Image Reasoning Specialist - AI Lab - AI Trainer

Mercor; Washington (District of Columbia); Remote; USD 41,000 - 55,000; Part time

Job summary

Mercor is seeking an Image Reasoning Generalist to join a remote-first team delivering high-quality prompts and image tests for AI models. You will work independently to find challenging images, craft clear instructions, and ensure labeling accuracy across multiple tasks.

Ideal candidates have strong written English, excellent attention to detail, and experience with photography, image editing, or data labeling in detail-focused roles.

Qualifications

• Strong written English skills for clear and understandable prompts.
• Good attention to detail to identify potential model confusion in images.
• Consistency and reliability in handling multiple tasks.
• Preferred experience with photography, image editing, data labeling, or other detail-focused work.

Responsibilities

• Find images that may cause AI models to count incorrectly.
• Use original, public domain, or AI-generated images.
• Identify elements that might confuse models, such as misleading text or similar-looking objects.
• Test images with AI models using the prompt: "How many X are there? Return an integer."
• Write additional prompts to provide context and label tasks effectively.
• Work independently and asynchronously to ensure accuracy and consistency in task completion.

Skills

Strong written English

Attention to detail

Reliability

Photography

Data labeling

Tools

Image editing software

Job description

About the job

Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark, General Catalyst, Peter Thiel, Adam D'Angelo, Larry Summers, and Jack Dorsey.

Position

Image Reasoning Generalist

Type

Contract

Compensation

$35/hour

Location

Remote

Role Responsibilities

• Find images that may cause AI models to count incorrectly.
• Use original, public domain, or AI-generated images.
• Identify elements that might confuse models, such as misleading text or similar-looking objects.
• Test images with AI models using the prompt: "How many X are there? Return an integer."
• Write additional prompts to provide context and label tasks effectively.
• Work independently and asynchronously to ensure accuracy and consistency in task completion.

Qualifications

• Strong written English skills for clear and understandable prompts.
• Good attention to detail to identify potential model confusion in images.
• Consistency and reliability in handling multiple tasks.
• PreferredExperience with photography, image editing, data labeling, or other detail-focused work.