
3 The project takes three use cases into account: crowd management, high impact crimes, and public nuisance. The role of this conversational agent is to allow deliberation and reflection between various stakeholders. In this way, it will be designed to help the various stakeholders to reflect on the values that they find important within the domain of public safety. In an interaction with the conversational agent, we hope to be able to allow the human to reflect upon their values and why those values are important in the domain of public safety. The stakeholders can then based on those reflections participate in a deliberation session with other stakeholders. In this way, a more complete reflection and representation of the public's mood around AI, and particularly the values that AI should be aligned with in the domain of public safety can come to light. In the vein of aiming to create trustworthy AI, this project can function as a template or framework for other developers when developing AI tools in the domain of public safety.
4
5 This particular conversational agent that we are designing here will put the focus on a crowd management use case, particularly in the domain of climate protests.
6
7 We hope that through interacting with the conversational agent, users can be able to deliberate with their fellow stakeholders and reach various consensus in a civil manner about the topics concerning crowd management.