Transparency is a concept that currently has almost universally positive connotations: What is transparent is transparent, understandable, publicly accessible and comprehensible. From architecture to action in political spheres, transparency is seen as a positive feature and traded as a counterpart to the hidden and inaccessible. In the context of the use of artificial intelligence, transparency has become a criterion that algorithms are supposed to fulfil in order to be able to clarify questions of attributing responsibility and liability.
In an interdisciplinary workshop, we would like to discuss the role of transparency as a manageable criterion from historical, technical, legal, techno-philosophical and ethical perspectives.
A first part will provide the broad view of transparency and contrast it with the application-related demand to ensure transparency of algorithms in such a way that it can be adjudicated legally.
In a second part, the more >technical< perspective on possible interpretations of transparency against the background of so-called “black box algorithms” will be discussed, and fundamental questions about the feasibility of the transparency requirement and possible consequences will be raised from the perspective of technology assessment.
A third part chooses medicine as an illustrative area of application: What does transparency mean in the application of technology in medical practice and what role does it play for acceptance both on the part of those being treated and on the part of medical staff? What conclusions can be drawn from this for the demand for transparency?
The workshop is aimed at researchers who, on the one hand, want to place politically and ethically motivated demands such as the creation of algorithm transparency in a larger context and, on the other hand, want to include the concrete feasibility in the consideration. The event will be held hybrid.
Programme
Thursday, 09 March 2023
14:00: Part 1: Transparency as an ideal and as a criterion
14:15-15:30: “Transparency as a key concept and challenge for democracy and privacy” – Dr. Lea Watzinger (Graz)
16:00-17:15: Transparency as a regulatory criterion – Prof. Dr. Hannah Ruschemeier (Hagen) Part 2: On the Operationalisability of Transparency
17:30-18:45: “AI thinks differently. Why is that?”- Prof. Dr. Markus Kollmann (Düsseldorf)
Dinner
Friday, 10 March 2023
10:00 – 11:15 a.m.: Transparency from a practical and regulatory perspective: A contribution to AI technology assessment? – Prof. Dr. Karsten Weber (Regensburg) Part 3: The demand for transparency in medicine
11:30 – 12:00: Transparency in Medicine – Introductory Reflections – Prof. Dr. Heiner Fangerau (Düsseldorf)
Lunch break
12:45 – 14:00: “Transparency in medical AI: An epistemological foray” – Dr. Thomas Grote (Tübingen)
14:15 h: Closing