Outlier Virtual — AI Evaluation Specialist (Aether Project)
Evaluated AI-generated text-to-image and text-to-video outputs for quality, accuracy, and prompt compliance. Identified visual errors, inconsistencies, and content defects against established quality standards. Verified that generated content aligned with user prompts through structured evaluation and annotation. • Assessed high volumes of AI-generated multimedia content while maintaining accuracy and consistency. • Compared multiple AI outputs to select the highest-quality results. • Logged findings and applied human feedback to support improvements to generative AI models. • Ensured prompt compliance and rating consistency across evaluations.