Newshot
Account

OpenAI Wants General AI Like GPT-6 Astra to Be Robots' Brains: Announces Reorganization of Embodied AI Team

Source: IT之家 Published: AI 中文 AIGPTOpenAIGPT-6Astra
AI summary

Fortune reported on November 13 that OpenAI has disbanded its standalone robotics team and reassigned its staff to other departments; OpenAI confirmed the organizational change, stressing that robotics remains core to its long-term vision and that the restructuring aims to accelerate R&D. The robotics researchers are now folded into the core research team reporting to Chief Research Officer Mark Chen (Mark Chen), while the applied robotics and hardware team moves to the larger applications division led by President Greg Brockman (Greg Brockman). OpenAI said general models including GPT-6 Astra can already handle some robotics tasks directly, so it is integrating robotics research into general model development; a study cited in the report found that an Astra-driven agent achieved a 74.8% success rate in a simulated manipulation test, above the 51.0% of the leading specialized model on that benchmark.

Translation(model translation, not the source's own text)

To Make General AI Like GPT-6 Astra the "Brain" of Robots: OpenAI Announces Reorganization of Its Embodied Intelligence Team

IT之家 November 14 news: 《财富》 yesterday (November 13) published a blog post revealing that OpenAI has disbanded its independent robotics team and reassigned the relevant personnel to other departments within the company. OpenAI confirmed this organizational adjustment to the outlet, but stressed that this adjustment is not an exit from the robotics field — robotics remains central to its long-term vision, and the reorganization is intended to accelerate R&D.

Regarding the structural adjustment, robotics researchers have been merged into the core research team, reporting to Chief Research Officer Mark Chen (Mark Chen); the applied robotics and hardware team has been moved into the larger applications division, which is led by company President Greg Brockman (Greg Brockman). IT之家 attaches the relevant screenshot below;d

As for the root cause of this structural adjustment, OpenAI's official response stated that, including GPT-6 Astra, general models can already be used directly for some robotic tasks, and therefore it decided to further integrate the relevant research into general model development.

OpenAI hopes to make AI that can understand text, images, and code also the "brain" of robots, rather than relying entirely on models developed specifically for robots. The report cites a study by 英伟达 stating that an Astra-driven agent achieved a 74.8% success rate in a simulated manipulation test, higher than the 51.0% of the leading specialized model on that leaderboard.

Whether general models can fully take over robotics still faces practical challenges. The scholars interviewed pointed out that the running speed of large models and the need for cloud connectivity constitute limitations, and that the on-site experience and data accumulated by robotics companies still hold value. According to reports, OpenAI is developing its own humanoid robot, with the goal of its first public demonstration in 2027.

Related Articles

  • Chengdu's fifth-generation Gongga humanoid robot unveiled, claiming to possess "the world's most leading core technology for humanoid robot brains"
  • Figure AI is hiring night-shift supervisors to look after humanoid robots, with an hourly wage of up to 45 USD
  • Apple discloses its humanoid robot HALO system: calls on the Qwen model for validation on the Unitree G1, with an average success rate of 91% per single task
  • 6 AgiBot robots work 24-hour shifts livestreamed at a convenience store, with 4 successfully "becoming regular employees"
  • Xiaomi robots' "dual-line breakthrough": successfully meeting the challenge of 8-hour factory operations and long-range household tasks
  • Results of the first CMG World Robot Mount Tai Climbing Competition announced, with the champion finishing in as fast as 27 minutes 26 seconds

The translation is machine-generated and individual terms may be off. Nothing is rewritten or cut; the original is authoritative.

Published by IT之家
This site only summarises and aggregates; copyright stays with the original author. Headlines and summaries follow the source — where they differ, the original wins.
Read the original ↗