- Contractors training Google’s AI chatbot, Bard, express dissatisfaction with convoluted instructions and problematic working conditions.
- Trainers often have to audit responses within three minutes but are concerned about Bard’s inaccuracies despite not having to perform rigorous fact checks.
- Google insists on its extensive work to ensure factuality and reduce biases while trainers allege that rushing may lead to inaccurate responses.
Menu
