'I got drafted': Inside Meta's push to move 7,000 staff into its AI task force

In a significant organizational shift, Meta has announced a major realignment of its workforce focused on artificial intelligence. As the company navigates recent layoffs affecting 8,000 of its 78,000 employees, an alternative path has opened for 7,000 others who have been invited to join a newly formed AI initiative. This initiative, spearheaded by CEO Mark Zuckerberg, aims to enhance Meta's competitive stance in the rapidly evolving AI landscape. The selected employees will be integrated into a group known as Applied AI (AAI), which is overseen by engineering vice president Maher Saba and reports directly to chief technology officer Andrew Bosworth. Additional teams, such as the “Agent Transformation Accelerator” and “Agent Data and Optimization,” have also been established to focus on specific AI projects. An internal communication to those chosen for these roles emphasized their strong performance and technical skills, stating that they were recognized as individuals capable of making a significant impact on the team. For many, these reassignment emails have provided a sense of relief during a tumultuous time, although uncertainty remains about what their new responsibilities will entail. As Meta continues to embrace AI, the company has implemented sweeping changes, including the formation of a superintelligence lab and the creation of smaller, AI-focused teams. The impact of these developments is evident in internal discussions, where some employees have humorously referred to their new assignments as being “drafted.” Despite its ambitious efforts, Meta's AI models have yet to match the performance levels of competitors like OpenAI and Google. The company recognizes the need to gather insights and expertise from its own workforce to improve its AI systems. However, the transition to these new roles is not voluntary, with reactions from employees ranging from anxiety to confusion. In addition to the AAI initiative, Meta has been working on internal AI tools to monitor employee productivity, including software that tracks keystrokes and mouse movements. This data is intended to assist in training AI models to better understand human-computer interactions. Zuckerberg has highlighted the intelligence of Meta employees compared to typical data-labeling contractors, suggesting that leveraging their expertise could provide a competitive edge in AI development. As the company pushes forward, the future of these AI teams will be closely watched by both employees and industry observers alike.