The landscape has been tumultuous for xAI over recent days, witnessing a wave of exits among its leadership and staff. Notably, earlier this week, cofounder Yuhuai (Tony) Wu announced his departure, signaling it was time for him to embark on a “new chapter.” Just hours later, fellow cofounder Jimmy Ba echoed this sentiment, stating it was time to “recalibrate [his] gradient on the big picture.” These significant departures have left xAI with only half of its original 12 cofounders, as other employees have also shared their intentions to leave, some even venturing to launch their own AI startups.

Elon Musk’s ambitious AI venture, xAI, finds itself under the same expansive umbrella as SpaceX and the social media entity, X, following a series of mergers and acquisitions. The recent merger with SpaceX has sparked a flurry of speculation, especially regarding the company’s future trajectory and its staggering $1.25 trillion valuation. Musk has hinted at grand plans, including “space-based AI” data centers and an unprecedented “innovation engine” designed to operate both on Earth and beyond. During a recent internal meeting at xAI, Musk reportedly unveiled bold aspirations to construct an AI satellite factory and even establish a city on the Moon.

Amid these sweeping changes, it’s not uncommon for companies to experience a reshuffling of teams post-merger. Musk himself acknowledged that some departures were part of a reorganization that, while necessary, led to “parting ways with some people.” However, it seems there are also undercurrents of dissatisfaction with Musk’s strategic direction.

According to a former employee who spoke anonymously with The Verge, there is a growing sense of disillusionment within xAI. The source cited concerns over the company’s prioritization of NSFW Grok creations and what they perceived as a lack of focus on safety. Moreover, there was a sentiment that xAI was trapped in a perpetual state of catching up, rather than forging new paths distinct from competitors like OpenAI or Anthropic. “We iterated rapidly,” the source noted, “but never achieved a breakthrough moment that outpaced what others had released.”

The merger with SpaceX has reportedly resulted in xAI shareholders receiving $250 billion in new shares, potentially providing employees with the capital to explore their own entrepreneurial ventures. Former employee Vahid Kazemi expressed his views on X, stating, “all AI labs are building the exact same thing, and it’s boring. I think there’s room for more creativity. So, I’m starting something new.” Another ex-staffer shared his plans to create a venture aimed at “accelerating science,” underscoring the drive among former xAI personnel to channel their expertise into fresh, innovative projects.

Yet another former employee said he was launching an AI infrastructure company called Nuraline alongside other ex-xAI employees. He wrote, “During my time at xAI, I got to see a clear path towards hill climbing any problem that can be defined in a measurable way. At the same time, I’ve seen how raw intelligence can get lobotomized by the finest human errors … Learning shouldn’t stop at the model weights, but continue to improve every part of an AI system.”

Musk posted a recording of xAI’s 45-minute internal all-hands meeting that announced the changes, adding that xAI will be categorized into four main areas: Grok Main and Voice (the main Grok AI model), Coding, Imagine (image and video), and Macrohard (“which is intended to do full digital emulation of entire companies,” Musk said).

The source who departed earlier this year said Grok’s turn toward NSFW content was due partly to the safety team being let go, with little to no remaining safety review process for the models besides basic filters for things like CSAM. “Safety is a dead org at xAI,” he said. Looking at the restructured org chart Elon Musk shared on X, there’s no mention of a safety team. The source also said that during his time at xAI, he felt leadership had a lot of differing opinions on which product features to prioritize and that infighting sometimes stalled progress. A lot of decisions on what to ship happen via an all-company group chat on X with Musk in it, he said.

A second source, who departed xAI before the recent restructuring and requested anonymity, echoed the thoughts 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’s almost zero risky bet. If something hasn’t been done before we’re not going to do it.”

He mentioned xAI’s lack of safety focus too, an issue that was also highlighted in reporting by The Washington Post earlier this month.

“There is zero safety whatsoever in the company — not in the image [model], not in the chatbot,” the second source said. “He [Musk] actively is trying to make the model more unhinged because safety means censorship, in a sense, to him.”

The second source also said engineers at xAI immediately “push to prod[uction]” and that for a long time, there was no human review involved.

“You survive by shutting up and doing what Elon wants,” he said.