Research Experts for Frontier AI/ML Open Questions at Mercor

Mercor is working with a leading intelligence AI lab to identify the most important open questions in core AI/ML fields and to build structured knowledge bases that could meaningfully accelerate progress over the next decade. We’re looking for exceptional PhD students and PostDocs with a clear point of view on what problems truly matter in their field and the depth to define how those problems could be tackled.

Eligibility Requirement

You will need to fill a short form in order to be eligible for this role: You will see this in addition to the AI interview in your application process. Below is guidance for what you will need to have in order to fill the form:

Consider the biggest open questions in your field, for example, the 10–15 questions where a breakthrough would make headline news. From this set, select those closest to your area of expertise: questions within or adjacent to your specialty, or those where you could mentor an expert toward meaningful progress.

Specifically, we are looking for questions where:

• A major breakthrough would be widely recognized as transformative (e.g., headline news in Nature, Science, or top field-specific venues)

• Meaningful progress is plausible within the next decade (not purely speculative or dependent on unknown technology)

• The question is concrete enough that progress can be evaluated (avoid umbrella questions like “How does the ML work?”)

• You have the relevant expertise to assemble a comprehensive knowledge base directly relevant to the question

What You’ll Do

1. Identify high-impact open questions

• Propose major open questions where a breakthrough would be transformative

• Focus on problems that are concrete, tractable, and close to your expertise

2. Build a knowledge base for selected questions

• Seminal papers, key datasets, methods, recent advances, and “hidden gems”

• Assume an extremely strong expert all knowledge up until 6 months ago (1st October, 2025)

Time commitment: ~8–16 hours per selected question

Who We’re Looking For

• PhD candidates or PostDocs from top-tier institutions

• Deep expertise in AI/ML/Engineering

• Strong judgment about significance, tractability, and research quality

• Ability to synthesize large bodies of literature into clear learning paths

• Openings: ~50 experts per domain

Expected Outputs

• Clearly articulated, high-impact open research questions

• Structured reading lists (~30–100 sources per question)

• Brief expert commentary on why each source and approach matters