Search Results → Reading Queue

Turn a large literature search into a transparent, prioritized reading queue.

AI Pill #0141 min readLiterature scansElicitRayyanResearchRabbit

This week's problem

Finding papers is easy. Deciding what deserves attention is the bottleneck. Without explicit inclusion rules, AI-assisted screening becomes a black box and important evidence can disappear behind a confident relevance score.

3 Useful AI Links

01Elicit

Find papers, screen candidate studies, and extract structured evidence for comparison.

02Rayyan

Run systematic review screening with inclusion decisions, labels, and collaboration.

03ResearchRabbit

Expand a seed set through related papers, authors, and citation networks.

Workflow To Steal

Research question + inclusion criteria
Build a seed set
Expand and deduplicate results
Screen with explicit reasons
Create a prioritized reading queue with human spot-checks

Copy-paste prompt

Act as a literature screening assistant. Prioritize the citations I provide using the explicit criteria below.

Research question: [Question]
Population / intervention / comparator / outcomes: [PICO or relevant framework]
Date and study-design limits: [Limits]
Inclusion criteria: [Criteria]
Exclusion criteria: [Criteria]
Citations or abstracts: [Paste results]
Classify each record as Include, Maybe, or Exclude.
Give one criterion-based reason for every decision.
Assign reading priority as High, Medium, or Low.
Flag records that require full-text review before a decision.
Return a table and a short list of likely evidence gaps.

Do not exclude a record because an abstract is incomplete.

Do not invent study details.

Human review remains required for consequential evidence decisions.

Best for

Literature scansJournal articlesGuidelinesMedical strategyContent development

Verdict

AI can shrink the pile. It should not hide the rules. Keep every include and exclude decision explainable, or the speed gain is not worth much.