Ben Recht's seminar: all polls are models as statistical prediction blurs into sampling
beenwrekt · x · 2026-10-22
In a live-blogged lecture from his graduate seminar "Forecasting: A Critical Retrospective," Ben Recht argues all polls are models, examining how survey designers avoid interacting with actual people. He explains the weak law of large numbers behind random sampling (error scales inversely with the square root of sample size, independent of population size) and how statistical prediction blurs into statistical sampling when surveying society.
More from Research
- Hands-on: building protein design workflows with RFD4-Proteina on Hugging Face Spaces — AllThingsApx · 2026-10-22
- 'Agentic Inequality' paper published in Nature Machine Intelligence: who captures AI's surplus agency — IasonGabriel · 2026-10-22
- 158 of 188 RFD4-Proteina-designed TrkA binders hit sub-micromolar affinity in the lab — AllThingsApx · 2026-10-22
- Team spends $5K benchmarking GPT, Claude, Gemini on LIBERO-10 — best success rate just 44% — YuXiang_IRVL · 2026-10-22
- ToM2Vec: continuous theory-of-mind tokens improve multi-agent coordination — UNC-ChapelHill · 2026-10-22
- From 10 nodes to elaborate structures: neural cellular automata art at EMERGENCE — zzznah · 2026-10-22