The work
You will review prompts and responses produced by an assigned AI assistant. Your task is to judge how useful, accurate, and high-quality each response is from the perspective of an everyday user.
Evaluations are completed in English. You will use your own personal account on an assigned assistant and draw on conversation history from an account that has been used for more than two weeks.
- Evaluate the quality and usefulness of AI assistant responses.
- Review English prompts and responses from an everyday-user perspective.
- Apply consistent judgment across response-evaluation tasks.
- Use a personal account on Gemini, Instinct, or Muse with sufficient conversation history.
What it pays and takes
This is a 16-week remote project. The project is paid in USD, but the payment amount and payment unit are not provided in the role details.
- Pay: Paid in USD; the amount and payment unit are not specified.
- Time: 20 or more hours per week.
- Location: You must currently be based in the United States and able to work remotely.
- Language: Fluent English and comfort evaluating English prompts and responses.
- Education: Your highest completed degree must be a high school diploma, some college, an associate degree, or a bachelor's degree.
- Eligibility: Applicants with master's, professional, or doctoral degrees do not qualify.
How it works
Apply on OpenTrain with your resume and then complete the application on the hiring site. OpenTrain is where people find AI training work and begin their application.