Elon Musk, who has hinted for months that he wants to build an alternative to the popular ChatGPT artificial intelligence chatbot, announced the formation of what he’s calling xAI, whose goal is to “understand the true nature of the universe.”
On a website unveiled Wednesday xAI said its team will be led by Musk and staffed by executives who have worked at a broad range of companies at the forefront of artificial intelligence, including Google’s DeepMind, Microsoft Corp. and Tesla Inc., as well as academic institutions such as the University of Toronto.
Musk was involved in the creation of OpenAI, the highest-profile AI startup and developer of ChatGPT. But he has frequently and publicly criticized OpenAI since he left the board in 2018, especially after it created a for-profit arm the following year. He has said he believes it to be “effectively controlled by Microsoft.” Microsoft has invested some $13 billion into OpenAI.
Despite his work in AI, Musk has expressed deep reservations about the technology. The billionaire was among a group of researchers and tech industry leaders who in March called for developers to pause the training of powerful AI models.
The 12 men listed on the website as of Wednesday morning (including Musk) include Jimmy Ba, an assistant professor at the University of Toronto who studied under AI pioneer Geoffrey Hinton, and Christian Szegedy, who spent years as a research scientist working on AI at Google.
Though Musk is a frequent critic of San Francisco, the xAI website says that the company is “actively recruiting experienced engineers and researchers” to work “in the Bay Area.” So far, most AI development has been concentrated in Silicon Valley.
Musk and Jared Birchall, who operates Musk’s family office, incorporated a business called X.AI in March, according to a Nevada state filing with the Secretary of State.
In April, the Financial Times reported that Musk was holding discussions with Tesla and Space Exploration Technologies Corp. investors about helping fund an AI startup, citing unidentified people familiar with the matter. The billionaire has acquired thousands of processors from Nvidia Corp. for the new project, the paper said.
The xAI website said the company is being advised by Dan Hendrycks, who is the director of the Center for AI Safety — a group that has warned about what it sees as existential dangers of developing AI quickly. This spring, it released a letter of caution signed by chief executive officers of some of the leading companies in AI, including Alphabet Inc.’s DeepMind and OpenAI.
Musk, 52, now oversees six companies: Tesla, SpaceX, Twitter, Neuralink, the Boring Co. and now xAI. In regulatory filings, Tesla says the auto giant is “increasingly focused on products and services based on artificial intelligence, robotics and automation.” Tesla’s website invites people to help “build the future of artificial intelligence” with a variety of products, from the “Tesla Bot” known as Optimus to AI interface chips that will run the electric automaker’s automated driving software.
Musk has a long history of borrowing engineers from one company to help out at another, as the contours of his ever-expanding empire bleed into one another. Tesla and SpaceX share a vice president of materials engineering, for example, and engineers from Tesla “volunteered” to work at Twitter after Musk bought the company for $44 billion in October.
The xAI website says that it is a “separate company from X Corp,” the parent company that Musk merged Twitter into earlier this year, but that it will “work closely with X (Twitter), Tesla, and other companies.”
--With assistance from Dana Hull and Sean O'Kane.
(Updates with context starting in the second paragraph.)
