Human-Centered Trustworthy AI / zu Modul CS5076-KP12

 

Lecturer:
Prof. Dr. Nele Rußwinkel, Rebecca von Engelhardt

Content:

• Introduction to perceived trust and trustworthyness in a system, in an interaction, in an artificial interaction partner
• Introduction into the field of Human Centered AI
• How to construct the understanding of our environment / system mechanism / interaction partner?
• Beliefs, Expectations & Intentions – What is Human-Aware AI?
• Structured Representations – Human-in-the-loop & Situation Awareness
• Interacting with “intelligent” Systems – Autonomous Driving – Agency – Task Analysis
• Different Types of Mental Models – e.g. Person Models
• Transparency – Beneficial and Human Compatible AI
• Control – Sense of Control – Shared Control – Teaming (Team awareness) – Team Fluency
• Coordination – Collaboration – Joint Action – Role of Expectations
• Bounded Rationality – Real-Time Systems – Heuristics – Workload Capacity
• Turn Taking – Perspective Taking
• Explainability – Explicability
• Intentions – Anticipation – Theory of Mind – Trust
• Challenges for Explanation on Demand
• Reaction to Mistakes and Misalignments

Start of the lecture:
First semester week

Additional information:

• This course is only available for the Online AI Master!
• This lecture is part of the module Human-Centered Trustworthy AI CS5076-KP12.
• Further information (scripts, organisation of the exercise groups, exercise material, etc.) on the lecture can be found in the corresponding Moodle course at the University of Lübeck. You must enrol in the course yourself to access it. Note that there are separate Moodle courses for the modules' other parts.