What the work involves

Projects vary, but the core tasks repeat. You write a question rooted in your own field — supplying the primary sources, the context a reader needs, and the interpretive problem itself — then draft the reference answer a model's attempt will be graded against. You grade model-produced interpretation and argument: does the source actually say what the model claims, does the inference hold, is the historical or cultural framing right. And you flag the characteristic failure mode of frontier models in the humanities — prose that reads well and cites confidently while misreading the passage, collapsing a live scholarly debate into a settled consensus, or inventing a reference outright.

Most of the deliverable is written justification. A rating without a defensible reason attached is close to useless to the research team, so the bar is a short, precise account of where the reading fails and what a stronger one would have done.

What the platform screens for

  • Verifiable scholarly background. A graduate degree in a humanities field, or equivalent scholarly or professional experience, plus a resume that lets Mercor confirm it.
  • Discrimination between readings. Not whether an answer is wrong, but whether you can distinguish a strong reading from a merely plausible one and name the precise point of failure.
  • Written clarity under time pressure. Explaining reasoning is the job, not an adjunct to it.
  • Tolerance for ambiguity. Task instructions are sometimes underspecified; the platform wants people who flag that rather than guess silently.

Logistics

Fully remote and asynchronous, with hours set per project. Rates on listings in this field have posted at $45–60/hr, set per project by scope and depth — that is an observed range, not a guarantee, and the binding number appears on the specific project listing.

This is a standing pool, not an opening. Applying means a resume, a work-location confirmation, and roughly a 20-minute AI interview. You will not receive an accept or reject decision on this listing; you are contacted when a project matches your field, which may be within a week or several months. Completing the domain expert interview and joining every network you qualify for improves match odds.