Machine Learning and Society (3 ECTS)

What it is about

You learn how machine learning influences societal structures, values and decisions and which ethical, social and technical challenges arise from it. In the end you will be able to critically assess ML systems, identify possible biases and harms, and apply approaches to alignment, auditing and safeguarding (red teaming) of ML systems.

What you will be able to do

• Critical evaluation of ML technologies and their societal impacts
• Recognition and analysis of algorithmic biases
• Understanding formal fairness definitions and their limitations
• Knowledge about alignment issues in Large Language Models
• Practical skills in auditing and red-teaming approaches
• Assessment of the limits of predictive models
• Identification and evaluation of various algorithmic harms
• Understanding ML as a socio-technical system and stakeholder roles

What the module consists of

• VorlesungIntroduction to central concepts; lectures, slides and presentations; short exercises and Q&A to secure understanding

Teaching method

• VorlesungDelivery of key concepts and discussion of case studies
• Kurzübungen/FragenEnsuring understanding and applying what was learned during the session
• PräsentationenPresentation of topics and case studies by instructors or students