PeopleofSuperintelligence

Researcher

Alex Ray

Alex Ray is an AI researcher whose work spans reinforcement learning, robotics, and language models. In 2011, he studied textile engineering at NC State and helped design an inflatable habitat for a Mars mission. His 2017 and 2018 research included domain randomization, Hindsight Experience Replay, and Dactyl, which trained a robot hand in simulation. In 2021 and 2022, he coauthored the Codex and InstructGPT papers on code generation and language models trained with human feedback.

Works
8
Badges

Publication numbers from OpenAlex, as of 18 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2011

    Designing a Mars habitat

    In 2011, he studies textile engineering at North Carolina State and joins a student team designing an inflatable living space for a Mars mission. Its flexible textile layers shield the habitat against radiation, and a metallic film reflects UV rays.

  2. 2017

    Domain randomization

    In 2017, he coauthors a study on domain randomization, which trains object-localization models on simulated images with randomized rendering so they can transfer to real images. The team uses its detector to grasp objects in a cluttered scene.

  3. 2017

    Hindsight Experience Replay

    In 2017, he coauthors Hindsight Experience Replay, a method for sample-efficient learning from sparse, binary rewards. The team tests it on robotic-arm pushing, sliding, and pick-and-place tasks.

  4. 2018

    Dactyl and dexterity

    In 2018, he coauthors a study that trains a Shadow Dexterous Hand in simulation to reorient objects with a physical robot. Dactyl transfers its policy to a real hand and completes 50 consecutive cube rotations.

  5. 2021

    Writing code with Codex

    In 2021, he coauthors the Codex paper, which introduces a GPT model fine-tuned on public GitHub code and studies its Python code-writing ability. A distinct production version of Codex powers GitHub Copilot.

  6. 2022

    Coauthoring InstructGPT

    In 2022, he coauthors the InstructGPT paper, which fine-tunes GPT-3 with labeler demonstrations and human rankings. In human evaluations, outputs from its 1.3-billion-parameter model are preferred to those from the 175-billion-parameter GPT-3.

A language model matched the links in this timeline to people, organizations and papers (provisional, model-judged).

  1. The median h-index is 55 at universities and 20 at frontier AI labsIncludes Alex Ray. The median h-index is 55 at universities and 20 at frontier AI labs. We compare publishing careers, citations and moves to estimate how universities and frontier labs contribute to AI research.
  2. 11 of 73 people shaping AI with a former OpenAI role now work at AnthropicIncludes OpenAI. Our estimates of where OpenAI recruits came from and where people with former OpenAI roles work now.
  3. Google alumni founded or led 79 other organizations, the most among the labs we compareIncludes OpenAI. We trace 181 organizations founded or led by AI lab alumni, and 187 career moves between the labs.
  4. 44% of AI career moves go from one university to anotherIncludes OpenAI. We estimate how 12,174 AI career moves link universities, companies, labs and government research centers, with net flows and the most common paths.

Sources and credits

Photo: We have no published photo of Alex Ray yet. Send us a free-licensed photo.

  1. Student solutions for manned trip to Mars innovationintextiles.com. Checked 18 October 2026.
  2. Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World arxiv.org. Checked 18 October 2026.
  3. Hindsight Experience Replay arxiv.org. Checked 18 October 2026.
  4. AI Helps Improve the Dexterity of Robots developer.nvidia.com. Checked 18 October 2026.
  5. Learning Dexterous In-Hand Manipulation arxiv.org. Checked 18 October 2026.
  6. Evaluating Large Language Models Trained on Code arxiv.org. Checked 18 October 2026.
  7. Training language models to follow instructions with human feedback proceedings.neurips.cc. Checked 18 October 2026.