AI Trainer (Physical Sciences)

About the Role

We are looking for expert computational physical scientists to support a key initiative focused on improving modern artificial intelligence. This position is tailored for experts who combine extensive scientific knowledge with a track record in AI model training or assessment. You will contribute to expanding the limits of machine intelligence by ensuring these systems can manage intricate scientific challenges with precision and reliability. Your contribution will focus on using your specialized expertise to design rigorous tests for AI systems, while also evaluating generated simulations, code, and technical data to maintain high quality standards. We seek individuals with a demonstrated history of technical excellence and previous engagement in physical science AI-training initiatives.

Requirements

• Documented professional history in the training, evaluation, benchmarking, or advancement of artificial intelligence models within technical or scientific fields.
• Advanced practical ability in Python programming.
• Extensive specialization in a physical science field, such as physics, astronomy, chemistry, materials science, or computational simulation.
• Demonstrable expertise in using computational tools for data processing, numerical analysis, or research reproducibility.
• Evidence of technical competence through resources like open-source scientific software, peer-reviewed journals, research portfolios, or reproducible notebooks.
• Strong familiarity with scientific computing libraries including PyTorch, SciPy, or NumPy.

Responsibilities

• Design challenging and authentic computational physical science prompts to evaluate sophisticated AI models.
• Develop detailed task guidelines, create reproducible testing frameworks, and provide reference solutions with explicit success criteria.
• Conduct thorough reviews of AI-produced code, simulations, and numerical calculations.
• Detect and report any inaccuracies, including flawed assumptions, numerical mistakes, impossible physical outcomes, or incorrect methodologies.
• Verify that AI outputs align with established experimental findings, analytical data, or scientific observations to confirm technical accuracy.

Benefits & Perks

• Flexible work arrangement requiring a minimum commitment of fifteen hours per week.
• Competitive pay rates ranging from $60 to $120 per hour, calculated based on individual task completion.
• Fully remote position available to candidates worldwide, allowing for work from any location.