Home

Home

795 results
  • intranet
  • deen

Artificial Intelligence

What can AI do and which previously human tasks could it take over? To what extent should it do or be allowed to do this and what consequences would and does this have?

Ein abstraktes Foto von kreisförmigen Lichtzeichnungen
Foto: Gertrūda Valasevičiūtė

Artificial intelligence (AI) is a hot topic. Its actual and potential applications are legion, for example in the fields of medicine, industry, urban planning or the application of law. The various branches of science discuss not only what AI can do and which formerly human tasks it could take over, but also to what extent it should or may do so and what the consequences might be. Each science has found its own approach to the topic of AI.

The "Artificial Intelligence" research group of Die Junge Akademie would like to approach the topic through concrete applications of AI methods and thus enter into an interdisciplinary conversation about the possibilities and limits of selected AI applications. In doing so, the RG would also like to discuss technologically, socio-scientifically and legally informed regulatory proposals for AI use.

In accordance with its interdisciplinary composition, very different perspectives determine the interest of the RG. Across the disciplinary boundaries, a positive and a normative perspective can be distinguished.

From a positive perspective, the RG is concerned with the prerequisites, potentials and consequences of the use of AI in selected areas of application. Questions include what an AI-supported evaluation of digital disease and patient data can do to make diagnoses or improve therapies. What would this look like? What requirements do AI innovations place on data quality and data generation in order to be able to fulfil their optimisation promises? At this level, the RG also considers indirect consequences of the use of AI. For example, the question arises as to which "new" tasks arise for which professional groups in order to make data available for AI applications ("data work"). How might the use of legal tech affect the legal system as a whole, like autonomously driving vehicles affect traffic as a whole? How does the use of AI in diagnosis and therapy change people's relationships within the healthcare system? Sociologically, the RG is also interested in the connection of AI with concrete utopias of optimisation and innovation. It asks about the concrete direction of these utopias, which alternatives for action they open up and close off.

From a normative point of view, the RG asks at what point AI use seems sensible and ethically and legally possible, and under what conditions. Can the problems of AI use in concrete applications be captured by regulation? In which areas should AI applications be generally prohibited? Staying with the above example: How should misdiagnoses or treatment errors resulting from AI applications be dealt with? What influence may AI-supported systems have on sovereign decisions? What requirements does data protection law impose? This level is thus primarily concerned with the regulation of AI applications. This, too, will have to be discussed primarily in relation to applications, since the use of AI applications in industrial production gives rise to different regulatory needs than their use in clinical diagnostics or for fighting crime.

2022

Workshop "Interdisciplinary Conversations on the Fairness, Explainability and Uncertainty of AI“

From 19 - 20 September, the members of the RG Artificial Intelligence will meet for an interdisciplinary dialogue on the chances and limits of AI and its applications. Different disciplines bring different perspectives to the table and similar sounding concepts do not necessarily mean the same. Moreover, even if concepts are the same across disciplines, different fields might focus on distinct aspects. The workshop “Interdisciplinary Conversations on the Fairness, Explainability and Uncertainty of AI” tries to bridge these gaps and brings researchers from different disciplines in conversation about these three major concepts of current research on AI.

At the beginning of each of three panels (representing Fairness, Explainability and Uncertainty Quantification), a presentation from the field of computer sciences or statistics will provide a conceptual basis for further interdisciplinary exchange. This presentation will be followed by presentations reflecting current research related to the respective concepts from different disciplines, such as computer sciences, political sciences, sociology, social psychology, or law. Each presentation will take the broad background of the audience that might come with different preconceptions into account. With this set up, the research group hopes to stimulate an engaging interdisciplinary debate across disciplines and applications; and hopes to encourage the dialogue between groups who substantively work on the same topics.

The workshop is organised by Nadja Klein and Pascal Langenbach.

With contributions by:

  • Sebastian Lapuschkin
  • Gabrielle Ras
  • Carlos Zednik
  • Katharina Weitz
  • Telmo de Menezese
  • Silva Filho
  • Sebastian Lerch
  • Christian Wirth
  • Asia Biega
  • Nina Grgić-Hlača

2022

Workshop “Artificial Intelligence in Biobanking”

From 29 - 30 July 2022, the Artificial Intelligence research group is organising the workshop "Artificial Intelligence in Biobanking" in Berlin. The aim is to bring biobanking and artificial intelligence closer together and to shed light on its further implications. The focus will be on aspects such as IT processes, organisational workflows or quality management measures, as well as on broader issues of ethical, legal and social significance.

The workshop is organised by Isabel Nahal Schellinger and will be held in cooperation with the Central Biobank Charité (ZeBanC).

With contributions by:

  • Dr. med. Carol Geppert (MIA) (Institut für Pathologie | Universitätsklinikum Erlagen)
  • Dr. Dr. Christian Matek, M. Sc. (Institut für Pathologie | Universitätsklinikum Erlangen)
  • Prof. Dr. techn. Wolfgang Nejdl (Institut für Verteilte Systeme | Leibniz Universität Hannover)
  • Dr. Christina Schüttler (Central Biobank Erlangen | Universitätsklinikum Erlangen)
  • Thomas Endres und Jonas Mayer (TNG Technology)

Fireside Talks

26.01.2022

Together with the research group Transfer of Innovation in Academia, RG Artificial Intelligence hosted a fireside evening with astrophysicist and Erium founder Theo Steininger. The members of Die Junge Akademie talked to him about his scientific and entrepreneurial path. Among other things, Theo Steininger addressed technical, social and legal points in the field of artificial intelligence.

23.03.2022

At this fireside talk with the Artificial Intelligence research group, the guest was Die Junge Akademie alumna Ulrike von Luxburg. Von Luxburg is a professor of computer science at the University of Tübingen and a fellow at the Max Planck Institute for Intelligent Systems. She works on the theoretical foundations of machine learning and is also interested in the influence of artificial intelligence (AI) research on future society. The members of the RG talked to Ulrike von Luxburg about her work and the potentials and challenges of AI applications for society and science. They were also interested in what experiences she has had as a scientist in the public discussion about the social significance of AI.

2021/22

Understanding blood-brain dynamics with statistical learning

The brain serves only as a cooling system for the blood - at least that's what Aristotle believed. Even though medicine and neuroscience have made some progress since then, the dynamics between neuronal activity and the fluctuations of various messenger substances and metabolic products in the blood are still insufficiently understood. The reason for this is, among other things, the complexity and effort of a parallel measurement of brain and blood parameters over several hours, and furthermore the existence of manifold interactions and feedback loops between these parameters, which quickly push classical statistical analysis methods to their limits.

The project aims to combine biomedical and mathematical-statistical expertise to decipher and understand complex data structures. For this purpose, the neuronal activity and the dynamics of various blood parameters of test subjects with balanced age and gender ratios were recorded in time with parallel EEG/blood measurements. Even in condensed form, this results in hundreds of snapshots of physiological parameters in the brain and blood per test subject and sleep lead. Accurate understanding of these interactive blood-brain dynamics requires correspondingly complex modelling and analysis as well as methods that are sufficiently scalable for such data volumes and allow relevant influencing variables and interpretable effects to be identified.

To bring together different expertise from neuroscience, medicine, statistics and mathematics, the project team, which consists of Martin Dresler, Nadja Klein, Isabel Schellinger and Timo de Wolff, is planning mutual visits and project meetings in 2021 and 2022 with participating PhDs and post-doctoral students. A joint publication is also planned.

2020

Workshop „AI and University“

On 10 and 11 January 2020, the RG Artificial Intelligence is organising its first workshop in Frankfurt am Main entitled "AI and University". On the first day, the workshop will deal with the topic "AI research despite basic data protection regulations". On the second day, the focus will be on "Using AIs in Education" (so-called Learning Analytics).

The first workshop day starts with an impulse from Prof. Nikolaus Marsch. Marsch teaches constitutional law and data protection law at Saarland University and is one of the best experts on the subject in Germany. He has habilitated on the fundamental rights and European dimensions of data protection. Prof. Dirk Ifenthaler could be won for a keynote lecture on Learning Analytics on the second day. Ifenthaler heads the Chair for Learning, Design and Technology at the University of Mannheim and is UNESCO Deputy Chair of Data Science in Higher Education Learning at Curtin University in Australia. He is also the editor of several journals and special volumes on learning analytics and heads several relevant research projects, such as the BMBF project STELA (Utilizing Learning Analytics for Study Success).

In the future, the research group aims to make its voice heard - not only, but especially - in the field of education with concrete proposals for regulation (model rules) in science policy.

Spokespersons

participating Members

participating Alumnae / Alumni

Activities

    • Workshop „Interdisciplinary Conversations on the Fairness, Explainability and Uncertainty of AI“

      Members of the RG Artificial Intelligence will meet for an interdisciplinary dialogue on the chances and limits of AI and its applications. Different disciplines bring different perspectives to the table and similar sounding concepts do not necessarily mean the same. Moreover, even if concepts are the same across disciplines, different fields might focus on distinct aspects. The workshop “Interdisciplinary Conversations on the Fairness, Explainability and Uncertainty of AI” tries to bridge these gaps and brings researchers from different disciplines in conversation about these three major concepts of current research on AI.

      Topics:

      Starts on
      19.09.22
      Ends on
      20.09.22

      Event access: Internal

      Berlin

    • Workshop "Artificial Intelligence in Biobanking"

      In diesem Workshop möchten Mitglieder der AG Künstliche Intelligenz Biobanking und Künstliche Intelligenz stärker zusammendenken. Es sollen weitere Anwendungsfelder identifiziert werden, um mit Hilfe von Künstlicher Intelligenz das Biobanking zu optimieren und in seinen weiterführenden Implikationen zu beleuchten. Dabei soll der Fokus sowohl auf Aspekte wie IT-Prozesse, organisatorische Arbeitsabläufe oder Qualitätsmanagementmaßnahmen, als auch auf übergeordnete Fragestellungen ethischer, rechtlicher und gesellschaftlicher Bedeutung gelegt werden.

      Topics:

      Starts on
      29.07.22
      Ends on
      30.07.22

      Event access: Internal

      Berlin

    • Preparatory meeting for the Workshop "Fairness, Explainability, Uncertainty & AI"

      In preparation for the workshop "Fairness, Explainability, Uncertainty & AI" in spring 2022, the members of the research group Artificial Intelligence meet via Zoom.

      Topics:

      Starts on
      15.12.21

      Event access: Internal

      online

      16:00 — 17:00

    • Medizinische Diagnostik, Polizeieinsätze, demokratische Willensbildung - nicht ohne KI

      Auf dem Sommerplenum wurde die Arbeitsgruppe „Künstliche Intelligenz“ gegründet. Was die Ziele der AG sind, klärt unser Interview.

      Topics:

      date
      06.08.19