$50–63/hr · Mercor · Hourly
You write and refine advanced philosophy assessment content that benchmarks next-generation AI model reasoning.
What you would do
- Compose original philosophy questions requiring conceptual synthesis across multiple subfields and texts
- Develop comprehensive step-by-step solutions using formal logical reasoning and markdown notation
- Select nine competitive wrong answers that target plausible errors or partial understandings
- Evaluate peer-written materials for accuracy, internal consistency, and solvability within time constraints
Who they want
- PhD or active doctoral candidacy in philosophy with demonstrated research output
- Master's degree acceptable only with exceptional specialization in a single subdomain
- Publications or university-level teaching experience in philosophy strongly preferred
- Native or near-native English proficiency with ability to express nuance without ambiguity
- Commitment of 10+ hours weekly for sustained, asynchronous contribution
Main skills
What the interview asks about
1.Conceptual precision and argumentation
Preventing ambiguous or trivial questions from entering benchmarks ensures AI evaluation materials remain challenging and academically defensible.
For example: “You've written: 'Artificial systems can never possess genuine intentionality because...' A reviewer flags this as containing a controversial premise disguised as foundational claim. How would you revise it?”
2.Distracter construction for expertise
Weak alternatives undermine assessment value; expert-level distractors reveal if test-takers actually understand the material versus lucky guessing.
For example: “For a question about Kripke semantics, provide the correct answer and four alternatives. Explain why each wrong option represents a meaningful misconception rather than obvious nonsense.”
3.Cross-domain integration
AI ethics and philosophy of technology draw on classical traditions; integrating multiple subfields signals you can handle interdisciplinary synthesis expected at this level.
For example: “Construct a question combining modal logic with responsibility attribution in autonomous systems. Why did you choose that particular integration point?”
4.Peer review judgment
Identifying flawed peer submissions without unnecessarily harsh critique maintains academic standards while fostering collaborative improvement.
For example: “A colleague's question about consciousness asks: 'Which theory best explains qualia in AI systems?' The phrasing conflates phenomenal consciousness with information integration. Draft feedback that acknowledges their intent while suggesting revision paths.”
A task you may get
Author a philosophy question with complete solution chain and nine alternatives. Explain why each wrong answer targets a specific misconception.
How to prepare
- Assemble five example questions from published philosophy assessments and analyze their structure, distractors, and why each answer tier works
- Review recent publications in AI ethics or philosophy of technology to refresh domain knowledge and identify emerging conceptual debates
- Practice writing formal definitions and necessary conditions that would withstand scrutiny from philosophers with different school-of-thought affiliations
The facts
- Pay
- $50–63/hr
- Hours
- Hourly
- Where
- Remote
- Field
- Humanities
- Posted
- 8/13/2026
- Places left
- 2
We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.