IRIS Forschung

Project Focus

Using thought experiments as a context for creating and exploring a novel artificial intelligence (AI) system that captures dynamics of human cognition in the form of mental models. A representation and prediction system is trained to emulate the basic causal chain-of events occurring when humans engage in thought experiments. Basic operations are cast into simple dynamical rules and objects, defined in a similar way as in cellular automata and
inspired by production rules in symbolic cognitive architectures, using an autoencoder-like architecture for enhanced abstraction. Potential outcomes of thought experiments can thus be “played out”, generated by the physical dynamics, and iteratively refined. These “AI augmented” thought experiments provide a direct means for self-reflection and critique of one’s mental models (i.e., unconscious biases and presumptions), and stimulate a creative process. At the same time, this enables building an AI model consistent with our own understanding. Cross-disciplinary implications are discussed upon evaluating the system from the perspectives of cognitive science, physics of non-equilibrium, complex systems, as well as for AI performative aspects like knowledge transfer.

Project Members

• Dr. Miriam Klopotek, SimTech Cluster of Excellence
• Jun.-Prof. Dr. Maria Wirzberger, Interchange Forum for Reflecting on Intelligent Systems
• Max Weinmann, Institute for Computational Physics
• Patrick Egenlauf

Project Focus

Online political discussions may help to weigh arguments and shape opinions potentially including considerations from all stakeholders. Thereby, discussion platforms may act as intelligent brokers for a fruitful exchange of opinion. In practice, resentful discussions abound and often they disparage specific groups from equal participation. In this research project, we plan to develop methods of natural language processing and network data analysis to identify 
such unfair treatment. Using the Reddit Politosphere dataset as a case study, we will provide specific statistical analyses and evaluation of developed methods. Based on such methods and analyses, we plan to inform discussion platforms about problems and potential means of containment.

Project Members

• Jun.-Prof. Dr. Raphael Heiberger, Computational Social Science, Institute for Social Science
• Prof. Dr. Steffen Staab, Analytic Computing, Institute for Parallel and Distributed Computing (IPVS) and Institute for Artificial Intelligence
• Sara Hanke, Institute for Social Science
• Farane Jalali, Institute for Parallel and Distributed Computing (IPVS)

Project Focus

The construction industry needs to urgently become more productive and sustainable. Automation is a common approach to increase efficiency. But this approach is challenged by a lack of qualified workers in the construction industry. This project at the intersection of architectural computing with social science addresses this challenge with a novel, AI-based method of human-robot collaboration that (1) replaces demands for human physical labor with demands for technical skills and that (2) engages workers by stimulating creativity and ensuring agency. Shifting from physical endurance to professional input and intellectual contributions opens construction to broader demographics that are otherwise excluded and allows them to contribute more meaningfully and with higher-value skills.

The project integrates perspectives from feminist technoscience—a transdisciplinary field offering distinct ways of thinking about societies, technologies, bodies, power, and environments—to develop a HRC method that attracts skilled staff and supports skills and decision-making while considering uncertainties inherent in construction. Specifically, the project uses methods from experimental democracy to ensure a fair and responsible development process for the HRC method that engages a diverse set of potential users

Project Members

• Prof. Dr. Cordula Kropp, Institute for Social Sciences (SOWI) and Centre for Interdisciplinary Risk and Innovation Research at the University of Stuttgart (ZIRIUS)
• Prof. Achim Menges, Institute for Computational Design and Construction (ICD)
• Tenure-Track-Prof. Dr. Thomas Wortmann, Institute for Computational Design and Construction (ICD)
• Gili Ron, Institute for Computational Design and Construction (ICD)
• Amelie Schreck, Institute for Social Sciences (SOWI)

Project Focus

The project investigates the potentials and challenges for democratic systems that come with the proliferation of generative AI applications (gAI) within the realm of public spheres and political deliberation. It focuses on three central aspects: First, the epistemic dimension of gAI both as a source of knowledge in political debate as well as its increased potential to further disinformation. Second, the ethical dimension of the quality of training data scraped from public sources and its implications for the representation of social groups and political opinions within these data and thereby also within answers provided by gAI models. And third, the democracy theoretical dimension of gAI applications as “participants” of political discourse and its repercussions for public deliberation and public reason.

With a genuinely interdisciplinary work plan, the project is able to analyze these three dimensions from a philosophical as well as a machine learning perspective and to discover crucial limitations as well as conditions for institutional and technology design to reap possible benefits of the technology. The overarching goal of the project is to develop and submit an interdisciplinary grant proposal (VW, ERC) for a larger project on the topic of “Generative AI and Public Reason”.

Project Members

• Prof. Dr. Christian Martin, Institute for Philosophy
• Prof. Dr. Mathias Niepert, Institute for Parallel and Distributed Systems
• Prof. Dr. Wulf Loh, Institute for Philosophy
• Dr. Marvin Tritschler, Institute for Philosophy

Project Focus

Online platforms connect individuals from diverse backgrounds across the globe, fostering a sense of global community. However, this vast reach also carries a potential risk: the formation of echo chambers that inadvertently reinforce existing stereotypes and contribute to polarization. This project uses the revision history of the collaboratively edited wikiHow1 platform to investigate the extent to which such a risk can be evidenced by measurable dynamics and how artificial intelligence methods can be used to mitigate polarization. Specifically, we study instructional texts written in a version for women and a version for men and examine how audience-specific changes over time contribute to making articles more polarizing and in how far such changes actually suit the needs and preferences of the corresponding target groups. Finally, we test whether large language models (LLM) can remedy undesired polarization effects by merging articles written for specific audiences into more balanced articles for a general audience. We bring together methods from computational linguistics and experimental psycholinguistics to carry out the project and to measure its success.

Project Members

• Dr. Michael Roth, Institute for Natural Language Processing
• Jun.-Prof. Dr. Titus von der Malsburg, Institute of Linguistics
• Nicola Fanton, Institute for Natural Language Processing
• Sidhath Rajan, Institute of Linguistics

Project Focus

Feedback enables the maintenance of motivation and the correct self-assessment of learners' individual progress. As individual tutoring by instructors is not always possible due to limited resources, Intelligent Tutoring Systems (ITS) can supplement such instructors to a certain extent. However, ITSs so far only adapt to the status of the learner, but not to the learners' diverse backgrounds (e.g., culture, gender, and socioeconomic status). Furthermore, gamification can be a suitable tool to increase learners' motivation. To investigate the differential impact of adaptive feedback and gamified learning elements on diverse learners in terms of learning outcomes, learner acceptance, and learning motivation, an ITS with these characteristics will be developed in this project. Such an adaptive and gamified ITS aims to address diverse students' needs and thus reduce course heterogeneity, thereby offering new opportunities for inclusive teaching. Therefore, this project forms the basis for future research on the ethically guided usage of ITS in classrooms, cyber-human learning, and the trustworthiness of AI in education.

Project Members

• Jun.-Prof. Dr. Maria Wirzberger, Interchange Forum for Reflecting on Intelligent Systems (IRIS)
• Prof. Dr.-Ing. Steffen Becker, Institute of Software Engineering
• Niklas Meißner, Institute of Software Engineering