The Peculiar Persistence of the AI Denialists (yaschamounk.substack.com)

A provocative essay argues that the AI moment is less a media fad than a structural break—more like the Industrial Revolution than a passing novelty—and laments widespread “AI denialism” among mainstream commentators. The author identifies three common forms of denial: claims that AI is fundamentally incompetent (pointing to hallucinations), that models are merely “stochastic parrots” without real understanding, and that AI will have limited economic impact. Despite occasional flaws, the piece stresses, modern models (ChatGPT, Claude, Grok, Gemini) already rival highly skilled humans across translation, coding, diagnostics, creative work and research summarization, so dismissing them on a few failure modes ignores their broad, demonstrated capabilities. Technically, the essay rebuts the “stochastic parrot” charge by noting that prediction-based pattern matching can produce behavior very similar to human intuition—contrasting rule-based Deep Blue’s brute-force chess with LLMs’ pattern-driven moves that even surprised Kasparov. Economically, the author warns that productivity gains require both capable AI and organizational change; adoption lags can mask eventual transformative effects (as happened after the dot-com bubble). The overall implication: the AI era poses profound risks and benefits, and treating it as merely another cultural trend risks being unprepared for deep social and economic upheaval.