Beyond Interpretability: An Interdisciplinary Approach to Communicate Machine Learning Outcomes

Explore an innovative approach to Explainable AI (XAI) in this 28-minute talk by Merve Alanyali, PhD, Head of Data Science Academic Partnerships at Allianz Personal. Discover how Alanyali's team collaborates with the University of Bristol to examine AI decision-making through a socio-technical lens, moving beyond traditional technical interpretations. Gain insights into their interdisciplinary method for explaining machine learning outcomes, drawing from Alanyali's extensive experience in both academia and industry. Learn about text-to-image models, the evaluation of cultural competence in AI systems, and the importance of cultural diversity as a new aspect of AI assessment. Delve into the future path of XAI research and its potential impact on the field of artificial intelligence. This talk is ideal for professionals and enthusiasts in machine learning, data science, and AI who seek to broaden their understanding of interpretability and communication in AI systems.

- Introduction
- Text-to-Image Models
- Evaluating Cultural Competence
- Cultural Diversity: A Brand New Evaluation Aspect
- Path Ahead