Designing Conversational AI for Aging: A Systematic Review of Older Adults’ Perceptions and Needs

Research Background and Issues

What problems or challenges did the authors identify?

1. Limitations of the target population: Existing studies often focus on a small, specific group of older adults or particular devices, neglecting the broader understanding of conversational systems among the elderly.
2. Technical challenges: The rapid development of conversational AI may make it difficult for older adults to keep up with technological changes. Moreover, current systems exhibit significant issues in data privacy, language comprehension, and interaction quality.
3. Social factors: Older users express fears about technology, including concerns about losing autonomy due to reliance on it.

Why is this issue important?

With the global aging population increasing and growing attention to the technological needs of older adults, the AgeTech market is projected to reach $2 trillion. Conversational AI not only has the potential to enrich the lives of older adults but also plays a critical role in areas such as health management and loneliness prevention. Therefore, it is essential to ensure that these technologies are tailored to the specific needs of older adults.

Research Motivation and Related Work

The authors point out that, although there have been reviews of conversational systems and the technology acceptance of older adults, their study is the first to analyze conversational AI design approaches and user experience requirements from the perspective of older adults, explicitly exploring future technological directions. Additionally, this paper provides a deep understanding of design challenges and opportunities through a systematic review of existing literature, calling for more inclusive and personalized technology design.

Solutions

What methods or solutions did the authors propose?

1. Systematic literature review: The authors conducted a comprehensive screening of relevant literature from 2010 to 2024, initially selecting 720 studies, narrowing them down to 86 articles relevant to the research objectives, and focusing on 18 highly relevant papers for in-depth analysis.
2. Thematic analysis: Based on the experiences and needs of older users, several themes were extracted, including interaction barriers, privacy concerns, and expectations for future systems.

What is innovative about this solution?

Unlike previous studies, this paper adopts the perspective of older users, refining their specific requirements for future conversational AI, such as emotional understanding, cultural awareness, and personalized interaction design. These insights extend beyond technical implementation to include user-technology interaction, social acceptance, and ethical considerations.

Implementation Steps

1. Paper selection: Conduct systematic searches across three major databases (IEEE Xplore, ACM Digital Library, and PubMed) to ensure the inclusion of academic literature on relevant topics.
2. Qualitative coding analysis: Develop a coding framework to extract key themes based on the technical types, methodologies, participant characteristics, and interaction experiences with AI described in the papers.
3. Key theme extraction: Identify multiple focus areas, including interaction barriers (e.g., repeated commands and inaccurate outputs), privacy and trust issues, and expectations for future technologies.

Research Findings

What specific findings were achieved?

1. Interaction barriers:

   • Older adults often worry about how to initiate interactions (e.g., proper use of wake words).
   • Systems struggle to understand older users' accents, phrasing, and cultural contexts, leading to poor interactions and frustration.
   • Responses lack emotional flexibility, making interactions feel cold and mechanical.

2. Privacy and trust:

   • Older adults have significant concerns about data privacy, such as the transparency of data collection and resistance to information leakage.
   • Conversational systems in the healthcare domain are suspected of being unable to provide reliable information.

3. Future system requirements:

   • More "human-like" interactions: Support dynamic adjustments in interaction behavior, such as addressing user backgrounds and remembering conversation content.
   • Emotional responses and cultural awareness: Including the ability to understand diverse language styles and contexts.
   • Personalization options: Allow users to customize voice styles, communication tones, and how the system handles personal data.

What advantages does it have compared to existing solutions?

Compared to traditional conversational systems, this study suggests shifting the design focus toward:

• Acceptance by older users: Optimizing interaction quality through cultural adaptation.
• Privacy control: Designing systems that provide users with direct control over data transparency and functionality switches.
• Multimodal interaction: Incorporating diverse methods such as visual and textual interaction in addition to voice, to personalize user experiences.

Experimental or evaluation results

By analyzing 18 key papers, the study clarified older adults' expectations for future AI technologies and the pain points of existing technologies. These findings provide clear design directions for future technological improvements, although the study itself did not include independent experimental results.

Limitations and Future Directions

Limitations:

1. The literature review primarily focused on three major databases, potentially overlooking relevant studies from other fields.
2. The study relied mainly on existing technological examples, which may limit the scope of user feedback.
3. It is challenging to completely eliminate selection bias and reporting bias.

Future Directions:

1. Design and evaluate further customized conversational systems to meet the needs of older adults.
2. Explore how to embed emotional intelligence and cultural sensitivity into AI interactions.
3. Conduct long-term studies to monitor the actual benefits of technology use on health management and social participation.

Through this systematic review, the study highlights the rich design potential and challenges of conversational AI systems in the context of large-scale aging. This provides a clear path forward for technology developers, designers, and academic researchers.