AI News

OpenAI Dissolves Robotics Team Formed Just 166 Days Ago

Sina TechUpdated Nov 16 20:45

AI summary

According to Fortune, OpenAI has dissolved its independent robotics team formed just 166 days ago; researchers have been merged into the core research organization, while applied robotics and hardware staff have moved into the Applied department, though robotics positions on its careers page have not yet been taken down. The article traces OpenAI's history from the 2016 Dactyl dexterous hand and the 2020 halt of robotics research to the 2025 restart and its investments in 1X and Figure. It notes that NVIDIA research shows GPT-6 Astra achieved a 74.8% success rate on the RoboLab-120 benchmark, higher than the 51% of the specialized policy Lakeside; OpenAI's adjusted operating loss in the third quarter of 2026 reached $10.9 billion.

Why read this

AI score 66

The article reviews how OpenAI adjusted its robotics business three times over ten years, and provides two pieces of background: general model manipulation capabilities and operating losses.

Topics

Tags

Body · AI translation

Add crowdsourced testing promotional copy Add crowdsourced testing promotional copy end Add App-exclusive traffic routing

This is not the first time OpenAI has dissolved a standalone robotics team.

166 days ago, Sam Altman announced the founding of OpenAI Robotics and released 19 job openings all at once, which at the time was once called by the outside world "the ChatGPT moment for humanoid robots." And just recently, OpenAI dissolved this robotics team that had only just been established. According to Fortune, this Robotics team, founded only half a year ago, has been dissolved, with researchers merged into the main model line and hardware personnel merged into the applications department. At this point, the robotics positions on OpenAI's recruitment page with annual salaries of up to $485,000 have not yet been taken down. During these 166 days, investment across the embodied AI industry has continued to increase. NEURA announced financing of up to $1.4 billion; Google DeepMind released Gemini Robotics 2 to advance full-body control of robots; DeepSeek spent approximately 141 million yuan to participate in Unitree Robotics' IPO strategic placement; and Figure reached a partnership with Nscale, planning to use up to 100,000 GPUs to train robot models. Almost every company is preparing for the next round of expansion. Against this backdrop, OpenAI's decision sounds dramatic, but it also fits OpenAI's consistent style. First, this is not the first time OpenAI has dissolved a star team; the last time was Sora. That team, once regarded as OpenAI's multimodal ace, was scaled back and merged into the main line under Altman's directive to "cut all side quests." OpenAI has an almost cold-blooded habit: no matter what happens in the outside world, and no matter how promising this direction looks at the time or how strong the team's paper capabilities are, as long as it is internally determined that this is no longer the shortest path to AGI, or as long as a general model can soon incidentally accomplish this direction, independent side quests will be merged into the main quest, and independent teams will lose their reason to exist. Dissolving the robotics team follows the same logic. And in fact, these 166 days were not an impulsive and clumsy retreat. Looking back 11 years reveals that this is not the first time OpenAI has dissolved a standalone robotics team. What happened during these ten years? OpenAI's first foray into robotics came six years earlier than ChatGPT. In 2016, the newly founded OpenAI listed several things it wanted to do in a research plan. One of them was to build a robot capable of completing basic household chores. At the time, the company had not yet focused its main efforts on language models. Robotics, games, and reinforcement learning were all directions in which it explored artificial general intelligence. That year, OpenAI released the reinforcement learning toolkit Gym; the following year, it launched the robot simulation tool Roboschool. The team hoped to first let machines learn movements in simulated environments, and then apply the learned capabilities to real hardware. Compared with letting robots trial and error repeatedly in a laboratory, simulation can run a large number of experiments simultaneously and does not have to worry about breaking equipment. In 2018, this approach produced a striking result called Dactyl. It was a dexterous robotic hand, and its task was to adjust the orientation of a block in its hand. A block turning in a human hand looks effortless, but it is difficult for a machine: how the fingers coordinate, where force is needed, and how to recover after the block slips all have to be solved. OpenAI trained the control policy in simulation and then transferred it to a physical robotic hand. A year later, the task changed from turning a block to solving a Rubik's cube. On October 15, 2019, OpenAI published a video of Dactyl manipulating a 3x3 Rubik's cube. To adapt policies trained in simulation to reality, the team continuously varied physical parameters in the simulated environment, including friction, object mass, and the motion conditions of the robotic hand. This method, known as "automatic domain randomization," was meant to teach the system to cope with changes that could not be anticipated one by one during training. In the demo, researchers put a glove on the robotic hand, tied up some of its fingers, and even used a plush giraffe to interfere with its movements. The robotic hand would make mistakes and need adjustments, but it could still continue operating. This result had an easily overlooked boundary: how to solve the cube was calculated by another algorithm; what Dactyl had to solve was how to use the robotic hand to execute those steps. Even so, having a robotic hand complete such a complex operation in a real environment was still a notable advance at the time. But this did not make robotics continue to be a focus for OpenAI. Important events for OpenAI from 2016 to 2026. In 2020, OpenAI stopped this line of research. It was not until 2021, when co-founder Wojciech Zaremba discussed the matter on a podcast, that the outside world learned more clearly that the team had been disbanded. The explanation he gave was data. For a robot to learn an action, it needs hardware, space, and experiments. Equipment breaks, collection speed is limited, and if the robotic hand or environment is changed, the original experience may not be directly reusable. By contrast, the internet already has massive amounts of text that can be continuously used to train language models. For a company aiming at artificial general intelligence, the research efficiency of the two directions began to diverge. In 2020, GPT-3 was released, and language models had already demonstrated the possibility of gaining more capabilities by scaling up training. OpenAI concentrated its resources on this route. Over the next two years, the robotics industry continued to move forward. Tesla announced the Optimus plan in 2021 and showed a prototype in 2022. Also in 2022, OpenAI launched ChatGPT. ChatGPT's success quickly changed the way the robotics industry discussed problems. Since a single model can acquire multiple capabilities such as conversation, coding, and reasoning by learning from text, could robots also have a general-purpose "brain" and not need to be programmed separately for each task? OpenAI once again appeared on the funding lists of robotics companies. In March 2023, it led a funding round for the Norwegian robotics company 1X. In February 2024, it participated in Figure's $675 million Series B financing and signed a cooperation agreement with the company to develop models for humanoid robots. This time, OpenAI did not build the hardware itself, but instead partnered with companies that own robot bodies. The demo Figure released in March 2024 made this division of labor intuitive: the robot could answer people's questions while identifying objects on the table, handing over an apple, and clearing the dishes. OpenAI provided model capabilities, and Figure was responsible for connecting them to a real body. But the partnership did not last long. In February 2025, Figure announced the end of its cooperation with OpenAI, shifted to developing its own models, and subsequently released Helix. Figure CEO Brett Adcock emphasized that robot AI needs vertical integration with hardware. This exposed a difficult problem in the partnership model. Model companies can provide language and vision capabilities, but how a robot moves every day, where it fails, and how it completes tasks—these experiences are generated on the hardware side. Who collects the data and how the data is used for the next round of training require close coordination between both parties. For Figure, controlling the model itself can shorten this process. Meanwhile, OpenAI was already rebuilding its own robotics business. In 2024, news gradually emerged that it was hiring robotics researchers. In November of the same year, Caitlin Kalinowski, who had worked on Apple hardware and the Meta Orion glasses, joined OpenAI to lead robotics and consumer hardware. By early 2025, job postings began to involve specific roles in sensors and mechanical structures. OpenAI's plans gradually extended beyond model research to robot bodies. Subsequent media reports also mentioned that it had set up a robotics testing site in San Francisco to collect real motion data through teleoperation. When OpenAI first built robots, it tried to use simulation to make up for the shortcomings of real-world data. After re-entering this field, it began investing personnel and equipment to collect data itself. The industry is also advancing in a similar direction. In 2025, Google DeepMind released Gemini Robotics, and Nvidia launched GR00T N1. Robot manufacturers, model companies, and chip companies are all expanding their investments, trying to connect vision, language understanding, and motion control together. However, OpenAI's internal robotics business has always been in a state of change. In March 2026, Kalinowski resigned over controversy surrounding OpenAI's cooperation with the Pentagon. On May 31, Altman publicly announced that OpenAI Robotics was hiring. In his post, he said that the world simulation research led by Aditya Ramesh had evolved into this team over the past year. This was also the first time the Robotics team made a public appearance. At that time, Altman's goals were very specific: in the short term, build robots that can help skilled workers build infrastructure; in the long term, he hopes everyone can own a personal robot. In the following months, the business appeared to still be expanding. Altman confirmed that OpenAI would develop humanoid robots and proposed a plan to hold a demo in 2027. On the careers page, software, control, mechanical, and hardware positions were still being continuously recruited. Until November 13, Fortune reported that OpenAI had disbanded its independent robotics team and reassigned personnel to other departments. It had been 166 days since Altman publicly recruited for the team. Researchers were transferred into the core research system, while applied robotics and hardware staff entered the Applied department. The independent team was no longer retained, but the related work continued to exist within the company. This differs from how it was handled in 2020. That year, OpenAI stopped robotics research and shifted resources to language models, which were easier to obtain data for and easier to scale up training for. This time, the reason is that general models can already handle some robotics tasks, and robotics remains a long-term direction for the company. Between the two adjustments, OpenAI also tried investment and partnerships to let external companies provide hardware. So over these ten years, OpenAI has been constantly changing the way it participates in robotics—for example, which work it does itself, which it leaves to partners, and which is worth keeping an independent team for. It has never stopped moving forward. OpenAI does not consider this a "retreat." Why did OpenAI choose to disband its robotics team today? Of course, we do not yet know the full internal decision-making process at OpenAI. But one month before the news was announced, a study provided a noteworthy backdrop: general models had begun to achieve results in robot manipulation tests that surpassed specialized policies. On October 14, NVIDIA researchers released Astranomy to test whether GPT-6 Astra could complete manipulation tasks in simulated environments without additional robot policy fine-tuning. On the RoboLab-120 benchmark, the best-performing agent configuration achieved a 74.8% success rate, higher than the 51% of the specialized policy Lakeside on the leaderboard. However, this result relied on a Codex configuration that continuously retained context; directly calling the model to generate actions yielded a 66.3% success rate. As recently as early 2026, the robotics team's internal roadmap still assumed that to give general models manipulation capabilities, it would first be necessary to accumulate physical intuition through a Sora-style video model and then distill it into robot policies. In less than a year, this intermediate step was skipped. In this regard, Carnegie Mellon University professor Guanya Shi's interpretation of OpenAI's disbanding of its robotics team is that, in OpenAI's view, robotics is no longer a special problem; it is merely a downstream application of general intelligence. In other words, judged purely from a technical perspective, this is not a retreat but an absorption, folding the exploration of robotics back into the main mission. But no decision at OpenAI has ever been based on technology alone. Another more realistic backdrop is money. According to media reports, OpenAI's adjusted operating loss in the third quarter of 2026 reached $10.9 billion. The company is also preparing for a possible IPO. Revenue is growing rapidly, and spending continues to expand; management must constantly decide which businesses are worth dedicating separate staff and compute to. In March of this year, Fidji Simo, head of the applications business, already told employees to focus on core businesses and avoid being distracted by side projects. The subsequent contraction of Sora also occurred against this backdrop. In other words, the disbanding of the robotics team is hard to understand entirely apart from the pressure of resource allocation. Technological progress and business pressure may have jointly driven this decision. General models beginning to demonstrate manipulation capabilities gave OpenAI reason to reassess the necessity of separate investment; continuously expanding spending made that assessment more urgent. OpenAI's past success makes it easy to attribute some kind of foresight to this adjustment as well. Leaving robotics in 2020 and concentrating investment in language models did indeed later yield enormous returns. But that outcome cannot prove today's choice in advance. A 74.8% simulation success rate is still a long way from a robot that can work reliably at controllable cost. In this regard, after conversations with many practitioners in the embodied intelligence industry, Huxiu found that they all said today's GPT only shows us that large language models are helpful for solving problems in the embodied industry, but "helpful" does not mean it can replace embodied models; it is still very far from the goal. An important window for observation is 2027; Altman previously said OpenAI plans to showcase its self-developed robot that year. As for whether OpenAI has found a faster path or reduced investment too early, the answer still lies outside the laboratory. Statement: This is an exclusive article from Sina.com. Reproduction without authorization is prohibited. tech_web_article_v2015_block_bottom Non-targeted 300*250 button 17/09 wenjing begin

Non-targeted 300*250 button end

Source

Sina Tech