Code and language assessment
I worked in a platform with multiple projects that involved training AI. The projects involved supervising the responses of an AI chatbot to prompts requiring both coding and Spanish language skills. The focus was on identifying issues related to truthfulness, instruction-following, and localization. Additionally, the role included reviewing assessments made by other workers during their interactions with the AI model. The project followed a multi-turn conversational structure. Each turn presented a prompt for which the model generated two responses. My task was to review both responses, choose the better one, and edit it if any issues were present.