38% of new arXiv papers read as AI-written; CS hits 62%
kimmonismus · x · 2026-11-16
A scan of 13,500 arXiv papers (2021–2026) finds 38% of papers sampled in Aug–Oct 2026 read as machine-written, up from a 0.4% pre-ChatGPT baseline, rising every year since 2023. By field: CS highest at 62%, then Quantitative Biology 58% and EE 51%; Math at 3.7% since equation-heavy papers give the detector less prose. Tuned to avoid false positives, it misses 20% of AI text, so the real share is likely higher. Citing Dwarkesh Patel's RSI podcast, the author expects the figure to climb and asks how knowledge will be generated going forward.
Related event: Study Finds 38% of New arXiv Papers Appear AI-Written(3 posts)→
More from AGI Musings
- Academic: AI-generated recommendation letters have zero signal, replace with 10-min calls — anshulkundaje · 2026-11-16
- Bo Wang: Mathematics enters its 'ImageNet era' as AI reshapes research — BoWang87 · 2026-11-16
- BlueDot founder: the superintelligence race forces me to build a company that may only matter for years — austinc3301 · 2026-11-16
- VraserX: arguing about AI content on Instagram while AI could cure aging is a broken priority — VraserX · 2026-11-16
- Pangram can't tell AI-written from AI-edited text, sparking career-ruining false labels — TuhinChakr · 2026-11-16
- Viral quip: Building things brutally reveals which of your smarts were merely verbal — StewartalsopIII · 2026-11-16