What is AI training work, and who gets paid for it?
A plain-English guide to AI training work: what the tasks are, who the labs hire, what it pays and how to get started.
Every AI model you use was shaped by people. Before a model can answer a tax question, review a contract or fix a bug, it needs thousands of examples of what a good answer looks like, and people who can tell a good answer from a plausible-sounding wrong one. That's AI training work.
What the tasks look like
Most tasks fall into a few types:
- Rating and ranking. You compare two or more answers from a model and explain which is better. This feedback, often called RLHF, is one of the main ways models improve.
- Writing. You write an ideal answer to a hard question in your field, so the model has a gold standard to learn from.
- Testing. You try to make the model fail: find the edge cases, the outdated facts or the unsafe advice, and document them.
- Building. For technical fields, you create problems, test suites or small environments that measure how well an AI agent does real work.
Who gets hired
Early AI training work was mostly general labelling. Today, labs need experts. The most in-demand profiles are software engineers, doctors and nurses, lawyers, finance professionals, scientists and mathematicians, and native speakers of many languages. A degree or a few years of professional experience in your field is often the main requirement. You don't need to know anything about machine learning.
What it pays
Pay is usually hourly and depends on how specialised the work is. General tasks often pay $15 to $30 an hour. Expert tasks commonly pay $50 to $150 an hour, and some specialist roles pay more. Some platforms also run one-off paid studies or interviews for a fixed amount.
Where the work comes from
Labs rarely hire experts directly. They work through human-data platforms, companies that recruit experts, check their skills and manage the work. Each platform posts its own openings, runs its own assessment and pays you directly. That's why it can take a while to find the right task: the openings are spread across many sites.
How to get started
- Find tasks that fit your background. Look for openings that ask for your field and level of experience. humaven matches you with the ones you're most likely to get.
- Apply on the platform. Expect a short skills test or interview, and an ID check.
- Start small. Quality matters more than speed. Careful, well-explained work leads to more tasks and better-paid projects.
A few things to watch
Work is flexible but not always steady: projects start and end, and hours can vary week to week. Apply only through the platform's own site, and never pay to get work. If you have a full-time job, check your contract for rules about outside work.