About this role
Develop original, executable scientific computing challenges that help evaluate advanced AI systems. This remote role focuses on creating rigorous physics problems that strong frontier models struggle to solve.
Key Responsibilities- Select or create source material, including published papers, Kaggle datasets, open-source repositories, or original scenarios.
- Write scientific coding prompts based on the selected material.
- Define grading criteria that establish what a correct answer requires.
- Test and calibrate tasks against frontier AI models, with tasks released only when strong models fail more often than they succeed.
- PhD in physics, applied physics, or a closely related field.
- Demonstrated depth in at least two of these physics subdomains: condensed matter, optics, quantum information or computing, computational physics, astrophysics, and particle physics.
- Scientific computing coding experience, with working proficiency in Python, R, or another relevant programming language.
- Comfort using Git or GitHub and running code in Docker. Work is completed through a pull-request workflow with automated quality checks.
- Publications in peer-reviewed journals.
- Previous scientific software development or research engineering experience.
- Remote, hourly engagement.
- Six-week project duration.
- Part-time commitment of at least 20 hours per week.
- Immediate start.
- $70 per hour.
- Submit a resume and completed application form.
- Complete a 25-minute conversational interview covering your background, experience, and motivations.
- Receive follow-up within a few days regarding next steps and onboarding.