Reasoning
Major social platforms have historically shown remarkable user retention despite repeated cycles of user discontent (Facebook maintained ~3 billion users despite 2018 privacy scandals, Twitter retention despite 2022-2023 management turmoil). A 20% decline on any major service by 2029 (3 years away) would require either: (1) a coordinated mass exodus comparable to MySpace's collapse (which took ~5 years and required a superior alternative), or (2) regulatory action forcing divestment. Current AI saturation concerns are elevated but lack the network-effect breaking catalyst that previous platform shifts required. Users typically demonstrate preference-revealing behavior through reduced engagement rather than account deletion. However, if the predicted "AI assistant fatigue" materializes into concrete usability degradation on 2-3 major platforms simultaneously, and emerging alternatives (Bluesky, Threads, or unforeseen platforms) achieve critical mass adoption, this could accelerate decline. The 28% probability reflects a modest but non-negligible tail risk that AI-driven content quality deterioration + generational platform switching could compound into one major service experiencing 20%+ user loss by 2029.Key uncertainty
Whether AI-generated content degradation will trigger measurable engagement migration before platforms can implement effective content filtering—this behavioral shift pattern remains empirically untested at scale and could either materialize rapidly or remain a niche concern limited to power users.