4/5 of 1,945 reviews

Challenge AI with Domain-Specific Visual Reasoning

Join a global community of Math experts to shape the future of AI. Earn up to $50/hr.

CHOOSE YOUR EXPERTISE

Arabic
Earn Upto $50/hr
Shape the future of AI
Work Remotely

LANGUAGE

Arabic

LANGUAGE

Arabic

PAY

Earn Upto $50/hr

PAY

Earn Upto $50/hr

LANGUAGE

Arabic

PAY

Earn Upto $50/hr

Rated 4/5 from 1,945 reviews on

Outlier is powered by Scale AI to accelerate AI development at leading enterprises

Earn an average $20 - $50/hr, paid out weekly

Earn an average $20 - $50/hr,
paid out weekly

Rates vary based on your
area of expertise

Remote work,
flexible hours

Work from
anywhere, anytime

Define
AI's future

Define AI's future

Define
AI's Future

Leverage your expertise
and train frontier AI models

About Outlier

Outlier is committed to improving the intelligence & safety of AI models. Owned and operated by Scale AI, we’ve recently been featured in Forbes for partnering experts with top AI labs to provide the high quality data for LLMs.

About the role
Our mission is to challenge AI with Humanity’s Last Exam, a rigorous new way to test AI’s intelligence and reasoning. To do so, we are recruiting a group of exceptional minds to craft PhD and Master's level problems that current AI models cannot solve correctly.

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

Mark Zuckerberg

Founder and CEO, Meta

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

Mark Zuckerberg

Founder and CEO, Meta

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

Mark Zuckerberg

Founder and CEO, Meta

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

Mark Zuckerberg, Founder and CEO, Meta

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

Mark Zuckerberg, Founder and CEO, Meta

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

Mark Zuckerberg, Founder and CEO, Meta

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta
"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."
-Forbes, A Growing Side Hustle For American College Grads: Fixing AI’s Wrong Answers
"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."
-Mark Zuckerberg, Founder & CEO Meta

What you will be doing

Create Challenging Prompts for AI

Write a valid question that AIs and average humans cannot answer.

Subject

Medicine

Question

Provide a challenging prompt that will stump the AI model or cause it to "break".

Provide Benchmark Solutions

Provide a definitive, high quality solution that serves as the benchmark for evaluating AI responses.

Question

Difficult

Answer/Solution

Provide a high-quality, thorough response that directly matches the intent of the original prompt.

Participate in Peer Review

Work closely with fellow experts to evaluate and refine each other's solutions. Provide feedback to ensure high accuracy.

Contributor A’s response

Edit

Approve

What we’re looking for

Pursuing or holding an Undergraduate, Master’s, or PhD degree in your field of expertise or a related field.
Attendance or a degree from a Top accredited University or institution such as Harvard, MIT, Stanford, Oxford.
2+ years of demonstrated expertise in your domain.
Deep subject matter expertise with the ability to create complex, graduate-level problems that challenge AI reasoning.
Strong analytical and problem-solving skills, with experience in crafting rigorous, high-quality questions and solutions.
Attention to detail to accurately assess AI capabilities and evaluate peer submissions.
Fluency/High proficiency in English.

What you will get from contributing

Earn up to $20-50 USD/hr, based on your area of expertise

Rates vary based on quality, accuracy and time spent. Paid out weekly via PayPal and AirTM

Earn up to $20-50 USD/hr, based on your area of expertise

Rates vary based on quality, accuracy and time spent. Paid out weekly via PayPal and AirTM

Earn up to $20-50 USD/hr, based on your area of expertise

Rates vary based on quality, accuracy and time spent. Paid out weekly via PayPal and AirTM

Join an exclusive network of top 1% folks in their domains

Collaborate with leading experts at an all-expenses-covered event

Join an exclusive network of top 1% folks in their domains

Collaborate with leading experts at an all-expenses-covered event

Join an exclusive network of top 1% folks in their domains

Collaborate with leading experts at an all-expenses-covered event

Flexible schedule and time commitment

No contracts, no 9-to-5. You control your schedule. (Most experts spend 5-10 hours/week, up to 40 hours working from home

Flexible schedule and time commitment

No contracts, no 9-to-5. You control your schedule. (Most experts spend 5-10 hours/week, up to 40 hours working from home

Flexible schedule and time commitment

No contracts, no 9-to-5. You control your schedule. (Most experts spend 5-10 hours/week, up to 40 hours working from home

Combine domain-specific images and prompts to challenge AI models

Combine meaningful domain-specific images with carefully crafted prompts to challenge state-of-the-art language models

Combine domain-specific images and prompts to challenge AI models

Combine meaningful domain-specific images with carefully crafted prompts to challenge state-of-the-art language models

Combine domain-specific images and prompts to challenge AI models

Combine meaningful domain-specific images with carefully crafted prompts to challenge state-of-the-art language models

Choose your expertise

Hear from other experts

Read about how MIT PhD Daniel Z earns extra income on Outlier

"Daniel has enjoyed the combination of earning extra cash while contributing to AI development and has made over $4,000 on Outlier".

Read Daniel's Story

Working on Outlier has helped expose me to new, different areas and hone my research skills. Communication is really good for the projects I’m in. The support, the community channels, QM support, weekly webinars, and office hours for questions help things run smoothly.

Robin F.

Math

Robin F.

Math

Robin F.

Math

Working at Outlier has been an enlightening experience. The focus on data quality and analytics has allowed me to hone my skills in a fast-paced, innovative environment.

Chelsea S.

Math

Working at Outlier has been an enlightening experience. The focus on data quality and analytics has allowed me to hone my skills in a fast-paced, innovative environment.

Chelsea S.

Math

Working at Outlier has been an enlightening experience. The focus on data quality and analytics has allowed me to hone my skills in a fast-paced, innovative environment.

Chelsea S.

Math

Application and onboarding process

STEP 1

Create an account
on Outlier

STEP 1

Create an account
on Outlier

STEP 1

Create an account
on Outlier

STEP 2

Verify your identity and conduct a video interview

STEP 2

Verify your identity and conduct a video interview

STEP 2

Verify your identity and conduct a video interview

STEP 3

Contribute your expertise to outsmart models and earn

STEP 3

Contribute your expertise to outsmart models and earn

STEP 3

Contribute your expertise to outsmart models and earn

In the News

Scale AI

Outlier Updates to Empower Contributors

April 3, 2025

Scale AI

Advancing Frontier Model Evaluation

April 2, 2025

Scale AI

Scale’s Series F: Expanding the Data Foundry for AI

May 21, 2024

In the News

Scale AI

Outlier Updates to Empower Contributors

April 3, 2025

Scale AI

Advancing Frontier Model Evaluation

April 2, 2025

Scale AI

Scale’s Series F: Expanding the Data Foundry for AI

May 21, 2024

In the News

Scale AI

Outlier Updates to Empower Contributors

April 3, 2025

Scale AI

Advancing Frontier Model Evaluation

April 2, 2025

Scale AI

Scale’s Series F: Expanding the Data Foundry for AI

May 21, 2024

Ready to use your brilliance to outsmart AI?

CHOOSE YOUR EXPERTISE

in collaboration with Scale AI

in collaboration with

Challenge AI with Domain-Specific Visual Reasoning

Challenge AI with Domain-Specific Visual Reasoning

Earn an average $20 - $50/hr, paid out weekly

Earn an average $20 - $50/hr,paid out weekly

Remote work,flexible hours

Remote work,flexible hours

DefineAI's future

Define AI's future

Define AI's Future

About Outlier

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone.

As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models.

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

"As artificial intelligence becomes ever more capable of mimicking human reasoning, it's increasingly requiring more highly skilled humans, often with specific areas of expertise, to hone its models."

"We partnered with Scale AI to work with Enterprises to adopt Llama and train custom models with their own data. We are excited to collectively make Llama the industry standard and bring the benefits of AI to everyone."

What you will be doing

What you will be doing

What you will be doing

Create Challenging Prompts for AI

Create Challenging Prompts for AI

Create Challenging Prompts for AI

Provide Benchmark Solutions

Provide Benchmark Solutions

Provide Benchmark Solutions

Participate in Peer Review

Participate in Peer Review

Participate in Peer Review

What we’re looking for

What you will get from contributing

Earn up to $20-50 USD/hr, based on your area of expertise

Earn up to $20-50 USD/hr, based on your area of expertise

Earn up to $20-50 USD/hr, based on your area of expertise

Join an exclusive network of top 1% folks in their domains

Join an exclusive network of top 1% folks in their domains

Join an exclusive network of top 1% folks in their domains

Flexible schedule and time commitment

Flexible schedule and time commitment

Flexible schedule and time commitment

Combine domain-specific images and prompts to challenge AI models

Combine domain-specific images and prompts to challenge AI models

Combine domain-specific images and prompts to challenge AI models

Choose your expertise

Choose your expertise

Choose your expertise

Accounting

Agriculture

Architecture

Art

Basic Medical Science

Biology

Biomedical Science

Chemistry & Chemical Engineering

Civil Engineering

Computer Science & Engineering

Design

Diagnostics & Laboratory Medicine

Economics

Electrical Engineering

Electronics Engineering

Finance

History

Literature

Managerial Science

Marketing & Sales

Mathematics

Mechanical Engineering

Music

Operations Management & Research

Others (Geography, Sociology etc)

Physics

Earn an average $20 - $50/hr,
paid out weekly

Remote work,
flexible hours

Remote work,
flexible hours

Define
AI's future

Define
AI's Future