Head of Core AI

Physical Superintelligence•Boston, MA

1d•Onsite

About The Position

Physical Superintelligence is a startup with roots at Google, NVIDIA, Harvard, Meta, MIT, Oxford, Johns Hopkins, Cambridge, and the Perimeter Institute building AI systems to discover new physics at scale. Our mission is to discover and commercialize transformative physics breakthroughs at scale with artificial superintelligence, safely, verifiably, and for broad public benefit. We believe artificial superintelligence will unlock the next golden age of physics. We're creating the infrastructure to industrialize scientific discovery and usher in this new era. We have one product: new physics, at scale. We are seeking a Head of Core AI to lead Core AI: the team that builds the models, agents, and training systems that do physics. This is applied work in service of a product. Everything Core AI builds exists to make PSI's AI-for-physics capability better.

Requirements

A track record of leading AI research teams that shipped: you have run an AI research group at a frontier lab, a fast-moving AI company, or equivalent, and the work made it into production systems or into results the field recognizes.
Deep technical fluency in modern ML: reinforcement learning, agentic systems, large-scale training, and evaluation methodology.
You have personally built and trained systems that beat non-trivial baselines, and you can still do so.
Publications and standing at top ML venues (NeurIPS, ICML, ICLR, or comparable), or an AI research artifact the community has adopted, as evidence of the caliber of your work.
AI research judgment under commercial constraints. You know when a research direction is dead, when a benchmark is being gamed, and when to trade elegance for something that ships this quarter.

Nice To Haves

Enough physics or mathematics background to hold a substantive conversation with domain scientists and to understand whether the science is real.
Hands-on experience applying agents or RL to science, mathematics, code, or other complex-reasoning domains.
Experience scaling an AI research organization through rapid company growth, including hiring, leveling, and performance management.
Experience with AI-for-science specifically: simulators, verification, scientific tool use.

Responsibilities

Set the Core AI roadmap and own the resulting AI research portfolio in service of one goal: making PSI's AI systems better at doing real physics.
Decide how agents acquire physical reasoning, how training infrastructure scales, how reward structures survive long-horizon discovery tasks, and how automated discovery compounds.
Own the people and the delivery, set direction, kill weak lines of work, and double down on what works.
Hold the work to one bar: does it produce physics a real physicist would trust.
Own the training and evaluation infrastructure the rest of the company depends on to run Core AI at scale, and you are accountable for closing the gap between a result that looks good and one that is real.
Stay technical. This is a player-coach role: you carry individual technical work, review work and code, and set the bar in person.
Partner with Applied AI, which turns Core AI's work into products and customer engagements, and with Engineering, which owns the platform infrastructure your systems train, evaluate, and run on.
Hire and develop world-class talent, keeping the bar high while Core AI grows.