Side by Side Factuality Evaluator

Side by Side Factuality Evaluator - English

We’re looking for English-fluent experts to engage on a project about the factual accuracy of AI-generated content.

As an independent contractor, you’ll review AI-generated responses, split them into individual claims, and check each claim against reliable sources. Your judgments will help build AI systems that give more accurate, trustworthy, and well-supported information.

Experts may be asked to

• Review AI-generated responses on a wide range of topics

• Break responses into individual, verifiable claims

• Research each claim online and find credible, authoritative sources

• Decide whether each claim is accurate, inaccurate, or unverifiable, paying close attention to small differences in wording

• Make independent, evidence-based calls, including in ambiguous cases

• Follow detailed evaluation guidelines and record clear, consistent decisions with supporting evidence

This project may be a strong match for experts who

• Are fluent in English, with strong reading comprehension and writing skills

• Have strong critical-thinking and analytical skills

• Have strong web research skills and can tell reliable sources from unreliable ones

• Pay close attention to detail, especially when spotting individual claims and subtle wording differences

• Are comfortable researching many kinds of topics, including unfamiliar ones

• Can understand detailed instructions and apply them consistently