MPI-SP brings a diverse research lineup at this year’s CHI Conference on Human Factors in Computing Systems
The CHI Conference on Human Factors in Computing Systems takes place this year from April 26th to May 1st in Yokohama, Japan. Scientists from MPI-SP are presenting several studies at this event covering a broad range of topics from contact tracing to Virtual Reality advertising.
Contact Tracing – A New Cultural Perspective
Contact tracing has been crucial for reducing the number of infections during the Coronavirus pandemic. While digital contact tracing has been widely studied in the Western world, little is known about the relevance and application of contact tracing in Africa. To fill this gap, a team of international scientists, including employees from MPI-SP, conducted a field study in two African countries: Kenya and Côte d’Ivoire. They investigated the implications of wearing a physical contact tracing device for healthcare workers and populations from rural areas. The study highlights the barriers to adopting such devices, including low awareness, misconceptions, and social stigma. This paper also presents lessons learned from participatory design sessions, which showed that people prefer devices that are discreet, easy to wear, and that can be assimilated with accessories with cultural meaning (e.g., traditional bracelets).
How might consent, credit, and compensation protect the needs of creative professionals?
Creative workers have experienced a lot of harm due to the fact that generative AI models were trained on their original work without consent, credit, or compensation. The “3 Cs” – consent, credit, and compensation- have been proposed as a framework to mitigate these harms. The current study investigates whether this 3 Cs framework can protect the creative workers from 3 major fields: visual art and design, writing, and software development. The results of 20 semi-structured interviews show that creative workers wish to be consulted about the usage and governance of AI, expressing the need for regulations in differentiating between “AI-made” and “human-made work”. They also find that the role of the 3 Cs is more nuanced than current recommendations suggest. Based on the outcomes of the study, the authors from MPI-SP and Oxford University discuss in this article their recommendations for appropriately implementing the 3 Cs to guide efforts for responsibly training generative AI systems on creative work.
Young family members safeguard Chinese seniors from online fraud
RedNote is a Chinese social media platform similar to Instagram. An international team, including scientists from MPI-SP, performed an inductive thematic analysis of posts and comments on this platform that described users’ experiences with protecting seniors from online fraud. Seniors are targeted by many forms of online fraud, which might exploit their health concerns and their need for social relationships. The study revealed that younger members of the family step in and provide prevention, identification, persuasion, loss recovery, and education of fraud. The study also highlighted the challenges that the younger generation faces, especially due to the cultural aspect of filial piety: the seniors guard information about being a fraud victim, refuse help, and distrust their family members.
The relationship between systemic injustice and perceived algorithm fairness
Artificial Intelligence is trained on human decisions; thus, it is as biased as humans are. A great example to illustrate this is the use of AI to pre-select candidates for job interviews: The systemic injustice that people from minority populations experience in the hiring process is also equally present in the AI-based process. A new study with participants from the United States investigated whether educating participants about the systemic injustice of AI reduced their trust in such algorithms. The results shed a surprising light on people’s perceptions: participants from historically advantaged groups reported negative attitudes to the use of AI, while those from disadvantaged groups became more positive about the use of AI. The study also demonstrates how framing biased AI decisions in different ways can influence public opinions about AI.
*This work was recognized with the Best Paper Award Honorable Mention
How do Virtual Reality (VR) users feel about advertisements on the VR platforms?
Advertisements are the dominant way for firms to reach out to their customers in the digital age. With Virtual Reality (VR) becoming a new way of entertainment and interaction, companies are experimenting with introducing advertisements to the immersion experience. Scientists from MPI-SP and the University of Michigan conducted 22 semi-structured interviews to investigate how users experience ads in VR and what their attitudes towards VR advertising are. They found users thought advertising could be a good way to reward VR artists and VR creators. Having said that, users expressed numerous concerns, including future ads being “forced” on the user, along with privacy risks (VR ads can collect eye-tracking data) and physical harms (e.g., overstimulation, motion sickness, etc.). Moreover, users are very worried about how VR ads would impact the immersion of VR. If ads are to be introduced into VR, carefully balancing user immersion and enjoyment should be of top priority.