The Foolish Narrative of AI, 2015–2026: A Retrosynthetic Account of a Decade of Category Error

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Between 2015 and 2026, a specific technical artifact — the transformer, asequence-to-sequence next-token predictor — was progressively rebranded as cosmicintelligence, then as superintelligence, then as a strategic national asset. The rebranding was not a mistake. It was the product. This paper traces the arc chronologically, identifies the four physical walls the narrative collided with (energy, water, hallucination, liability), and examines the two 2026 events that mark the narrative's terminal phase: the executive rebranding of "AI" to "Superintelligence" on September 29, and the 120-day autonomous weapons project announced on September 30. The paper's claim is narrow: a five-year-oldcan see the category error, and the professionals who could not were not confused — they were compensated.