Economist proposes hybrid human-AI peer review workflow to save reviewers time
paulnovosad · x · 2026-11-09
Economist Paul Novosad outlines what he calls the best expert paper review under current time constraints: an expert reads the intro and forms their own ideas, asks a strong AI model for its top 3 critiques, discusses and cross-checks them against the paper, then judges which critiques are substantive before writing the report. A quoted commenter adds that ideal peer review would have the human and AI work on the paper's code together, saving everyone a lot of time.
More from AGI Musings
- Same AI, different game: getting answers vs directing engineering work — burhop · 2026-11-09
- Only 18% of US firms use AI, but federal statistics can't yet track its labor impact — robseamans · 2026-11-09
- METR task-horizon debate: logistic rollover by 2027 or hyperbolic singularity? — teortaxesTex · 2026-11-09
- 7-year CV engineer on the genAI hype: where are the hard non-LLM problems? — SchwingBesen · 2026-11-09
- VC argument: in the AI slop fest, entrepreneurs hold the leverage, not the rich — StewartalsopIII · 2026-11-09
- LLMs will scale apocryphal attribution losses across the humanities — StewartalsopIII · 2026-11-09