We summarize the first Workshop on the Design of Responsible Hybrid Intelligence (RHI2023), co-located with the 2st International Conference on Hybrid Human-Artificial Intelligence (HHAI 2022), held on Applications of hybrid intelligence systems in which humans and machines collaborate must behave responsibly in order to promote synergy and prevent unwanted or even harmful efects. For instance, the agents in a hybrid intelligence system (both human and automated) should communicate in a correct and reliable way, they should be able to provide their reasons for having a belief or opinion, and they should be able to explain their actions in terms of the values they apply. against them.
LGOBE The authors contributed equally.
Topics and issues on designing responsible hybrid intelligence we aimed to address included, but were not limited to, the following: • Methods and tools for designing responsible hybrid intelligence systems • Theoretical foundations for the design of responsible hybrid intelligence systems • Experiments with and innovative applications of responsible hybrid intelligence systems • Evaluation methods for implemented responsible hybrid intelligence systems • The role of various AI approaches in the design of responsible hybrid intelligence systems (data, knowledge, reasoning, argumentation, language, …) The aim of the workshop was to collect research and research ideas aimed at designing responsible hybrid intelligence systems. Hence we aimed to collect contributions and discussions that bridge the technological and ethical side of responsible hybrid intelligence design. For all contributors (whether technological or ethical), we asked to make explicit how and to what extent the research or research idea discussed aims to contribute to responsible hybrid intelligence design. We invited submissions of all levels of maturity (early stage, mid stage, completed). A selection was made on the basis of overall quality, relevance and diversity.
The workshop included two keynote lectures, one by Stefan Schlobach (Free University Amsterdam) and one by Lameck Mbangula Amugongo (Technische Universität München). A diverse set of papers were presented. The program was aimed at stimulating discussions between the participants, an interdisciplinary group of researchers, ranging in focus from technical implementation to societal implications of AI. The workshop concluded with a plenary discussion led by Virginia Dignum (Umeå University, Delft University of Technology) on how we could and should design Responsible Hybrid Intelligence, acknowledging the challenges and encouraging design.
The Program Committee (PC) received a total of 6 submissions. Following a single-blind reviewing process, each paper was peer-reviewed by at least two PC members. The committee decided to accept 5 short papers, containing original work.
The specifics of the program were as follows.
• Stefan Schlobach - Responsible Hybrid Intelligence Systems—a Knowledge Representation
Perspective • Lameck Mbangula Amugongo - Designing and Implementing Responsible AI
• Annet Onnes, Silja Renooij, Mehdi Dastani - Bayesian Network Conflict Detection for
Normative Monitoring of Black-Box Systems • Jonne Maas - Not a black box, but an empty one: Accounting for power in AI systems • Marco Zullich and Giovanni Santacatterina - Assessing Fairness in Open-Source Face Mask
Detection Algorithms • Tijn van der Zant - Trias Intelligentia • Virginia Dignum, Petter Ericson - Towards a feminist, relational, conception of Artificial
Intelligence
• Virginia Dignum - Moderator
Workshop Chairs • Ludi van Leeuwen, University of Groningen • Cor Steging, University of Groningen • Bart Verheij, University of Groningen Program Comittee
The organisation of the workshop was supported by the Hybrid Intelligence Center, a 10-year programme funded by the Dutch Ministry of Education, Culture and Science through the Netherlands Organisation for Scientific Research, https://hybrid-intelligence-centre.nl. The organizers would like to thank the HHAI 2023 workshop chairs and organization for providing an excellent framework for RHI2023.