Conference Explaining Machines

1st TRR 318 Conference Explaining Machines

Since 2022, TRR 318 hosts an annual international scientific conference. The first TRR 318 conference took place under the title "Explaining Machines" at Bielefeld University and focused on the social science perspective on Explainable Artificial Intelligence (XAI). Further conferences with other focal points are planned for the coming years.

Currently, one of the key questions is how comprehensible explanations of the machine processes underlying artificial intelligence can be created: should humans be able to explain how machines function, or should machines learn to explain themselves?

“The double-meaning of the name of our conference, ‘Explaining Machines,’ expresses these various possibilities: machines explaining themselves, humans explaining machines – or maybe both at the same time,” explains Professor Dr. Elena Esposito (Project B01).

The first conference took place in June 2022 at Bielefeld University and focused on research from the social sciences on explainable artificial intelligence.

As Professor Dr. Tobias Matzner (Project B03) describes: “If explanations from machines are to have an impact socially and politically, it’s not enough that explanations are comprehensible to computer scientists. Different socially situated people must be included in explanatory processes – from doctors to retirees and schoolchildren.”

Future professional conferences that will focus on the concept of explainability in AI from different scientific disciplines will be held in the coming years.

Abstract of the 1st TRR 318 Conference Explaining Machines 2022

Today's most advanced machine learning algorithms are often incomprehensible to humans, including those who designed them. How can we achieve an understandable explanation of their processes? This is currently one of the crucial questions for artificial intelligence projects. Should we be able to explain how machines work, or should the machines learn to explain themselves? The ambiguity in the title of our conference contains the coexistence of two possibilities: machines explaining themselves or humans explaining machines - or maybe both at the same time.

If explaining machines should have a social and political impact, it is not enough that they are understandable to computer science experts. Explaining machines needs to include different socially situated and diverse humans. The issues are complex and involve multiple skills. Computer scientists who design machines must collaborate with social scientists who study understanding (and the lack thereof), the process of explanation, and their conditions. Now more than ever, the challenge of artificial intelligence projects is as much social as technological, and our conference addresses this by stimulating the debate, presenting the variety of perspectives and insights developed by the social sciences in the context of XAI.