After suing OpenAI this month, claiming the company was too closed-minded, Elon Musk said he would make the “truth-seeking” answers to his chatbot Grok, ChatGPT, public for anyone to download and use.
“This week @xAI will open source Grok.” musk wrote today on his social media platform X. This means that his AI company, xAI, will make Grok’s entire code public and allow anyone to use or change it. In contrast, OpenAI makes its version of ChatGPT and the language model behind it freely available, but keeps the code private.
Musk previously said little about Grok or xAI’s business model, and the chatbot was only available to X’s premium subscribers. After accusing the OpenAI co-founder of breaking a promise earlier this month to deliver the company’s artificial intelligence, Musk felt he should open source his chatbot, perhaps to show he was committed to that vision.
OpenAI responded to Musk’s lawsuit last week by releasing email messages between him and others, appearing to support the idea of making the company’s technology more closed as it becomes more powerful. Musk ultimately invested more than $40 million in OpenAI before discontinuing the project in 2018.
When Musk first announced that Grok was in development, he promised that it would be less politically biased than ChatGPT or other AI models. Testing by WIRED and others quickly showed that while Grok can adopt a provocative style, it is not significantly biased one way or the other. This may perhaps reveal the challenge of consistently fitting AI models to a particular perspective.
Open-source Grok could help Musk increase his company’s interest in AI. Restricting Grok access to paying subscribers of X, one of the smaller global social platforms, means it doesn’t yet have the same appeal as OpenAI’s ChatGPT or Google’s Gemini. Launching Grok will encourage developers to use and build on our models, which can ultimately help us reach more end users. This can provide xAI with data that it can use to improve its technology.
Musk’s move to free Grok is something he sees as consistent with Meta’s approach to generative AI. Meta’s open source models, such as Llama 2, have become popular among developers because they can be fully customized and adapted to a variety of uses. But adopting a similar strategy could bring Musk further into the growing debate about the benefits and risks of giving anyone access to the most powerful AI models.
Many AI experts argue that open source AI models offer significant benefits, including greater transparency and greater accessibility. “Open models are safer and more powerful, and we’re excited to see more options from leading companies in this space,” says Emad Mostaque, founder of Stability AI, a company that builds a variety of open source AI models.