OpenAI Revives Robotics: Four Core Engineer Roles with Salaries Over $300K

OpenAI Robotics hiring

OpenAI Robotics announced four core engineering positions—electrical engineer, simulation‑environment engineer, actuator‑design engineer, and control‑system software engineer—with base salaries of $210,000–$310,000 (≈220 M RMB) plus equity.

Background: Dactyl project (2017‑2019)

Dactyl used a high‑DOF Shadow Hand, reinforcement learning and automatic domain randomization to train in simulation before transferring to the real hand. Achievements included block‑flipping, single‑hand Rubik’s‑cube solving, and robust operation under external perturbations, establishing a benchmark for dexterous manipulation.

The project was discontinued around 2020 because training data for robotics were scarce and iteration speed was slower than for language‑model research, prompting OpenAI to focus on large‑scale language models (e.g., ChatGPT).

Current strategic context

OpenAI is re‑entering the embodied‑AI track, competing with Google DeepMind’s robot‑model research, Tesla’s Optimus production line, and Figure AI’s humanoid benchmarks.

Team composition

Leadership by Aditya Ramesh’s world‑simulation group; senior researchers include:

Lin Xingyu – CMU PhD, contributed to GELLO and HumanoidBench.

He Tairan – former tech blogger, developed Omni H2O for whole‑body coordination.

Lawrence Chen – BAIR postdoc, former NVIDIA intern, focuses on robot learning.

Li Chengshu – Stanford, works on HumanoidBench.

Ying Hang – Stanford Vision & Learning, involved in BEHAVIOR‑1K.

Zhang Pengchuan – Meta FAIR, core contributor to SAM and Llama.

Zhao Jialiang – MIT CSAIL, researches world‑simulation integration with robotics.

Research focus

Robot learning & dexterous manipulation – building on Dactyl’s simulation‑to‑real pipeline.

Simulation, benchmark infrastructure, and dataset construction – e.g., GELLO, HumanoidBench, BEHAVIOR‑1K.

Transferring world‑simulation techniques to physical robots – combining visual perception, world models, and control.

Short‑term and long‑term objectives

Short term: develop robots that assist construction workers and infrastructure projects.

Long term: create personal robots for household chores, aiming for widespread deployment.

References

[1] https://x.com/sama/status/2061117302528188712

[2] https://cryptobriefing.com/openai-robotics-hiring-engineers/