news 
The Morning After: Google accused of using novices to fact-check Gemini’s AI answers
Google has been accused of using unqualified individuals to verify the accuracy of its Gemini AI-powered answer generation platform, according to a recent report. The alleged instruction suggests that contractors will no longer be allowed to skip prompts that require specialized domain knowledge, instead being forced to rate the parts they understand.
TechCrunch reports that internal guidance viewed by the publication reveals Google instructing contract workers not to skip any prompts, regardless of their expertise. However, this instruction has been updated, allowing testers to bypass sections where a significant amount of information is missing or if the content is harmful in nature.
The only exceptions to this rule seem to be instances where large portions of the information required are absent or the content poses a risk to users. In all other cases, contractors will need to rate the parts they comprehend while acknowledging any areas outside their knowledge.
Google’s response to the allegations emphasizes that its raters provide valuable feedback not just on the content itself but also on style, format, and other factors.
Source: www.engadget.com
