Nov 06 – MATH+ @ Berlin Science Week
Only one week left until the start of the 4th Berlin Science Week (1 to 10 November). Scientists from Berlin and around the world will present their research and discuss it with each other and with the public.
MATH+ will be part of this year’s Berlin Science Week.
01 November, 7 p.m.
Max Klimm will battle for MATH+!
Numerous support is welcome!
Roadrunner’s Paradise, Saarbrücker Straße 24, 10405 Berlin
06 Nov, 10:00-11:30 am Panel Discussion, Natural History Museum, Tristan Lounge
How reliable and safe is Artificial Intelligence and what role does mathematics play in it?
We currently witness how AI algorithms are infusing numerous aspects of the public sector. Besides applications such as speech recognition in cell phones or self-driving cars, such approaches are also being used, for instance, for prescreening job applications or revolutionizing the healthcare industry.
A similarly strong impact can be observed on science itself, ranging from biology over engineering and material sciences to politics. However, the working horse of AI, namely deep neural networks, is still an unpredictable black box with sometimes erratic behavior; and a profound theoretical understanding is almost completely missing. This is a tremendous problem, in particular, for sensitive applications such as healthcare.
In this panel discussion, we welcome as panelists AI experts from mathematics, computer science and beyond. We will discuss which role mathematics can and has to play for developing predictable, interpretable and robust AI approaches. In the beginning each panelist will give a short statement on this topic. After some moderated questions, the audience will be invited to ask questions to the AI experts.
On the panel:
Sabine Ammon (Professor for Knowledge Dynamics and Sustainability in the Technological Sciences, TU Berlin)
Ulrike von Luxburg (Head of Cluster of Excellence “Machine Learning: New Perspectives for Science”, Uni Tübingen)
Christof Schütte (MATH+)
Sebastian Pokutta (MATH+)
Presenter: Gitta Kutyniok (MATH+)