What’s behind the mass exodus at xAI?

The past few days have been very tumultuous for xAI, which has been racking up employee and co-founder departure announcements left and right. On Tuesday and Wednesday, co-founder Yuhui (Tony) Wu announced his departure, saying it was time to [his] “Next Chapter,” co-founder Jimmy Bae wrote later that day, with a similar post saying it was “time to recalibrate.” [his] Shield on the bigger picture.” The departures mean xAI now has only half its original 12 co-founders on staff. Several employees also went to XAI to announce that they were leaving XAI, with some announcing that they were starting their own AI companies.

Elon Musk’s AI startup, through some merger/acquisition or some other means, is under the same umbrella as both his space company, SpaceX, and his social media platform, X. Since the SpaceX merger announcement last week, rumors have swirled about the $1.25 trillion valuation and future plans for the all-in-one company, which Musk announced will include a “space-based AI” data center and “the most ambitious, vertically-integrated innovation engine on (and off) Earth.” At an internal meeting at xAI on Tuesday, Musk reportedly talked about plans to build an AI satellite factory and city on the Moon.

Mergers are often a natural departure point for companies, and Musk has announced that some of the departures were a restructuring that “unfortunately required parting ways with some people.” But there are also signs that people are not liking the direction Musk has taken things.

a source who spoke The Verge A person familiar with the events inside the company, who left earlier this year and requested anonymity due to fear of retribution, said many people at the company were frustrated by XAI’s focus on NSFW grok creations and disregard for security. The source also felt that the company was “stuck in the catch-up phase” and was not doing anything new or fundamentally different from its competitors. “Although we were iterating really fast, we never got to the point of, ‘Oh, we’ve made a step function change compared to what OpenAI or Anthropic or other companies have released,'” he said.

The SpaceX merger meant that a reported $250 billion of new shares were issued to XAI shareholders, so employees with equity have more runway to fund their ideas. A former employee, Wahid Kazemi, wrote on X that “All the AI ​​labs are building exactly the same thing, and it’s boring. I think there’s room for more creativity. So, I’m starting something new.” Another former employee said he left the company to “build something new with a focus on accelerating science.”

Yet another former employee said he was launching an AI infrastructure company called Noorline with other ex-XAI employees. He wrote, “During my time at

Musk posted a recording of xAI’s 45-minute internal all-hands meeting in which the changes were announced, saying that xAI would be categorized into four main areas: Grok Main and Voice (the main Grok AI model), Coding, Imagine (image and video), and Macrohard (“which aims to create full digital simulations of entire companies,” Musk said).

The source, who left earlier this year, said Grok’s trend toward NSFW content was partly due to the security team being let go, which left no security review process for models other than basic filters for things like CSAM. “Security at XAI is a dead organization,” he said. Looking at the reorganized organization chart shared by Elon Musk on X, there is no mention of a security team. The source also said that during his time at XAI, he realized that leadership had many different opinions on which product features should be prioritized and that progress sometimes stalled due to infighting. A lot of the decisions on what to send are made with Musk through an all-company group chat on X, he said.

A second source, who left XAI before the recent restructuring and requested anonymity, echoed the view that Musk’s company was playing catch-up. “Trying to do what OpenAI was doing a year ago is not how you beat OpenAI,” he said. “Everything is a catch-up. There are almost zero risky bets. If something hasn’t been done before we’re not going to do it.”

He also noted XAI’s lack of security focus, an issue that was also highlighted in the reporting Washington Post earlier this month.

“There is no security of any kind in the company – not in the image [model]Not in a chatbot,” the second source said. ”He [Musk] He is actively trying to make the model more independent because for him security means a kind of censorship.

The second source also said that XAI engineers “immediately ramped up production.”[uction]”And for a long time, there was no human review involved.

He said, “You can survive by keeping quiet and doing what Elon wants.”