Generative AI Compensates for Age-Related Cognitive Decline in Decision Making (arxiv.org)

Researchers tested whether generative AI can offset age-related declines in decision-making by running a controlled experiment with 130 participants (56 younger, 74 older) who completed a music-selection task in two contexts (familiar “road trip” and novel “space travel”). In the AI-nonuse condition participants generated options from memory; in the AI-use condition GPT-4o produced preference-aligned recommendation lists. Cognitive function was measured with the Wechsler Adult Intelligence Scale–Fourth Edition (WAIS‑IV). Results showed that older adults with lower cognitive scores reported more choice difficulty and lower satisfaction when AI wasn’t used. Introducing AI recommendations significantly reduced perceived choice difficulty for both age groups and weakened the link between lower cognitive function and worse decision outcomes; overall choice satisfaction did not decline with AI assistance. The study suggests generative models that produce aligned, user-specific option sets can compensate for age-linked deficits in memory, processing speed, and information search—reducing cognitive load without hurting satisfaction. For ML practitioners and HCI designers this highlights the value of preference-conditioned generation and lightweight recommendation interfaces for older users, while flagging practical issues: ecological validity beyond a lab music task, dependence on model fidelity (here GPT‑4o), transparency, overreliance risks, and the need for robust preference elicitation and evaluation metrics in deployed systems.