By Titik Puspa 
In a dramatic personnel reshuffling that underscores the intense competitive pressures in the artificial intelligence landscape, xAI, Elon Musk’s ambitious AI venture, has seen a significant exodus of its founding team. Of the original eleven co-founders who embarked on this journey with Musk just three years ago, only two now remain. This drastic rebuilding effort, Musk asserts, is a deliberate strategy to establish a more robust foundation for the deep learning lab as it vies for supremacy against established giants like Anthropic and OpenAI. "xAI was not built right first time around, so is being rebuilt from the foundations up," Musk stated on his social media platform, X, on Thursday. However, the path to this reconstruction appears fraught with challenges, with the current upheaval suggesting a less-than-smooth transition.
The most immediate catalyst for this intensified overhaul is the fierce competition in the AI sector. This week alone, xAI co-founders Zihang Dai and Guodong Zhang departed the company. Their exit followed Musk’s vocal dissatisfaction with xAI’s AI coding tools, which he deemed insufficiently competitive with industry-leading offerings such as Anthropic’s Claude Code and OpenAI’s Codex. Musk revealed that an all-hands meeting was convened on Wednesday to address the urgent need to bridge this technological gap, expressing confidence that xAI could achieve parity by the middle of the current year.
The strategic importance of advanced AI coding tools cannot be overstated, as they represent a critical revenue stream for AI laboratories. While xAI initially garnered attention through its relatively lax content moderation policies for Grok, its LLM, which allowed for the generation of sexually explicit and even abusive imagery, the long-term financial viability of the company hinges on its ability to develop and monetize sophisticated developer tools. xAI’s current lag in this crucial area is therefore not merely a perceptual issue but a significant business problem that demands immediate and decisive action.
This extensive personnel restructuring is not an isolated incident confined to the past week. Approximately a month prior, eleven senior engineers, including two co-founders, left xAI. Musk characterized these departures as part of a broader reorganization aimed at aligning the company’s structure with its evolving business objectives. However, this initial effort apparently proved insufficient, as the Financial Times reported that executives from SpaceX and Tesla have been seconded to xAI to conduct rigorous employee evaluations and identify underperformers for termination. This infusion of leadership from Musk’s other ventures signals a high-stakes intervention to inject new strategic direction and operational efficiency into xAI.
The formidable task of rebuilding xAI now falls upon the shoulders of the two remaining co-founders, Manuel Kroiss and Ross Nordeen, alongside Elon Musk himself. Their immediate challenge is to not only retain existing talent but also to aggressively recruit new expertise to bolster the company’s capabilities.
In a testament to the urgency of his recruitment drive, Musk revealed on Thursday that he and colleague Baris Akis are personally reviewing rejected employment applications. This initiative aims to identify promising candidates who may have been overlooked during the initial screening process and extend them an invitation for an interview. "My apologies," Musk stated, addressing the multitude of applicants whose applications may have languished unaddressed, signifying a commitment to a more thorough and inclusive hiring approach.
For comparative context, LinkedIn data indicates that xAI currently employs just over 5,000 individuals. This figure positions it behind both OpenAI, with more than 7,500 employees, and Anthropic, which boasts a workforce exceeding 4,700. While xAI’s employee count is substantial, the disparity highlights the scale of talent acquisition required to compete effectively with its more established rivals.
On the hiring front, there are nascent signs of progress. Andrew Milich and Jason Ginsberg are set to join xAI from Cursor, a company specializing in AI coding tools. At Cursor, both individuals held joint responsibility for product engineering. A key distinction between Cursor and xAI is that Cursor relies on frontier labs for access to the AI models it utilizes. Their decision to transition to xAI may signal the perceived value of direct access to cutting-edge large language models (LLMs) and the necessary computing resources to operate them. This move could also suggest that xAI’s proprietary frontier model remains a compelling asset, capable of attracting top-tier talent.
Regardless of these individual hires, the pressure on xAI to demonstrate tangible results is immense, stemming from both internal expectations and external market demands. With xAI now integrated under the SpaceX umbrella and the anticipation of a SpaceX public offering, the cash-intensive AI unit faces heightened scrutiny to prove the uptake and efficacy of Grok, its flagship LLM. A faltering AI division would undoubtedly be an unwelcome narrative for investors considering SpaceX’s future capital raises.
Looking beyond the immediate focus on coding tools, Musk harbors a more ambitious long-term vision centered on xAI’s "Macrohard" project. Musk has humorously described the project’s name as "a funny reference to Microsoft" and its objective is to create an AI agent capable of performing any task a white-collar worker can accomplish on a computer. However, this initiative has also encountered significant setbacks. Toby Pohlen, who was appointed to lead the Macrohard project in February, reportedly departed the company within weeks of his appointment. Furthermore, Business Insider reported this week that the Macrohard project is currently on hold.
In response to these developmental hurdles, Musk has enlisted another of his ventures into the Macrohard effort. He revealed for the first time that Macrohard is a collaborative endeavor with Tesla. Tesla is concurrently developing a complementary AI agent known as "Digital Optimus," a nod to its ongoing work on the Optimus humanoid robot. According to Musk’s description, the xAI language model would serve as the directing intelligence for the Tesla agent, orchestrating its task execution.
This integrated approach, while ambitious, is not entirely novel in the AI landscape. The vision of an AI agent acting as a "digital proxy" for users shares similarities with Perplexity’s new "Everything is Computer" offering, which aims to provide enterprise users with a dedicated assistant capable of managing their digital tasks. This concept also echoes the work being undertaken by entrepreneur Peter Steinberger at OpenAI, following his success in developing popular personal agents for OpenAI’s platform. The AI industry is rapidly evolving, and the race to develop intelligent agents that can seamlessly integrate with and augment human capabilities is intensifying. xAI’s strategic pivot and collaborative efforts with Tesla suggest a concerted push to carve out a significant niche in this burgeoning market, even as it navigates internal challenges and the relentless pace of technological advancement.