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

There's no arguing that AI still has quite a few unreliable moments, but one would hope that at least its evaluations would be accurate. However, last week Google allegedly instructed contract workers evaluating Gemini not to skip any prompts, regardless of their expertise, TechCrunch reports based on internal guidance it viewed. Google shared a preview of Gemini 2.0 earlier this month.  

Google reportedly instructed GlobalLogic, an outsourcing firm whose contractors evaluate AI-generated output, not to have reviewers skip prompts outside of their expertise. Previously, contractors could choose to skip any prompt that fell far out of their expertise — such as asking a doctor about laws. The guidelines had stated, "If you do not have critical expertise (e.g. coding, math) to rate this prompt, please skip this task."

Now, contractors have allegedly been instructed, "You should not skip prompts that require specialized domain knowledge" and that they should "rate the parts of the prompt you understand" while adding a note that it's not an area they have knowledge in. Apparently, the only times contracts can skip now are if a big chunk of the information is missing or if it has harmful content which requires specific consent forms for evaluation. 

One contractor aptly responded to the changes stating, "I thought the point of skipping was to increase accuracy by giving it to someone better?" 

Google has not responded to a request for comment. 

This article originally appeared on Engadget at https://www.engadget.com/ai/google-accused-of-using-novices-to-fact-check-geminis-ai-answers-143044552.html?src=rss https://www.engadget.com/ai/google-accused-of-using-novices-to-fact-check-geminis-ai-answers-143044552.html?src=rss
Creato 28d | 19 dic 2024, 15:10:13


Accedi per aggiungere un commento

Altri post in questo gruppo

Google decides it won't comply with EU fact-checking law

Google has told the EU it will not comply with a forthcoming fact-checking law, according to a copy of a let

16 gen 2025, 22:30:05 | Engadget
CFPB fines Block $175m over Cash App's lax fraud controls

The Consumer Financial Protection Bureau (CFPB) announced today that's

16 gen 2025, 22:30:04 | Engadget
AGDQ just ended, but there's already a schedule for Frost Fatales and it owns

Awesome Games Done Quick has already wrapped up for 2025 (with a cool

16 gen 2025, 22:30:03 | Engadget
China-linked hackers accessed over 400 US Treasury computers

The US Treasury Department announced in a letter back in December that it had been the

16 gen 2025, 20:10:15 | Engadget
MoviePass made a film trailer app for the Oculus Quest and Apple Vision Pro

If you're a cinephile who misses the old Apple TV app for movie trailers, MoviePass CEO Stacy Spikes knows your pain. So he decided to build a trailer app of his own, one that could easily

16 gen 2025, 20:10:14 | Engadget
TikTok, Temu and more face complaints alleging GDPR violations in EU

Austrian privacy advocate NOYB has launched its first GDPR complaints against Chinese businesses. The organization has filed complaints against TikTok, Xiaomi, Shein, AliExpress, Temu and WeChat, a

16 gen 2025, 20:10:13 | Engadget
Apple pauses AI notification summaries of news alerts in latest iOS beta

Some significant changes are coming to Apple Intelligence notification summaries. With the latest slate of developer previews for iOS 18.3, iPadOS 18.3 and macOS Sequoia 15.3, Apple has suspended t

16 gen 2025, 20:10:12 | Engadget