Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

Google accused of using novices to fact-check Gemini's AI answers

Dec 19, 2024 - engadget.com
Google has reportedly instructed contract workers evaluating its AI model, Gemini, not to skip prompts outside their expertise, according to internal guidance viewed by TechCrunch. Previously, contractors could skip prompts that required specialized knowledge they did not possess, such as legal questions for non-lawyers. The new guidelines now require evaluators to rate the parts they understand and note their lack of expertise, only allowing skips if information is missing or if the content is harmful and requires specific consent for evaluation.

This change has raised concerns among contractors, who believe that skipping prompts was intended to improve accuracy by assigning tasks to more knowledgeable evaluators. Google has not commented on these new instructions, which were reportedly communicated to GlobalLogic, the outsourcing firm responsible for evaluating AI-generated output. The move comes after Google shared a preview of Gemini 2.0 earlier this month.

Key takeaways:

  • Google allegedly instructed contract workers evaluating Gemini not to skip prompts, regardless of their expertise, according to internal guidance viewed by TechCrunch.
  • Previously, contractors could skip prompts outside their expertise, but now they are instructed to rate what they understand and note areas outside their knowledge.
  • Contractors can only skip prompts if a significant portion of information is missing or if the content is harmful and requires specific consent forms.
  • Google has not responded to requests for comment on these changes.
View Full Article

Comments (0)

Be the first to comment!