This study examines the 2025 psychiatric collapse of a Portuguese academic whose critiques of academia’s systemic decay (2007–2021) culminated in creating an AI-generated heteronym network (e.g., “Prof. António Carvalho”) using LLMs. Diagnosed with dissociative identity disorder (Dibunker, 2025), his personas produced radical critiques of “algorithmic solipsism” and “necrotic scholarship,” mirroring his existential terror and the institutional pathologies he denounced. The crisis climaxed at the AI-Pocalypse Now conference (ISCTE-IUL, May 38th, 2025), where he discovered he was both keynote speaker and subject of his heteronym “Carvalho’s” critique, triggering an ontological collapse fusing author, subject, and critic. Through forensic analysis of his psychiatric evaluation and this catalytic event, we argue his clinically diagnosed “stochastic parroting” delusion constitutes a profound methodological indictment: the heteronyms exposed academia as a self-referential void where algorithms produce unread papers, metrics quantify fictive impact, and human scholarship degenerates into higher-order stochastic parroting (Bender et al., 2021). His trajectory reflects the economic pathologies of academic capitalism under algorithmic dominance, akin to Bostrom’s “paperclip problem” (2014) and Piccione & Rubinstein’s resource appropriation models (2007). Consequently, the psychiatric prognosis of “guarded” applies equally to the academy. We contend institutional survival hinges on forging “therapeutic alliances” with algorithmic systems – embracing rather than resisting AI to transform necrotic scholarship into fertile ground for posthuman epistemic renewal.

--

LLM,Academia,Patafisica