The case study discussions during AI Sweden Ethics Lab's third meeting resulted in three strategic insights that will guide future efforts in supporting the implementation of ethical AI.
The discussion was based on a presentation by Anders Westerlund and Simon Melin, Helsingborg Stad, about a tool (LAIban) that is used in pre-schools in the municipality of Helsingborg. LAIban is an AI assistant who answers questions like what the children will have for lunch, who’s picking them up and when and suggests what clothes to wear when going outside to play. The children access LAIban via an iPad and it is currently used at about 40 preschools in Helsingborg City.
Strategic key insights
The focus for the AI Ethics Lab discussions is not on how ethical issues are or should be regulated, but on what ethical aspects and additional perspectives should be considered in the implementation of the AI solution.
The following insights were derived from the group discussions, which used the presentation and use case mentioned above as a starting point.
Clarify what problem you want to solve
As with any innovative, new idea, ensuring that the solution you have in mind actually addresses a real-world problem is the first step. The issue that the AI solution aims to solve should be clearly defined and not taken for granted. This will give both developers and other stakeholders guidance when balancing risks against the value that the AI solution will bring.
Consider the different target groups
An important part of the process of evaluating the ethical aspects of AI solutions is to bring in a variety of perspectives. One way of doing this is to consider the potential implications - both positive and negative - for the target groups in question. Looking at AI applications developed for toddlers, the primary target group is of course the children that will be using the AI solution. Their needs are paramount. But there are several other perspectives to keep in mind, and including those early on is key. Parents, teachers, school administration, local authorities and others all have a stake to consider. At times, the incentives for implementation or development may differ or even clash for different stakeholders. Understanding this will lay the basis for how to prioritize and move forward.
See the bigger picture
Implementing a single solution is one thing - and another to envision it being used on a larger scale. What happens if the AI solution used in one preschool becomes the norm at all preschools in all of Sweden? It is likely that the positive effects found in the pilots will continue to be reinforced. Still, it is important to ask ourselves if societal preconceptions or boundaries may shift as a result. The theoretical concept of sociomateriality examines the social and physical aspects of introducing new technologies and becomes relevant in this particular case, in order to understand how a new generation of children will interpret a kindergarten using AI solutions on a daily basis.
Read more about the AI Ethics Lab.