On Monday, June 26th, 2023, Professor Vasilis Kosteas, along with Konstantinos Charonis and Yannis Kotsanis, conducted a workshop on the Elements of AI Ethics for two International Baccalaureate (IB) classes at Doukas School. This workshop was organized as part of the Theory of Knowledge course, and it utilized materials developed by the research and innovation department of the school. The initiative was carried out within the Erasmus+ research project FACILITATE-AI.
This workshop builds upon prior activities focused on research and source identification to support an argument. In small groups, students engage in research using articles and videos that reveal ethical pitfalls in an area of Artificial Intelligence (AI) of their choice. Subsequently, each group formulates at least one principle aimed at providing a solution to the identified issues in their selected domain.
For the workshop, a lesson plan was created through a Padlet where the activities that the students participated in were embedded, with several auxiliary material and with the ability to capture their thoughts and answers to it.
The activities included:
An example of a student’s comment that was added in the Padlet is: As it is widely believed, in order for one to be ethical, logic, understanding of society’s circumstances, as well as empathy/emotional intelligence are needed. When it comes to artificial intelligence, other than logic and understanding due to unlimited access to the internet, there’s still a great lack of emotional intelligence. Therefore, even though AI is still not capable of fully grasping the idea of ethics, one could say that the needed form of intelligence will grow simultaneously as AI adapts to human behaviour.
One of the conclusions of the workshop was that here is a fundamental distinction between ethics and ethical behaviour: ethics can be defined as a set of values, principles and rules which regulate human life and behaviour; ethical behaviour, on the other hand is the expected outcome of rule-following in the confines of a certain cultural framework. Behaviour is linked to social interaction and as such cannot be applied to AI. But the actual question is not whether AI can be ethical (which obviously cannot – even in the extreme scenario where can become both producer and acquirer of knowledge, unless – somehow – is able to socially interact in a meaningful way), as whether the rules and regulations that determine operational systems running on AI are designed in accordance with a certain ethical (even cultural) framework.
The workshop had a significant impact on the students, as they actively participated in the activities and engaged in thoughtful reflections on the ethics of AI. This positive response was evident in the evaluation, where the majority of students responded positively to questions regarding the clarity and alignment of the learning objectives, their motivation throughout the lesson, and the opportunities provided to develop new competences. One improvement suggestion from the students was to schedule the workshop at the beginning of the academic year.