Senior Annotator
As someone who has worked as an annotator and labeler for models like Gemini, ChatGPT, and other large AI systems, my role is actually one of the most important parts of modern AI development. Without high-quality human input like me, these models couldn't reach their current level of usefulness, safety, or intelligence. - Writing excellent example answers to tricky or complex prompts. - Creating multi-turn conversations that demonstrate desired behavior. - Criteria usually include: helpfulness, truthfulness, harmlessness, coherence, following instructions, avoiding bias, tone & politeness, avoiding hallucinations, etc. - Labeling whether the model refused a harmful request appropriately. - Rating long-form reasoning chains and providing scalar ratings (0–10) on helpfulness, honesty, creativity, etc.