About the Project
\n
¿Tiene las cualificaciones y habilidades adecuadas para este trabajo? Descúbralo a continuación y pulse en \"solicitar\" para ser considerado.
\n
Pareto.AI is a human data-collection platform connecting leading AI researchers with trusted industry experts to collaborate on AI alignment, safety, and training projects. We are partnering with a frontier AI lab to evaluate an AI model's ability to replicate empirical economics research findings.
\n
What You'll Do
\n
Identify suitable causal economics papers with publicly available replication data
\n
Write prompts asking the AI model to replicate findings given a research question, dataset, codebook, and context
\n
Write rubrics to evaluate the AI model's performance across each step of the empirical pipeline:
\n
Data cleaning
\n
Variable construction
\n
Specification choice
\n
Robustness judgment
\n
Who We're Looking For
\n
PhD in Economics
\n
(required)
\n
Hands-on experience with causal inference methods
\n
— DiD, IV, RDD, RCT, natural experiments
\n
Familiarity with replication-friendly microdata
\n
— NLSY, ACS, CPS, administrative data
\n
Proficient in
\n
STATA, R, or Python
\n
Strong understanding of empirical research workflow from raw data to published results
\n
Bonus: experience with
\n
AI/ML tools or interest in AI evaluation
\n
Ideal Background
\n
Active or former academic economist at a research university
\n
Published or working papers in applied microeconomics
\n
Fields: labor, health, development, public, environmental economics
\n
Why Join
\n
Contribute to cutting-edge
\n
AI safety and alignment research
\n
Adaptable part-time remote work
\n
— task-based engagement
\n
Collaborate with a global network of economists and AI researchers
\n
Competitive compensation per completed task
\n
Compensation - $100/hr USD
\n
Apply
\n
To apply, submit your CV. We review every application personally — no automated screening. xugodme If your background is a strong fit, you'll receive a direct link to join the project and complete your application.