Meta CTO Admits AI Reorganization Was 'Atrocious'

In an internal memo seen by WIRED on Monday, Andrew Bosworth, Meta's chief technology officer, admitted the company mishandled the rollout of its new artificial intelligence division. Meta created the Applied AI unit in March 2024 with approximately 6,500 engineers and product managers focused on improving generative AI models. However, employees reported dissatisfaction with the work assignments and organizational changes. Bosworth stated: "We obviously did an atrocious job explaining the vision, giving people a clear picture of how we would support them and their careers in the shift, and painting a picture of how it would change over time." He acknowledged undermining employee trust regarding career advancement, impact, and management stability. To address the unrest, Bosworth announced several measures: capping manager direct reports at approximately 20 each, limiting manager changes during restructurings, focusing managers primarily on management duties, and providing access to AI coaching tools. Bosworth attributed the problems partly to executive focus on broader strategy issues, including competition in AI coding tools, while losing sight of employee concerns. He noted that "there are going to be times where the work requires sacrifice" and that employees may need to work on projects they don't find personally fulfilling. The internal turmoil reflects broader morale issues at Meta following mass layoffs and worker surveillance concerns.