Antrittsvorlesungen von
Univ.-Prof. Dr. Jörg Peter Lücke
Univ.-Prof. Dr. Radu-Aurel Prodan
Institut für Informatik
Mittwoch, 29. April 2026
17:00 Uhr
Großer Hörsaal an der Technik, 6020 Innsbruck
Um Anmeldung wird bis spätestens 20. April 2026 gebeten.
Jörg Lücke received his doctoral degree from the Ruhr-University Bochum, Germany, in 2005, where he worked at the Institute for Neuroinformatics.
After his doctoral research, he joined the Gatsby Computational Neuroscience Unit, UCL, UK, as a senior research fellow. With grants from different funding agencies, Jörg Lücke then established his own lab at the Goethe-University Frankfurt in 2008, and later moved the lab to the TU Berlin. Since 2013, Jörg Lücke has been an associate professor for Machine Learning at Oldenburg University, Germany. In 2021 he received a call from a Bavarian University to become full professor of Theoretical Machine Learning, which he declined to become a full professor at Oldenburg University. In 2025 Jörg Lücke joined the University of Innsbruck as full professor in Computer Science and as head of the Artificial Intelligence Lab.
Abstract
Our time is probably the first time in human history that experiences the emergence of an artificial alternative to human intelligence. This alternative, commonly termed artificial intelligence, is having a profound impact on science, education, the economy and society.
Generative AI, in particular, with its many positive as well as many negative aspects, has a very strong transformative impact; and AI-generated data such as texts and images are often taken as evidence for "super-human" abilities of current AI systems. In my talk I first provide a critical view on generative AI, and highlight how it relies on very prosaic foundations: big computers and big data. However, both these foundations are currently reaching their limits. In my talk I, therefore, highlight how the research in my own and other labs allows for addressing the two limitations. First, when "big data" is available, I show our recent results on so-called sublinear learning algorithms which enable highly efficient training. Concretely, I show recent prototype algorithms that can learn large data representations (with billions of parameters) on relatively small computers in very short times (hours). This contrasts with mainstream representations of similar size, which usually require supercomputers and weeks of training.
Secondly, when big data is unavailable, I present results on stronger, smarter models that we and others have developed. Example applications of all models are shown for different tasks and on data such as images (including medical images), sounds and medical data. For all presented research I refer to my own research in Machine Learning and AI over the past years, and relate it to the general developments in the field of AI.
Radu Prodan holds an endowed professorship in Edge AI at the Department of Computer Science, University of Innsbruck, Austria, co-funded by the Austrian Research Promotion Agency (FFG), Land Tirol, Wirtschaftskammer Tirol, Industriellenvereinigung, and 10 local companies.
Between 2018 and 2025, he held a professorship in distributed systems at the University of Klagenfurt. He received his PhD in 2004 from the Vienna University of Technology and his tenure in 2018 from the University of Innsbruck. His research interests include AI methods and tools for performance, optimization, and resource management in distributed and parallel systems. He participated in numerous national and European projects and coordinated three European projects, securing funding of over €7.5 million. He authored over 300 publications and received three IEEE Best Paper Awards.
Distribute and Learn 1 Billion
The digital world hosts approximately 6 billion Internet users, representing about three-quarters of the global population, with over 8.5 billion mobile edge device subscriptions as the primary access method. Bringing nearly everyone online generates over 400 billion gigabytes of daily data, overwhelming the processing and storage capacity of modern supercomputers with millions of processing cores, operating at over 1 billion gigaflops per second. The emergence of GPU accelerators triggered the modern AI boom, enabling deep neural networks with billions of neurons and parameters to outperform humans in specific, supervised, structured tasks, such as internet search, image recognition, or speech detection. Today, edge devices like Iot industrial sensors, smart cameras, networking routers, switches, gateways, wireless access points, and consumer retail devices like smartphones, laptops, and healthcare equipment equipped with powerful neural processing units exhibit peak performance up to trillions of operations per second for specialized AI inference, comparable to that of parallel computers 20 years ago. The presentation provides an outlook on the research activities of the endowed professorship in Edge AI for learning on large numbers of distributed, interconnected, memory-constrained devices. The research addresses real industrial needs of 10 local funding companies and three large multimillion Horizon Europe projects: Graph-Massivizer using sampling for compressing graph-structured data with billions of nodes and trillions of edges for training and inference in graph neural networks, DataPACT addressing AI compliance according to the EU regulations (with over 19,000 total legal acts in force) using federated knowledge distillation, and CoAgent, resporting AI pipelines and distributed small language models for reasoning.
Um Anmeldung wird bis spätestens 20. April 2026 gebeten.
Esther.gezzele-lechner@uibk.ac.at
Esther Gezzele-Lechner
Sekretariat
Forschugsgruppe für Künstliche Intelligenz
Institut für Informatik
Fakultät für Mathematik, Informatik und Physik
Universität Innsbruck, Österreich
Technikerstraße 21A Raum 3M12
ICT Gebäude
+43 512 507-53257
https://informatik.uibk.ac.at/
Esther.gezzele-lechner@uibk.ac.at
Datenschutzhinweis
Im Rahmen dieser Veranstaltung können Fotografien (oder Screenshots) und/oder Filme erstellt werden. Mit der Teilnahme zur Veranstaltung nehmen Sie zur Kenntnis, dass Fotografien und Videomaterialien, auf denen Sie abgebildet sind, zur Presse-Berichterstattung verwendet und in verschiedensten (Sozialen) Medien, Publikationen und auf Webseiten der Universität Innsbruck veröffentlicht werden. Weitere Informationen zum Datenschutz entnehmen Sie bitte unserer Datenschutzerklärung unter: www.uibk.ac.at/datenschutz
Wir bitten um eine umweltfreundliche Anreise! Mit den öffentlichen Verkehrsmitteln, zu Fuß oder mit dem Fahrrad. Fahrradabstellplätze stehen ausreichend zur Verfügung. Informationen zu den öffentlichen Verkehrsmitteln finden Sie unter www.vvt.at