Welcome aboard!A Holistic Approach to AIProf. Dr. Ralf Möller strengthens the Faculty of Humanities.

4 March 2024, by Möller/Red.

Photo: DFKI/Jürgen Mai

Prof. Dr. Ralf Möller

Every year, Universität Hamburg welcomes numerous new researchers. This series introduces them and their areas of research. This time: informatics expert Prof. Dr. Ralf Möller.

Prof. Dr. Ralf Möller joined Universität Hamburg from Universität zu Lübeck in winter semester and has a professorship for artificial intelligence in the humanities in the Faculty of Humanities. He heads the new Institute for Humanities-Centered Artificial Intelligence (CHAI). Has has already been doing research since 2019 in the Cluster of Excellence Understanding Written Artefacts at Universität Hamburg. He is the spokesperson there for the research field F (data linking) and member of the board, the ethics working group, and the ethics council.

My research area in 3 sentences:

The long-term goal of the research on artificial intelligence, or AI, is the development of interactively operating systems that we see as “actors” or “agents.” They should be able to suitably interpret, on the basis of explicitly or implicitly described tasks and internal models of their environment, so-called “dynamically received sensory input.” This could be, for example, a task description. They should also be able to react to these perceptions and determine and execute actions in the world to solve problems posed to them as best as possible.

Because these actions always take place in a social context, however, the interpretations of task descriptions must always be further developed, logically by the agents themselves, for example by providing humans with preferences.

The research on AI as a science thus includes the analysis of social contexts—also called “social mechanisms”—with, if possible, many agents and their mutual interactions. And the goals, restrictions, and societal regulations also need to be taken into consideration.

This is how I explain my research to my family:

The science of artificial intelligence involves asking informatics-related questions to develop interactive technical systems that, to some extent, imitate human abilities such as perceiving the environment, learning, planning, creativity, and purposeful action. These abilities should be used to support people’s activities using intelligent technical systems.

The main thing is to facilitate what we think of as well-designed interaction between people and systems. Well-designed means in this case that human preferences must be suitably accounted for in the social mechanism. Knowledge about AI is used in different areas to increase efficiency and create new possibilities, for example, in medicine, in traffic systems, and in the industry.

In Hamburg, the city, and the University, I am looking forward to:

Hamburg is a wonderful city with an excellent research university and unimagined possibilities for knowledge exchange that we are just now discovering with other AI researchers in Hamburg.

These are my plans at Universität Hamburg:

AI systems have the potential to change the humanities in the long term, for example, by deriving new knowledge from very large text corpora and providing researchers with the information. The systems, meaning the agents, are as impacted by the researchers as they are from previous findings in the humanities. This is the only way for them to perform as desired.

Therefore, my research focuses on the interactive training of intelligent agents using the abilities of humanities’ scholars, for example, to process text and images. At the same time, we have to school researchers in the humanities to avail themselves of technology such as language processing or the automatic translation, summarizing, or linking of texts so that the limits and possibilities of current and future technologies can be duly appreciated, used, and further developed by selecting training material for special tasks.

To do this, we have established a new Institute for Humanities-Centered Artificial Intelligence (CHAI) at Universität Hamburg. The goal is to work on basic issues in the field of artificial intelligence and we work on the assumption that it is only by integrating findings in the humanities that we can make a quantum leap in AI research.

On the one hand, we want to focus on the possibilities of AI technologies, for example, when doing research on human culture on the basis of historical and contemporary sources that are highly complex and comprehensive. On the other, we want to concern ourselves with the implications of AI for our pursuit of knowledge, for example, with regard to strategies for evaluation distortions and recommendations for tried-and-true procedures.

This is why students should attend my courses:

Our world is strongly shaped by the interplay between humans and technical systems and is currently undergoing a time of change that makes the program extremely exciting now. Studying the consequences of these developments for the economy, science, and society also means mastering new challenges.

Students will learn to use findings from philosophy and psychology in the useful development and handling of AI. Current achievements in AI are no longer based on imitating people but on developing new, high-performing multimodal technologies. So it’s about improving extant technological systems so that they are trustworthy and can interact with people such that the distinct human abilities to take in information and make decisions and the malleability of human world views are taken into consideration.

My research is important to society because: 

The humanities play a crucial role in the development of intelligent systems. By researching topics such as language, culture, ethics, and history, the humanities provide important findings and perspectives that are indispensable to the form and use of intelligent systems.

Linguistics, a sub-field of the humanities, is, for example, vital to the development of language-recognition systems and machine translation. Ethics research, on the other hand, contributes to the development of ethical guidelines for using AI technologies. This is important for ensuring that intelligent systems used in a social mechanism work in harmony with moral principles.

Thanks to our special focus at CHAI and the incorporation of all of these areas of humanities expertise in designing intelligent systems, we will contribute to the holistic development of AI technologies.