Economist proposes AI-first journals: AI first-pass review, 1-click replication, formal proofs
Afinetheorem · x · 2026-10-23
An economist shares his proposal, made to the AEA two years ago, for journals in the AI era: 1) AI does the first-pass technical review with authors correcting what they agree with, 2) 1-click replication packages from raw data verified by a journal 'AI', 3) open, AI-readable data, 4) formalized proofs à la MathEconLib. A quoted email from Kevin Bryan broadly agrees but finds the report too pessimistic on double-blind bias, notes AI as a 'fourth reviewer' for checking tables and proofs is nearly superhuman, and argues AI-assisted fraud makes 1-click replication and Lean-style formal proofs essential.
More from AGI Musings
- Joe Weisenthal hails AI-driven intellectual flourishing, but most casual users don't use it that way — ArtificialOther · 2026-10-23
- Reddit Post Claims It's Near-Certain a Backdoor Already Sits in Popular LLM Harnesses or Inference Engines — Ori_553 · 2026-10-23
- waitbutwhy's classic AI explosion chart: from 'monkey tricks' to 'The fuck??' — pickover · 2026-10-23
- Formal methods community's embrace of AI: automated verification was always the dream — jefrankle · 2026-10-23
- Solving aging solves everything: unlimited time means unlimited solutions, says researcher — rand_longevity · 2026-10-23
- Reddit debate: Chinese open-weight models may win the AI era without beating the frontier — StandardLovers · 2026-10-23