Remembering with Reminiscope: Codesigning with Generative AI for Reminiscence Among Older Adults
This is a credible CHI honorable-mention contribution because it reframes generative AI from a tool for older adults into a co-design material for reminiscence, then validates that idea through two qualitative studies and a tangible artifact. The novelty is strongest at the artifact-and-practice level, not as a generalizable algorithmic advance.
Axes Lens
Rare contribution shape, typical evidence profile. The point here is not a score. It is to show what kind of claim the paper makes, and whether the evidence pattern is unusual or baseline in this 268 -review set.
Contribution shape
- Knowledge form
- descriptive knowledge typical · 92/268
- Novelty type
- artifact typical · 20/268
- Abstraction level
- artifact typical · 19/268
- Generalization target
- user population typical · 75/268
- Validation mode
- qualitative study typical · 63/268
Evidence profile
- Evidence strength
- moderate typical · 105/268
- Claim alignment
- strong typical · 231/268
- Overclaim risk
- medium typical · 210/268
Review Summary
The paper’s main value is conceptual and interactional rather than technical: it challenges the common assumption that older adults should simply consume AI-generated outputs and instead invites them to co-create memory content through a tangible, embodied workflow. That shift is visible in the combination of textile collaging, AI transformation into video, and the Reminiscope artifact for group reminiscence. The novelty is therefore not a new model or algorithm, but a new way of staging generative AI within a participatory reminiscence practice. The evidence base is appropriate for that claim: the authors report an individual co-design study with 16 older adults and a second workshop study with 15 participants, 14 of whom returned from Study 1. This supports a qualitative, design-oriented contribution about how participants experienced the process and how the artifact shaped shared reflection. At the same time, the paper’s claims should be read with care. The studies are small, context-specific, and rooted in care-home recruitment, so they do not establish broad generalizability. The prototype was also customized to the workshop setting, which means the findings are best understood as boundary-setting design insights rather than universal principles. The paper is strongest when interpreted as an HCI contribution about designing with generative AI for older-adult reminiscence, especially in group settings, and weaker if read as evidence that the approach will transfer unchanged across populations or deployment contexts. The limitation evidence is explicit: the authors note that participants were recruited from care homes, had diverse conditions and low AI literacy, and that the prototype was customized to the workshop context. That makes the work valuable as a situated design study, but not as a broadly validated intervention.
What Changed
Canon before
Prior CHI work on older-adult reminiscence and AI typically frames older adults as recipients or end-users of systems, rather than as co-designers working with generative AI as a material in an embodied, tangible reminiscence workflow.
Departure from common sense
The paper argues against the default framing of older adults as passive users of AI tools and instead positions generative AI as a co-design material in a reminiscence process that combines textile collaging, AI-generated video, and an interactive artifact for shared reflection.
Actual novelty
Reminiscope extends individual co-creation with generative AI into a group reminiscence setting via a tangible interactive artifact, including a “cinematic cabinet” that lets participants view each other’s AI-generated memory content and turns participant-made collages and videos into shared social prompts for memory exchange.
Evidence
The paper’s contribution is supported by two qualitative studies with older adults: an individual co-design study and a follow-on group workshop study using the Reminiscope artifact. The evidence shows the authors explored how participants engaged with generative AI for memory-related content and how the artifact supported shared reminiscence, but the claims remain grounded in small, context-specific studies rather than broad causal validation.
“ Reminiscope is an interactive artifact, featuring a "cinematic cabinet" that facilitates memory sharing in group settings by enabling users to view each other’s AI-generated memory content (Figure 5”
actual novelty · 3.4.1 Designing Reminiscope / 6 Discussion · confidence 0.66
“he internet.” This orientation aimed to ensure participants were familiar with the materials and procedures. Step 2: Memory Recall Exercise. The researcher encouraged the older adult to share a past memory that represented their "happiest time" or a moment they felt "most proud of". Follow-up questions, such as "why is it so special to you?”
departure from common sense · 2 Related Work / 3 Method / 6 Discussion · confidence 0.60
“[56] Emily L Mroz, Maya Shah, Hanzhen Lan, Ajua Duker, Maria Sperduto, Becca R Levy, and Joan K Monin. 2022. When Harry Met Sally: Older Adult Spouses’ First Encounter Reminiscing and Well-Being. The Gerontologist 62, 10 (2022), 1486–1495. Crossref Google Scholar [57] Elena Mugellini, Elisa Rubegni, Sandro Gerardi, and Omar Abou Khaled. 2007. Using personal objects as tangible interfaces for memory recollection and sharing. In Proceedings of the 1st international conference on Tangible and embedded interaction (Baton Rouge Louisiana, 2007-02-15). ACM, 231–238. Digital Library Google Scholar [58] Allen Pincus. 1970. Reminiscence in Aging and Its Implications for Social Work Practice. Social Work 15, 3 (1970), 47–53. Crossref Google Scholar [59] Sinikka Hannele Pöllänen and Reetta Marja Hirsimäki. 2014. Crafts as Memory Triggers in Reminiscence: A Case Study of Older Women with Dementia. Occupational Therapy In Health Care 28, 4 (2014), 410–430. Crossref Google”
limitation · 6.5 Limitations and Future Work · confidence 0.80
“ In Study 2, we conducted four rounds of workshops where Reminiscope , an interactive artifact created based on the creation of participants from the previous study, was introduced to facilitate reminiscence among 15 participants (including 14 participants from Study 1, and one joining for the first time)”
validation scope · Abstract / 3.2 Participants / 3.4 Workshop Procedure · confidence 0.72
Limits
Method limits
Evidence comes from qualitative co-design and workshop studies rather than controlled comparison; the findings are interpretive and context-bound, so they support design implications more than general causal claims.
Deployment limits
The prototype was customized to the workshop context, and the approach depends on participants being willing and able to engage with textile collages, AI-generated videos, and shared viewing in a care-home setting.
Boundary conditions
The work is bounded by care-home recruitment, older adults’ varying AI literacy, and the specific Reminiscope interaction format; transfer to other populations, settings, or reminiscence practices may require redesign.
Position in field
This sits at the intersection of older-adult HCI, reminiscence technologies, and generative-AI co-design, pushing beyond end-user adoption toward participatory use of AI as a design material and toward a tangible artifact for group memory sharing.