Ethics in AI: Addressing Bias and Lack of Diversity

Neuland

AI@HPI Conference 2024 (2)

As artificial intelligence (AI) increasingly integrates into various aspects of our lives, addressing the ethical implications of its use becomes more crucial. Algorithmic decisions are often made at such a massive speed and volume that human oversight is no longer practical. In this panel, experts from various disciplines will explore how biased data and a lack of diversity in development can lead to discriminatory outcomes. This is particularly troublesome for recent AI models with many millions of parameters that are far too complex to be interpretable by humans. The discussion focusses on how biases arise in datasets and algorithms, the ethical standards that need to be established, and how to create fair, inclusive, and transparent AI systems.