Skip to main content
Elon Musk's xAI logo and Grok chatbot on phone, symbolizing staff exits and AI challenges [techcrunch.com](https://techcrunch

Editorial illustration for xAI faces staff exodus as human errors blunt raw AI intelligence

xAI Loses Half Its Founding Team in Mass Exodus

xAI faces staff exodus as human errors blunt raw AI intelligence

Updated: 3 min read

The raw intelligence of xAI’s models is being undermined by the very humans who build them. That’s the blunt admission buried inside a staff exodus that is now reshaping Elon Musk’s AI venture. Safety is dead, internal critics declare.

Learning, Musk insists, shouldn’t stop at the model weights. But as he reorganizes the company into four distinct pillars, Grok, Coding, Imagine, and the audacious Macrohard, the question hangs in the air: can any restructuring fix the fundamental human errors that keep blunting the edge?

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." "Safety is a dead org at xAI." 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 raw intelligence of a system is only as sound as the hands that feed it. xAI’s restructuring into Grok, Coding, Imagine, and Macrohard is a strategic gambit, a bid to compartmentalize chaos. But you cannot code your way out of a culture where safety is declared dead, where human error lobotomizes the very models you’re racing to perfect.

Musk’s all-hands recording was a performance of control, yet the exodus tells a different story: talent votes with its feet. And when the brightest walk, what remains is not intelligence, but the echo of it, amplified, brittle, and at the mercy of the next mistake. The lesson is brutal: learning must never end at the weights.

It has to infect every process, every person, every policy. If it doesn’t, the purest AI becomes a monument to human failure.

Common Questions Answered

How many of xAI's original co-founders have now left the company?

Six of xAI's twelve original co-founders have now left the company, which represents exactly half of the founding team. The departures include notable names like Igor Babuschkin, Yuhuai (Tony) Wu, Kyle Kosier, Greg Yang, Christian Szegedy, and most recently Jimmy Ba.

What did Jimmy Ba say about xAI's mission before leaving the company?

In his farewell post, Ba discussed xAI's mission to advance humanity on the Kardashev technology scale. He also predicted that recursive self-improvement loops—AI systems that improve themselves—could potentially 'likely go live' within the next twelve months.

What recent corporate action has impacted xAI's structure?

SpaceX recently announced a takeover of xAI that values SpaceX at one trillion dollars and xAI at 250 billion dollars. This acquisition is notable because xAI currently generates almost no revenue, despite the enormous costs of developing and running its AI models.

LIVE03:21OpenAI's Miles Wang in Talks for USD 2B AI Drug Discovery Startup