Design hard domain-specific tasks that expose AI reasoning flaws and check model outputs against primary sources, across history, law, finance, journalism and tech. About 15 hours a week at $30–60/hour, paid per task.
$30 – $60 / HourWorldwide
Remote AI training and data labeling jobs
Every role here has been checked against the platform that posted it. Pay is shown as reported, and marked when it is an estimate rather than a firm rate.
Filter jobs
Location
- Worldwide1 job
- United Statesno roles alongside your other filters
- Indiano roles alongside your other filters
- United Kingdomno roles alongside your other filters
- Mexicono roles alongside your other filters
- Canadano roles alongside your other filters
Language
- English1 job
- Spanishno roles alongside your other filters
- Japaneseno roles alongside your other filters
- Germanno roles alongside your other filters
- Frenchno roles alongside your other filters
- Portugueseno roles alongside your other filters
Field: AI Safety & Evaluation
- AI Safety & Evaluation1 job, applied. Activate to remove
- Business & Finance13 jobs
- Languages & Linguistics2 jobs
- Audio & Voice9 jobs
- Software & IT13 jobs
- Health & Medicine8 jobs
- Engineering15 jobs
- Science & Math2 jobs
- Law, Policy & Security13 jobs
- General & Data Collection2 jobs
- Data, AI & ML6 jobs
- Video, Image & Design3 jobs
- Writing & Education2 jobs
- Other fields4 jobs
Level: Medium
- Entryno roles alongside your other filters
- Juniorno roles alongside your other filters
- Medium1 job, applied. Activate to remove
- Senior3 jobs
Company: micro1
- micro11 job, applied. Activate to remove
- Mercor23 jobs
- AfterQuery28 jobs
- Turing20 jobs
- Invisible Technologiesno roles alongside your other filters
- Terac7 jobs
- Vetto1 job
- Alignerrno roles alongside your other filters
Newest
1 open role matching these filters
Nothing that fits today?
New roles land most days. Get them on Telegram or Discord as they are added, or read how the platforms pay before you apply.