Outsmart Artificial Intelligence with Your Human Brilliance
Join the top 1% of scholars in shaping the future of AI
Earn up to $118/hr, paid out weekly
for qualified problem and solution pairs
Remote work, flexible hours
Work from anywhere, anytime
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.
Read more about Scale AI and Humanity's Last Exam
040725 - Landing Page Video_16-9
Click for sound
"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
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 a PhD or Master’s Degree in your field of expertise or a related field.
- Attendance or a degree from a Top 200 accredited University or Institution such as Harvard, MIT, Stanford, Oxford.
- 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 $118/hr, paid out weekly
Rates vary based on quality, accuracy and time spent
Co-author a research paper on how data enhances AI reasoning
Analyze how effectively this data impacts industry-leading models
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
Join an exclusive network of top 1% folks in their domains
Collaborate with leading experts at an all-expenses-covered event
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".
Being able to discuss the nuances of intricate genetic systems with people from around the world that I've never met before is super cool… and of course the money is a good bonus!
Josh Cole
Biologist, United States
As an engineer, I often use Al models to get the right answers, but this experience is the opposite: I need to make the models generate wrong answers. It is strange, exciting, and unforgettable.
Cuong N.
Computer Scientist, Vietnam
Application and onboarding process
STEP 1
Create an account
on Outlier
STEP 2
Verify your identity and conduct a video interview
STEP 3
Contribute your expertise to outsmart models and earn
Frequently Asked Questions (FAQs)
What is Humanity's Last Exam (HLE) in simple terms?
What is the relationship between Outlier and Scale AI?
How will contributors be paid for their HLE submissions?
What are the main benefits of participating in HLE?
What's the expected time commitment for participants?
In the News
Outlier Updates to Empower Contributors
April 3, 2025
Advancing Frontier Model Evaluation
April 2, 2025
Scale’s Series F: Expanding the Data Foundry for AI
May 21, 2024