Elon Musk’s artificial intelligence company, xAI, has laid off around 500 employees from its data annotation team, according to a report by Business Insider. The move, communicated late on Friday evening, affects workers who were responsible for training the company’s generative AI chatbot, Grok.
xAI lays off 500 data annotation staff
As per the report, in an email sent to staff, xAI said it was reducing its focus on developing general AI tutors and would instead concentrate resources on specialist AI tutors. “After a thorough review of our Human Data efforts, we’ve decided to accelerate the expansion and prioritisation of our specialist AI tutors, while scaling back our focus on general AI tutor roles,” the message stated. “As part of this shift in focus, we no longer need most generalist AI tutor positions and your employment with xAI will conclude.”
Employees were told their system access would be revoked immediately. However, salaries would continue to be paid until the end of their contracts or until 30 November, adds the report.
Expansion of specialist AI roles
The company has reportedly made clear it is ramping up investment in specialist AI tutors across fields such as video games, web design, data science, medicine, and STEM. On 13 September, xAI announced plans to expand this team tenfold, saying the roles were “adding huge value”.
Notably, the layoffs follow recent reports that senior members of the data annotation team had their Slack accounts deactivated before the formal announcement was made.
In other news, earlier this month, Musk once again put the spotlight on artificial intelligence, as he highlighted the predictive abilities of X’s AI chatbot, Grok. On his official X account, the billionaire shared a link to a live benchmark platform, urging users to test Grok’s forecasting prowess.
In his first tweet, Musk wrote, “Download the @Grok app and try Grok Expert mode. For serious predictions, Grok Heavy is the best.” He followed up with, “The ability to predict the future is the best measure of intelligence.”
The link pointed to FutureX, a platform designed to evaluate how well large language models (LLMs) can predict real-world events. Developed by Jiashuo Liu and collaborators, FutureX presents AI agents with tasks spanning politics, economics, sports and cultural trends, scoring their predictions in real time.