PhD in Physics or related field with postdoctoral research experience.
Proven research-level reasoning and critiquing skills in physics.
Strong scientific programming expertise in Python and familiarity with research tools like Git and LaTeX.
Ability to communicate complex ideas clearly and effectively.
Experience collaborating in fast-paced, cross-disciplinary environments.
Entrepreneurial mindset with a passion for solving fundamental scientific challenges.
Responsibilities
Review and critique the model's reasoning in physics for errors and assumptions.
Provide corrections and alternative derivations to define quality scientific reasoning.
Translate domain knowledge into actionable model behavior and improvements.
Create reference outputs for model training and fine-tuning processes.
Rank model outputs to enhance scientific reasoning through feedback loops.
Cooperate in building an external Collaborators program for expert review.
Contribute to high-quality research outputs suitable for publication.
Benefits
Opportunity to solve foundational issues at the intersection of AI and physics.
High-trust, low-bureaucracy work environment promoting real ownership.
Remote-first workplace with flexible day structuring.
Exposure to advanced ideas in AI, scientific discovery, and technology.
Culture emphasizing curiosity, depth of thinking, and first-principles reasoning.
Opportunity to contribute to advancements in fundamental science.
Full Job Description
Why This Role Exists:
We are building an AI Physicist, but there is no established way to define, evaluate, or measure high-quality research when AI is part of the system. This role exists to bring deep domain expertise to shape how the system learns, what "good" looks like, and how we benchmark progress, ensuring the outputs are scientifically rigorous, meaningful, and advance the field.
What You'll Do:
Review and critique the model's reasoning in physics, identifying subtle conceptual errors and invalid assumptions.
Provide clear corrections, alternative derivations, and minimal counterexamples to shape what "good" scientific reasoning looks like.
Translate domain expertise into actionable improvements for model behaviour, reasoning patterns, and tool usage.
Create gold-standard solutions and reference outputs for training and fine-tuning.
Rank and evaluate model outputs to improve scientific reasoning through expert feedback loops.
Help build and operate our external Collaborators program, coordinating expert review cycles and integrating feedback into training and evaluation.
Contribute to high-quality research outputs (benchmarks, datasets, technical reports, preprints) that meet publishable standards.
Who You Are:
PhD in Physics, or a closely related field, with postdoctoral-level research maturity.
Proven ability to reason at a research level in physics and rigorously critique proofs, derivations, and scientific arguments.
Strong scientific programming skills (Python) and experience with research tooling (Git, LaTeX).
Clear, precise communicator with a strong ability to translate complex ideas into sharp critiques, guidance, and evaluation frameworks.
Comfortable working in a fast-moving, cross-functional environment with researchers and engineers.
Mission-driven, with an entrepreneurial mindset and interest in tackling foundational problems in science.
What You'll Get:
The opportunity to work on foundational problems at the intersection of AI and physics.
A high-trust, low-bureaucracy environment with real ownership.
Remote-first work with flexibility in how you structure your day.
Exposure to cutting-edge ideas across AI, scientific discovery, and emerging technologies.
A culture that values curiosity, depth of thinking, and first-principles reasoning.
The chance to contribute to work aimed at advancing fundamental science.
