Bob McGrew

AI researcher; former chief research officer, OpenAI

Bob McGrew is the founder of Arda, an AI manufacturing software company, and a former chief research officer at OpenAI. His research spans reinforcement learning, robotics and large language models.

Explore the coverage
Bob McGrew
Photo: U.S. Army photo by Leroy Council; cropped to Bob McGrewPublic domain · Displayed in grayscale

At a glance

Bob McGrew is the founder of Arda, an AI manufacturing software company, and a former chief research officer at OpenAI. His research spans reinforcement learning, robotics and large language models.

Current work
Founder, Arda
Known for
AI research leadership; reinforcement learning; robotics
Previously
Chief research officer, OpenAI
Research
Dexterous manipulation; language models

Sources checked . Selected career background and milestones.

Career, in context.

Selected research, leadership and career milestones, newest first.

  1. Launches Arda from stealth

    Accel confirms McGrew is building AI software for manufacturing and reports approximately $230 million in total funding. The investor co-led Arda’s seed and Series A rounds.

    Source: Accel ↗
  2. Reportedly co-founds Arda

    The Information reports McGrew is co-founding Arda, an AI manufacturing startup.

    Source: The Information ↗
  3. Joins the Army’s Executive Innovation Corps

    The U.S. Army commissions McGrew as a lieutenant colonel in the Army Reserve’s Detachment 201, drawing on his AI expertise.

    Source: U.S. Army / DVIDS ↗
  4. Contributes to OpenAI o1

    OpenAI lists McGrew among the leadership contributors to its o1 reasoning-model program.

    Source: OpenAI ↗
  5. Contributes to GPT-4o

    OpenAI credits McGrew among the research and organizational leaders behind GPT-4o.

    Source: OpenAI ↗
  6. Co-authors robotic Rubik’s Cube research

    OpenAI demonstrates a robotic hand solving a Rubik’s Cube, with McGrew credited as a research contributor.

    Source: OpenAI ↗
  7. Co-authors dexterous manipulation research

    Learning Dexterous In-Hand Manipulation studies reinforcement learning for a robotic hand trained in simulation and transferred to real hardware.

    Source: Research paper ↗
  8. Co-authors Hindsight Experience Replay

    The research introduces a way to learn from unsuccessful goal-directed experiences in reinforcement learning.

    Source: OpenAI ↗

The latest on Bob McGrew

Newest first
2 min read · Members

Bob McGrew's Arda Raises $232 Million

Bob McGrew's Arda raised $232 million for an assembly line agent that cut Anduril's instruction writing from 10 hours to 15 minutes.

AI Secret

AI Secret Membership

Choose your edge.

Independent AI analysis. Your own agent to take it further.

AI Secret

AI Secret Membership · Step 1 of 2

Your next chapter starts here.

Enter your email to choose your membership. Use the same email if you already read AI Secret.

Already a member? Sign in

AI Secret

Your free daily briefing

Check your inbox.

We've sent a confirmation link to . Follow the link to verify your email and complete your free subscription.

Free membership includes the daily briefing. Paid Deep Dives require a paid membership.