Physics AI Evaluation Expert

You will evaluate physics problems and AI-generated scientific outputs used to improve AI models. Your work will test whether answers use sound mathematical and physical reasoning and will help build dependable evaluation methods for scientific performance.

• Review complex or unclear physics problems and model responses.
• Check derivations, quantitative solutions, units, dimensional consistency, boundary conditions, approximations, and limiting cases.
• Assess interpretations of experiments, simulations, theoretical models, and scientific literature.
• Find incorrect assumptions, unsupported claims, mathematical errors, and gaps in reasoning.
• Create scoring rubrics, reference answers, supported commentary, and repeatable evaluation records.
• Help resolve scientific disagreements and explain evaluation guidance clearly.

What It Pays and Takes

The role is a remote, asynchronous independent contractor engagement with occasional synchronous calls. The initial engagement lasts four weeks, and the work is listed as part time with a 20-plus-hour weekly requirement.

• Pay: The listing does not provide a rate.
• Location: Open to candidates in the United States; the work is fully remote.
• Language: Fluent in English.
• Education: A PhD in Physics, Applied Physics, or a closely related discipline is required.
• Experience: At least three years of post-PhD research, teaching, or relevant industry experience, plus current or recent work in academic research, industrial research and development, or a recognized research institution.
• Physics expertise: Specialization in areas such as condensed matter, particle, nuclear, astrophysics, plasma, optical, atomic and molecular, quantum information, or computational physics.
• You must be able to assess scientific literature, experimental design, uncertainty analysis, data interpretation, and mathematical derivations.
• You must document assumptions, evidence, uncertainty, limitations, and scientific consensus clearly.
• Helpful experience includes AI evaluation, machine learning for physics, scientific computing, dataset development, journal or technical report review, reproducibility, numerical validation, or programming in Python, MATLAB, Mathematica, or C++.
• Faculty appointments, substantial research contributions, peer-reviewed publications, patents, and graduate mentoring experience are also relevant.
• An NDA and compliance with project protocols, confidentiality requirements, and quality standards are required.

How It Works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI Training Work

AI training is the human work behind better AI systems, including reviewing model answers, checking reasoning, and preparing reliable examples and evaluations. Physics specialists are needed because scientific evaluation depends on subject knowledge, careful evidence review, and accurate mathematical judgment.