Digital game-based learning has been found to be among the most effective teaching approaches in higher education in several meta-analyses. In some studies, digital games have been found to achieve more learning in one hour than courses that extend over an entire semester – including in ethics education. An area of ethics education of growing importance concerns the use of artificial intelligence (AI), particularly machine-learning (ML) applications. ML algorithms are prone to a number of ethical risks of which the general public is still unaware, including the possibility of discriminatory advice and decisions-making. At the same time, ML applications like ChatGPT, DeepL, or DALL-E have become so powerful that millions of people have begun to use them.
One area where decision-makers need to be aware of ethical risks of ML and AI applications is business, e.g., in view of algorithms for personalized marketing. To sensitize learners for related risks, we have developed the serious game “CO-BOLD” at Leuphana University Lueneburg. Learners slip into the role of a quality manager at a big tech company. Together with their team, they assess the goodness of an AI-based digital assistant for financial consultants.
The game engages learners in a realistic and suspenseful investigation of an AI application, which sensitizes them to the function of ML applications, their strengths, and shortcomings. Players’ success in detecting the AI’s shortcomings and addressing the main risks determines, which of the games’ different endings they will reach – with positive, problematic, or even tragic outcomes.
At the workshop, participants will be able to try out and discuss parts of the game and learn about how it can be used in class. In tests with 100 students, we have found that CO-BOLD sensitizes learners for relevant issues, and that it is perceived as highly engaging. We hope to attract people who want to use the game in their future teaching – free of charge.