Elon Musk Reveals: xAI Used OpenAI to Develop Grok

⚡

Key Takeaways

1Elon Musk admitted that xAI used OpenAI models to train its AI Grok, a common practice in the industry.

2OpenAI and Anthropic oppose third-party distillations, particularly in China, which threaten their competitive advantage.

3Musk is suing OpenAI for deviating from its original nonprofit mission, which was discussed during a trial in California.

💡Why it matters — This admission highlights the growing tensions around distillation practices, which could reshape the landscape in the AI sector.

Elon Musk's Revelations on xAI and OpenAI

During testimony in a federal court in California, Elon Musk revealed that his company xAI has used OpenAI models to train its own artificial intelligence model, Grok. This practice, known as distillation, involves using public chatbots and APIs to develop new AI models. Musk stated that this method is common among AI companies, highlighting a widespread practice in the industry.

Distillation: A Threat to AI Giants

OpenAI and Anthropic have recently ramped up their efforts to counter third-party initiatives that utilize distillation. This technique allows for the creation of open-weight models that are nearly as effective as those developed in the United States, but at a significantly lower cost. Discussions have particularly focused on Chinese companies that exploit this method to compete with American giants. However, it is widely assumed that American labs themselves are using these techniques to avoid being outpaced by their competitors.

Elon Musk's Lawsuit Against OpenAI

Elon Musk is currently in a lawsuit against OpenAI, its CEO Sam Altman, and Greg Brockman. He accuses them of straying from OpenAI's original non-profit mission by transforming the entity into a for-profit structure. This trial began this week, and Musk testified about the distillation practices used by xAI. When asked if this indicated full agreement with these practices, he replied: “Partially.”

The Implications of Distillation for the Industry

Musk's admission is significant as distillation threatens the competitive advantage of AI giants. By heavily investing in computing infrastructure, these companies have built an edge that distillation could diminish. This method allows other software publishers to create models that are nearly as effective at a lower cost. There is a certain irony in this situation, given that leading labs have themselves been accused of circumventing copyright rules to obtain the data necessary for training their models.

The Response from AI Leaders

To counter the threats posed by distillation, OpenAI, Anthropic, and Google have launched an initiative through the Frontier Model Forum. This initiative aims to share information to combat distillation attempts, particularly those coming from China. These attempts typically involve systematic querying of models to understand their internal workings. Leading labs are working to prevent users from making suspicious mass queries.

Ranking of AI Providers by Musk

During his testimony, Musk also ranked the top global AI providers. He placed Anthropic at the top, followed by OpenAI, Google, and Chinese open-source models. He described xAI as a smaller company, with only a few hundred employees. OpenAI has not yet commented on Musk's admission at the time of publication.