Job opportunity
Generative AI Voice Image Evaluator
Scale AI
Evaluate generative AI voice, image, and video outputs with sharp attention to detail. Outlier offers flexible freelance work from a remote location.
As an AI Evaluator, you will help train models through structured human feedback. You will judge whether outputs follow prompts, inspect realism and quality, and document visible or audible issues. Clear English communication is central to every evaluation task.
Key Responsibilities:
Conduct live, scenario-based English conversations with voice models
Judge paired images and videos for prompt adherence
Zoom and replay outputs to identify subtle defects
Assess anatomy, motion, text, and audio-video alignment
Rate outputs and explain decisions in English
Requirements:
Fluent English and everyday conversational comfort
Strong, clear written communication skills
Ability to notice problems through repeated review
Comfort making decisions when options have flaws
Short screening and onboarding, possibly including spoken English
Bring careful observation, conversational confidence, and precise written reasoning to Outlier’s generative AI evaluation projects.
#J-18808-Ljbffr
Apply on source site As an AI Evaluator, you will help train models through structured human feedback. You will judge whether outputs follow prompts, inspect realism and quality, and document visible or audible issues. Clear English communication is central to every evaluation task.
Key Responsibilities:
Conduct live, scenario-based English conversations with voice models
Judge paired images and videos for prompt adherence
Zoom and replay outputs to identify subtle defects
Assess anatomy, motion, text, and audio-video alignment
Rate outputs and explain decisions in English
Requirements:
Fluent English and everyday conversational comfort
Strong, clear written communication skills
Ability to notice problems through repeated review
Comfort making decisions when options have flaws
Short screening and onboarding, possibly including spoken English
Bring careful observation, conversational confidence, and precise written reasoning to Outlier’s generative AI evaluation projects.
#J-18808-Ljbffr