General Purpose Artificial Intelligence (GPAI) systems that can perform a growing number of tasks at near-human levels show great promise. My mission is to explore how current and future GPAI systems can radically accelerate scientific research, transform biomedical discovery and healthcare delivery, and ultimately contribute to human health and flourishing.
There are two complementary aspects to the research of our group. On the one hand, we aim to identify untapped opportunities and remove barriers to positive developments, such as the acceleration of biomedical research and the development of new, life-saving therapies. On the other hand, we seek to identify and mitigate potential large-scale, systemic risks associated with these powerful systems, such as the breakdown of established processes of scientific evaluation and oversight, or the empowerment of malevolent actors.
The current focus of our group is to examine how public and institutional policies can guide the effective and responsible development and deployment of GPAI. In this context, I am currently co-chairing the EU General-Purpose AI Code of Practice drafting process (Working Group 2, concerned with risk identification and assessment of frontier models).
My work is based on a cross-disciplinary understanding of both state-of-the-art AI and biomedicine, including large language models, knowledge representation, personalized & translational medicine, and regulatory/ethical aspects. I was co-initiator and component leader of the 15 million € Horizon 2020 project “U-PGx: Ubiquitous Pharmacogenomics”, which involved the deployment and evaluation of computational decision support solutions for personalized medicine in seven European countries, encompassing widely different healthcare institutions, regulatory frameworks and thousands of patients.
Our group is part of the Institute of Artificial Intelligence at the Medical University of Vienna, Austria.