The CEO of JoyIn Raises Concerns About Similarities with OpenAI
The CEO of an Ant-backed humanoid robotics startup, JoyIn, has recently published a letter to OpenAI in Chinese, questioning the technical and design similarities between the two companies’ recent models. Guo Renjie, the CEO of Suzhou-based JoyIn, expressed his concerns in a public statement, stating that the similarities between the two companies’ models are a direct reflection of JoyIn’s work without any modifications.
According to Guo, JoyIn had publicly presented its “extraterrestrial visitor” AI model framework in Silicon Valley a few weeks prior to OpenAI’s Chief Scientist publishing “An Alien Mind” on September 6. He pointed out similarities in core technological approaches such as “recursive self-improvement” and the use of AI to optimize computing power.
Guo also highlighted similarities in the design of OpenAI’s GPT-6 Astra webpage and JoyIn’s Aether model website, which he claims was released two months earlier. As a result, JoyIn has initiated the process of filing a lawsuit against OpenAI.
While some concepts mentioned in the claims are part of AI research more broadly, companies often implement them differently. OpenAI has not yet responded to the allegations made by JoyIn.
Anthropic, a U.S.-based company, has previously raised concerns about unauthorized use of its models by Chinese companies to enhance their AI capabilities. Recently, a U.S. cybersecurity agency identified six Chinese companies, including DeepSeek and Alibaba, for distilling models from companies like Anthropic, Google, and OpenAI.
JoyIn’s Aether model, which was publicly announced by Guo on Thursday, boasts a perceptive approach to robotic control that allows humanoids to achieve a 90% success rate on tasks on the first attempt. Zhu Mingxuan, the lead developer of the model, mentioned that she had previously worked on similar projects at a U.S. humanoid company and that Aether has significantly reduced training time.
Developing humanoids and AI models that can perform tasks as intelligently as humans has been a challenging endeavor, especially amidst the ongoing tech race between U.S. and Chinese companies. Garry Tan, the CEO of Y Combinator, expressed his stance on the issue, indicating that U.S. companies have trained their AI models on data covered under copyright law.
This ongoing dispute highlights the complexities and challenges present in the development and deployment of advanced AI technologies in the global market.
—Contributors: CNBC’s Kate Rooney and Isabel O’Brien
