New opportunity on Mercor!
If you previously worked as a reviewer on Project O, here's another opportunity worth checking out.
There are two exercises to complete:
1. Rubric Review Exercise
2. Task Review Exercise
What does the role involve?
You'll be reviewing AI-generated tasks, including prompts, grading rubrics, and model responses, to determine whether the model's failures are genuinely justified.
Your responsibilities will include:
- Verifying that the AI agent had access to all the information needed to complete the task.
- Checking whether grading rubrics accurately reflect the user's instructions.
- Identifying situations where a valid response could be unfairly marked as incorrect.
- Ensuring rubrics comply with the required evaluation guidelines.
- Rewriting unclear, unfair, or poorly structured criteria.
- Flagging tasks that cannot be fixed and briefly explaining your reasoning.
The goal: Ensure AI-generated tasks are evaluated fairly, consistently, and strictly according to the actual instructions.
If you have experience reviewing AI outputs, this is definitely worth exploring!
t.mercor.com/sOlQF
New opportunity on Mercor!
If you previously worked as a reviewer on Project O, here's another opportunity worth checking out.
There are two exercises to complete:
1. Rubric Review Exercise
2. Task Review Exercise
What does the role involve?
You'll be reviewing AI-generated tasks, including prompts, grading rubrics, and model responses, to determine whether the model's failures are genuinely justified.
Your responsibilities will include:
- Verifying that the AI agent had access to all the information needed to complete the task.
- Checking whether grading rubrics accurately reflect the user's instructions.
- Identifying situations where a valid response could be unfairly marked as incorrect.
- Ensuring rubrics comply with the required evaluation guidelines.
- Rewriting unclear, unfair, or poorly structured criteria.
- Flagging tasks that cannot be fixed and briefly explaining your reasoning.
The goal: Ensure AI-generated tasks are evaluated fairly, consistently, and strictly according to the actual instructions.
If you have experience reviewing AI outputs, this is definitely worth exploring!
t.mercor.com/sOlQF
Check your DM
you asked for a referral program, so we built one! π
earn $20 for every new Guild user who signs up through your link, qualifies, and completes their first approved, billable hour.
log in to Guild talents.getblend.com/ β Affiliate Program to get your unique referral link.
we're launching in beta, starting with AI Red-Team Specialists only.
share with friends, colleagues, or your community.
a few rules:
new Guild users only
registration must happen through your link
one referrer per person
no self-referrals or duplicate accounts
bonuses are credited once the qualifying work is approved
let's see what we can build together!
309











