Economist slams AI-written papers: Formulas are all wrong
joshgans · x · 2026-08-17
Johan Fourie analyzes the impact of AI on academic publishing, noting a high rejection rate due to the poor quality of LLM-generated submissions. He highlights issues with formulaic structures, incorrect math where authors fail to verify AI outputs, and instances of placeholder results being submitted.
The article mentions that the AI reviewing tool Refine has partnered with the American Economic Association to implement technical verification before publication, signaling an academic shift towards using AI to combat the flood of low-quality content.
More from AGI Musings
- AI research forms a self-reinforcing loop — ingliguori · 2026-08-17
- FT: Next China shock will come from open-source AI — pstAsiatech · 2026-08-17
- Embrace bots instead of blocking them as agents evolve — Daniel_Farinax · 2026-08-17
- AI can reduce top causes of death without needing AGI — davidmanheim · 2026-08-17
- Claude's writing differs from humans due to simultaneous perception and larger memory — Sauers_ · 2026-08-17
- High Energy Physics papers surge 30% signaling AI acceleration — Dr_Singularity · 2026-08-17