The essay questions whether retirees can foresee paradigm shifts. As a historical check, it cites Fritz Lang’s Metropolis (1926/27): it correctly imagined videophones and humanoid robots, yet portrayed future aircraft only as propeller planes, failing to anticipate jets. Turning to today, I argued that large language models can have immediate impact in rubber science by translating and summarizing papers and, more importantly, by practically organizing the Society’s accumulated knowledge such as journals, conference proceedings, and study-group materials, into a searchable corpus, provided data leakage is prevented. I also urge recording and interviewing “rubber legends” to capture tacit know-how. On long-term in silico development, I estimate that full all-atom simulation of a Banbury mixer at centimeter scale for an hour is far beyond current computing, even with optimistic Moore’s-law extrapolation. Finally, I conclude that Japan’s “nomi-nyucation” culture may spark disruptive ideas, and LLMs could capture and distill such conversations into actionable innovations.
Mikihito TAKENAKA (2026) studied this question.