ľ�ϸ���Ӱ��

Social robots, smart speakers, Google Home, or Amazon Alexa: they are making their way into our homes. But what role do these forms of artificial intelligence play within the family? Do parents and children see them as tools, little brothers or sisters, or more like pets? Ruud Hortensius and his team of researchers are exploring this question.

Tool, little brother or pet

To understand the role AI plays in families, researchers compare three different types of groups: families that already use AI, families that do not, and families with a pet. "Do families see their virtual assistant more as a tool? Or is it given a more social role, for example, as a little brother or perhaps more like a pet? To find out, we look at the conversations families have. Does their language suggest that they see the robot as a social being? But also, how does this new technology influence family dynamics?" Hortensius started this project a year ago, and three PhD students are now also working on it.

Cargo bike lab

How do you study something as complex as dynamics? Hortensius explains: "We have a mobile laboratory, which we take to the families' homes in a cargo bike. To properly map the dynamics, we also measure brain activity. Parents and children wear a kind of swimming cap to measure brain activity using a technique called functional Near-Infrared Spectroscopy (fNIRS)." This technique works like a kind of torch: it shines invisible light waves onto the brain and then captures them again. The amount of light we capture is determined by the amount of oxygen in the brain areas, and more active brain areas have a higher oxygen content. "This technique allows us to see which areas of the brain are active at the same time in different members of the family."

Once the bathing caps are on, the experiment can begin. "We let the families play with a social robot and observe how this affects the dynamics. The families play different games in their own living rooms to create as realistic a setting as possible. We look at social dynamics, such as cooperation and empathy. They play games such as Mikado and card games. We also look at the conversations the families have. How do they talk about the AI?"

Synchronous

“We investigate how connected families are at the level of brain activity by comparing how closely the brain activity of different family members matches each other.” All kinds of combinations can be made here: the brain of parent 1 with child 1, the parents with each other, or the children with each other. "By taking all this together, we can conclude that one family is more synchronised with each other than another family. You can also do this at the behavioural level. Ultimately, we are trying to answer the question of how AI changes this dynamic."

In addition to the fNIRS data, the researchers are also collecting video and audio recordings. "We study the interaction between family members and the robot we bring along. How does that work? How do they respond to each other, and how do they respond to the robot?" Both the recordings and the brain activity are an implicit measures, which means that there is no need to consciously ask about certain things, as is more often the case in a questionnaire or interview. "This is particularly useful when working with children, who are not always able to complete a questionnaire." The combination of brain and behavioural data and self-reporting also provides a unique insight into the complexity of family dynamics. Hortensius is enthusiastic about all the methods they use as a research group: "It's a candy store full of data."

Dynamics of Youth

The strategic theme Dynamics of Youth plays an important role for Hortensius in this project: "The subject of our research naturally ties in well with the theme of DoY and also builds a nice bridge between DoY and human-centred AI." The team was also pleased with the DoY network. "We were able to learn from the approach of other researchers working on completely different topics, such as the Luister project (ed.: giving children a voice in court during divorce proceedings). When we were setting up our research, we were able to brainstorm with them about collecting data within families."