The work
You will review original images alongside AI-generated versions and assess how accurately the generated image reflects the original. Your notes will help improve the quality of AI image systems.
The work requires careful visual comparison, consistent use of review guidelines, and clear descriptions of each issue and its exact location.
- Upload original images that meet project specifications.
- Compare original and AI-generated images side by side.
- Review lighting, perspective, texture, anatomy, color, shadows, patterns, surfaces, and text.
- Identify defects such as warped hands, smudged surfaces, inconsistent shadows, broken patterns, and misshapen text.
- Write concise, error-free notes that make each issue clear and actionable.
- Apply detailed quality standards consistently and prioritize accurate evaluations.
What it pays and takes
Experience in visual work can help, but prior AI experience is not required when you bring relevant expertise.
- Pay: $25 to $35 per hour.
- Work type: Part-time contract work.
- Location: Open to candidates in the United States.
- Language: Excellent written and verbal English.
- Required: Strong visual acuity and the ability to detect subtle differences between similar images.
- Required: Careful, reliable judgment when applying detailed review standards.
- Helpful experience: Photography, digital imaging, graphic design, photo editing, image evaluation, data annotation, image comparison, or visual quality assurance.
- Helpful knowledge: Technical vocabulary for describing visual defects and common AI-generated image failure modes.
How it works
You apply here in about two minutes and finish on the hiring platform. The next step is a short AI interview (technical roles include a coding challenge). Passing it is how you qualify for paid tasks, so take it soon after you are invited.
The hiring platform's application for this job asks for your resume, your LinkedIn profile, your phone number, your location and answers to 3 screening questions.
About AI training work
AI training is the human work behind systems that learn from examples, including reviewing images and rating model outputs. Careful evaluators help identify errors and improve how these systems perform.