Social RoboticsOlder adults set out what they want from an LLM companion robot
KTH co-design workshops put memory, privacy and discretion at the top of the list.
Bahar Irfan, Sanna Kuoppamaki and Gabriel Skantze at KTH ran co-design workshops with 28 Swedish-speaking adults aged 66 to 86, showing them a Furhat robot powered by GPT-3.5 and discussing everyday scenarios. Participants wanted the robot to listen actively when they were alone but stay quiet in company, to remember personal history, and to protect privacy with the ability to delete data. Many doubted it would ease loneliness for people with other social options and found its voice and face too flat. The authors turn these views into design recommendations for foundation-model-based companion robots.
Source: Frontiers in Robotics and AI
This archive entry summarises reporting published elsewhere; follow the source link for the full story. The summary was drafted with AI from that reporting.