ChatGPT audits 40,326 blog posts to name Marginal REVOLUTION's weakest arguments
joshgans · x · 2026-11-01
Economist Alex Tabarrok published an AI-generated critique on Marginal REVOLUTION, using ChatGPT to scan the site's full archive of 40,326 posts and shortlist the ones with the weakest reasoning.
- The methodology is notably careful: it only picks posts where there's a concrete gap between evidence and conclusion, explicitly disclaims being an objective ranking, and separates "poor reasoning" from "low value."
- The first named target is Tyler Cowen's 2022 post "Democracy and the Democrats": the AI argues it conflates two logically distinct positions — worrying about democratic breakdown vs. refusing to accept electoral defeat — and applies unequal evidentiary standards to Democrats vs. classical liberalism.
- The AI flags its formulations as illustrative rather than quotations, showing decent anti-hallucination discipline.
A rare public example of using an LLM to bulk-audit years of one's own writing, doubling as a test of LLM argument-evaluation ability.
More from Models
- Users Report ChatGPT Image Generation Suddenly Downgraded: Distorted Faces, Prompt Drift — RumbleversePlayer · 2026-11-01
- Google readies Chief of Staff voice agent for Antigravity, possible launch next week — testingcatalog · 2026-11-01
- Reddit user details Codex disaster: broken UI and botched logic in Unity game port — AcademicMistake · 2026-11-01
- Nvidia engineers detail Open Agent Safety Platform's three-layer approach to agent containment — LuizaJarovsky · 2026-11-01
- Dev rants: Google's gated model releases are training users to hate them — ricklamers · 2026-11-01
- Why labs may be hiding Lean math results: teaching AI to break encryption — rickasaurus · 2026-11-01