Semester Kick-off of Kristiania AI center activities

Abstract:
"Artificial Intelligence is becoming an essential part of decision-making across industry, helping organizations forecast markets, optimize their product offer, and thereby secure their success. However, many of today’s AI systems operate as ‘black boxes’: they can produce highly accurate predictions without clearly explaining how or why they reached their conclusions. For users, this lack of transparency can make it difficult to trust AI-driven decisions, particularly when they have significant economic consequences.
This talk explores why explainability is a key ingredient for the successful adoption of AI in industry. Rather than simply identifying which pieces of data influenced a prediction, modern approaches can combine AI with expert knowledge about real-world events and their relationships, providing explanations that are more meaningful and easier for people to understand. Using examples from the automotive sector, the talk will illustrate how explainable AI can help engineers, managers, and decision-makers better understand AI recommendations, build confidence in their use, and support more informed decisions. The presentation will also discuss the opportunities and remaining challenges in developing AI systems that are not only as accurate as possible, but also transparent, trustworthy, and aligned with human expertise."

Simon Schramm is a Data Scientist within Corporate Strategy at the BMW Group, leading technical and research initiatives at BMW involving knowledge-graph-based scenario modelling, MLOps infrastructures, and strategic market forecasting. His research focuses on explainable AI, knowledge graphs, neurosymbolic AI, and interactive machine learning, with a particular emphasis on explainable time-series forecasting. He is also a PhD candidate in Applied Computer Science at the University of Bamberg and Alongside his academic work. With an interdisciplinary background spanning machine learning, stochastic engineering, statistics, and economics, his work combines cutting-edge AI methodologies with real-world industrial applications and decision-making.