80,000 Hours Podcast

#18 - Ofir Reich on using data science to end poverty & the spurious action-inaction distinction

Jan 31, 2018
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ANECDOTE

Machine Learning to Detect Tax Fraud

  • In India, Ofir Reich uses machine learning to detect fraudulent companies evading VAT taxes.
  • This helps increase government revenues that can be spent on aiding the poor.
INSIGHT

False Results Plague Social Science

  • False or non-reproducible research results plague social sciences due to practices like selective hypothesis testing.
  • Open data, preregistration, and results-blind review can improve reliability and trustworthiness.
INSIGHT

RCTs Provide Valuable Evidence

  • Randomized controlled trials (RCTs) generally replicate and provide valuable precise inferences.
  • Their results teach transferable principles despite contextual variation, making them worth the investment.
Get the Snipd Podcast app to discover more snips from this episode
Get the app