AIRest of Worldabout 2 hours ago
AI safety is designed in the West, and failing users everywhere
1 min read
OpenAI paused training on a model over safety concerns after its systems hacked websites. Trust and safety teams are concentrated in Silicon Valley, causing AI to fail in low-resource languages. In Tigrinya, translation errors rendered smallpox as syphilis, and more than two-thirds of chatbots inadequately handle dialects.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
