Man and machine as team partners

Artificial intelligence (AI) can support people in everyday life. A research project at the Universities of Kassel and Hamburg is investigating what requirements must be in place for AI-based assistance systems to be accepted and at the same time enable individualized lifelong learning by humans and AI. The project is funded by the German Federal Ministry of Education and Research (BMBF) with 1.9 million euros. 

What are the requirements for the design of AI-based assistance systems so that they are accepted by humans and offer individualized user experiences? And how can processes be designed so that AI systems learn from humans and vice versa? These two central questions will be addressed by a research team from Kassel and Hamburg over the next four years. The co-leaders of the AI junior research group "Hybridization of Human and Artificial Intelligence in Knowledge Work - HyMeKI (FKZ 01IS20057)" Prof. Dr. Eva Bittner (University of Hamburg) and Dr. Sarah Oeste-Reiß (University of Kassel) primarily want to explore the collaboration and learning processes of humans with AI systems.

The starting point is the observation that tasks in a digitized working world are becoming increasingly complex and require lifelong learning - especially in the processing of knowledge. Artificial intelligence, for example in the form of text, image or speech recognition, can help knowledge workers solve tasks or take them over completely. HyMeKI thus offers great economic potential for organizations with knowledge-intensive activities. This applies equally to the IT, media and retail location of Hamburg as well as to Kassel and the region of northern Hesse, as a location for industry and medium-sized businesses. "For the University of Kassel and the Scientific Center for Information Technology Design (ITeG) located there, projects like HyMeKI are particularly nice successes to further expand the expertise of a socio-technical system design in the field of artificial intelligence with qualified female scientists and to develop socially desirable innovations" agree the ITeG directors Prof. Dr. Jan Marco Leimeister as well as Prof. Dr. Bernhard Sick, who support the junior research group as mentors.

Merging human and artificial intelligence

An important approach of HyMeKI is that humans sometimes perceive AI-based systems as social actors and therefore have similar expectations regarding the quality of their solutions, capabilities and communication behavior. However, these expectations are often not met, leading to rejection of AI systems. "Completely new challenges arise in the collaboration between humans and AI systems, and it is important to create added value for both, as well as to make the handovers - for both the human and the AI system - smooth," says Prof. Dr. Eva Bittner. "The fusion of human and artificial intelligence in collaboration plays a key role in this, so we need innovative collaboration practices that simultaneously promote human and machine learning," adds Dr. Sarah Oeste-Reiß. The ambitious goal of the junior research group HyMeKI is to develop and test techniques and reusable socio-technical design patterns for collaboration and mutual learning between knowledge workers* and AI assistance systems - and to do so while taking into account aspects such as transparency, agency and autonomy of the participants.

For their research, the business informatics researchers will use various methods. These include case studies, surveys and observations, analyses of human-AI interaction in multimedia-equipped team labs as well as online and crowd experiments.

Background

The project is funded within the framework of the BMBF's guideline "Förderung von KI-Nachwuchswissenschaftlerinnen". The purpose of the funding is the research of AI questions on novel and innovative topics by junior research groups led by women. The targeted support of women is intended to increase the participation and influence of women in AI research in Germany. The applicants had to prevail in a two-stage, competitive selection process. A total of twenty projects were selected nationwide, which will now receive funding.

Contact

Dr. Sarah Oeste-Reiß
University of Kassel
Department of Information Systems & Scientific Center for IT Design (ITeG)
Tel: +49-561-804 6154
E-mail: oeste-reiss[at]uni-kassel[dot]de
www.hymeki.de