Between Reality and Delusion: Challenges of Applying Large Language Models to Companion Robots for Open-Domain Dialogues with Older Adults-Reference-Cited by-同舟云学术

Between Reality and Delusion: Challenges of Applying Large Language Models to Companion Robots for Open-Domain Dialogues with Older Adults

Published:2023-05-05 Issue: Volume: Page:
ISSN:
Container-title:
language:
Short-container-title:

Author:

Irfan Bahar¹,Kuoppamäki Sanna-Mari¹,Skantze Gabriel¹

Affiliation:

1. Royal Institute of Technology

Abstract

AbstractThis work aims to provide initial guidelines towards developing companion robots with large language models (LLMs) to be part of everyday lives of older adults. Using iterative participatory design (co-design) approaches, we analyze the challenges of applying LLMs for multi-modal open-domain dialogue, deriving from older adults' (one-to-one) interactions with a personalized companion robot, built on Furhat robot with GPT-3.5. An initial study with 6 Swedish-speaking older adults (65 and older) showed that the robot frequently interrupted the users, responded slowly and repetitively, engaged in superficial conversations, and caused a barrier in the interaction due to foreign language (English). Upon incremental technical developments to address these issues, participatory design workshops were conducted with 28 Swedish-speaking older adults. While the interactions (in Swedish) were smoother, less disrupted, and more varied in topics and responses, further challenges were observed due to hallucinations and obsolete information, and disengagement cues, causing frustration, confusion, and worry.

Publisher

Research Square Platform LLC

Reference344 articles.

1. Zhang, Lei and Wang, Shuai and Liu, Bing (2018) Deep learning for sentiment analysis: {A} survey. WIREs Data Mining and Knowledge Discovery 8(4) https://doi.org/10.1002/widm.1253, en, 1942-4787, 1942-4795, July, Deep learning for sentiment analysis

2. Morley, Jessica and Machado, Caio C.V. and Burr, Christopher and Cowls, Josh and Joshi, Indra and Taddeo, Mariarosaria and Floridi, Luciano (2020) The ethics of {AI} in health care: {A} mapping review. Social Science & Medicine 260: 113172 https://doi.org/10.1016/j.socscimed.2020.113172, en, 02779536, September, The ethics of {AI} in health care

3. Giaccardi, Elisa and Redstr öm, Johan (2020) Technology and {More}-{Than}-{Human} {Design}. Design Issues 36(4): 33--44 https://doi.org/10.1162/desi_a_00612, en, Are we reaching the limits of what human-centered and user-centered design can cope with? Developing new design methodologies and tools to unlock the potentials of data technologies such as the Internet of Things, Machine Learning and Artificial Intelligence for the everyday job of design is necessary but not sufficient. There is now a need to fundamentally question what happens when human-centered design is unable to effectively give form to technology, why this might be the case, and where we could look for alternatives., 0747-9360, 1531-4790, September

4. Biduski, Daiana and Bellei, Ericles Andrei and Rodriguez, Jo ão Pedro Mazuco and Zaina, Luciana Aparecida Martinez and De Marchi, Ana Carolina Bertoletti (2020) Assessing long-term user experience on a mobile health application through an in-app embedded conversation-based questionnaire. Computers in Human Behavior 104: 106169 https://doi.org/10.1016/j.chb.2019.106169, en, 07475632, March

5. Christin, Ang èle (2020) The ethnographer and the algorithm: beyond the black box. Theory and Society 49(5-6): 897--918 https://doi.org/10.1007/s11186-020-09411-3, en, 0304-2421, 1573-7853, October, The ethnographer and the algorithm