Linguist Mocks LLM-Generated Research Papers With 'Supercazzola' Parody
preslav_nakov · x · 2026-11-15
NLP researcher Preslav Nakov shared a parody passage imitating how LLM-generated research papers sound, calling it "supercazzola" — Italian for words that sound sophisticated but mean nothing. The image spins classic AI academic clichés into gems like hallucinations "reclassified as anticipatory references," humans-in-the-loop whose job is "to click regenerate until the limitations become contributions," and "further work is needed, preferably by someone else."