Quick facts
- Pay
- $100/hr
- Type
- Contract
- Location
- Remote (US, Canada)
- Background
- PhD
- Platform
- Workada — 23 open jobs
- Typical pay at Workada
- $23–130/hour · median $100 (9 jobs on our site)
- Payout
- Stripe (3 times a week)
- Apply
- Free to apply — never pay to get AI training work
This role is also open in other regions
- Remote (US, Canada) — $100/hr
- Remote (US, Canada) — $100/hr
What You'll Do
As a contributor, you'll draw on your specialized research and published findings to stress-test advanced AI models. You'll design complex scenarios based on your own expertise to identify where models struggle with high-level scientific reasoning. By surfacing subtle reasoning errors and providing expert-validated corrections, you help set the standard that teaches AI to think more like a scientist. Your work ensures that the next generation of models is grounded in rigorous evidence and accurate technical logic.
Specialties We're Looking For
If your PhD is in Cancer Biology or Oncology, there's likely a place for you.
You're a Great Fit If
- You hold, or are completing, a PhD in Cancer Biology or Oncology. - You can draw on your own published papers (work you authored or co-authored and already have accessible) as the source material for your Tasks. - You read primary literature fluently and think critically about study design, statistics, and evidence. - You have a precise, adversarial eye; you enjoy finding the exact point where an argument breaks and explaining it cleanly. - You're honest about uncertainty, distinguishing what the evidence shows from what you'd like it to show. No AI or machine-learning background needed. Your scientific judgment is the expertise we're after.
Materials You Can Use
Almost any piece of your work can seed a Task, for example: - Published papers - Figures and the data behind them - Result tables and summary statistics - Datasets and supplementary files - Experimental images and readouts - Sequences, structures, and other domain-specific data - Methods, protocols, and the analysis code behind them