A recent internal Meta livestream took an unexpected turn when an employee launched into an expletive-laden outburst, expressing deep frustration about their role within the company. This dramatic incident, heard on a recording by WIRED, included a directive to tell a specific Meta AI executive that “he’s a piece of shit.” One presenter reportedly covered their face in response, while leaders quickly moved on, urging attendees to mute their microphones as employees commented on the “spicy” start.
This public display of anger, occurring during a call open to thousands of employees, isn’t an isolated event. It reflects a growing tide of dissatisfaction within Meta, particularly among its Applied AI team. Formed in March to bolster the work of AI researchers at Meta Superintelligence Labs, this unit of approximately 6,500 engineers and product managers is reportedly experiencing widespread discontent over how it was assembled and the “drudgework” they feel they’ve been assigned.
Life Inside Meta’s Applied AI Team: “The Gulag”
Three current employees, speaking anonymously, describe the Applied AI unit as a challenging environment. One individual starkly referred to it as “literally the gulag,” lamenting a sudden loss of purpose and minimal interaction, replaced by weekly tasks that feel unfulfilling. This sentiment highlights a profound shift for many who were previously engaged in more complex software development.
Some of the assigned tasks include generating intricate puzzles to test the reliability of AI models from Meta and other companies. While some employees admit the work itself can be easy compared to their prior roles, the projects are often perceived as menial and “soul-crushing.” The consensus among most employees seems to be one of profound unhappiness, feeling their extensive skills are being underutilized in these new, data-centric roles.
Engineers drafted into the Applied AI unit often had no choice but to join or leave the company—an unusual mandate for highly sought-after technical talent in Silicon Valley. This has led some members to describe themselves as “draftees,” a term that underscores their lack of agency in the transition. The phased growth of the organization has created a strange dynamic, with early members observing new waves of employees experiencing the “shock of it” as they arrive.
Company-Wide Tensions and Leadership’s Candid Remarks
The frustration isn’t limited to Applied AI; it’s a symptom of broader tensions impacting morale across Meta. The company’s extensive AI-focused restructuring, which included the layoff of 8,000 employees last month (about 10 percent of the company), has created significant stress and additional workload in various divisions, from data center engineering to Instagram.
Further exacerbating internal unease, more than 1,600 employees signed a petition demanding Meta halt a new initiative to monitor US employees’ clicks and keystrokes for AI training data. While Meta has since scaled back the program, allowing temporary pauses and specific exemptions, the initial rollout sparked considerable concern. These pressures contribute to what many workers describe as record-low morale.
In a recent all-hands meeting at Instagram, Meta chief product officer Chris Cox candidly addressed the “difficult” and “brutal” environment caused by the “insanity of this company” in recent months. Cox commended Instagram employees for launching new features and serving 2 billion users amidst what he likened to “running a marathon in the middle of a hailstorm and then, like, your teammate gets replaced and then we’re recording you.” He echoed employee sentiment, declaring, “It is like what the fuck,” drawing laughs from the audience.
Cox emphasized the need for leaders to “get in touch with the company again” and avoid being “overearnest” about AI’s power. He pragmatically stated that AI “is neither god, nor is it the devil,” acknowledging it’s “nowhere near as good as you think it is, and it is nowhere near as bad as you think it is.” His comments underscored the dynamic, often unpredictable nature of AI development.
Zuckerberg’s Acknowledgment and Path Forward
Meta CEO Mark Zuckerberg also acknowledged the widespread distress caused by recent organizational changes in an internal memo, seen by WIRED. He admitted that “we’ve made mistakes and will almost certainly make more” during this complex period. Zuckerberg committed to providing as much stability as possible going forward, a crucial assurance for a workforce reeling from recent shifts.
Reiterating a promise to avoid further mass layoffs this year, Zuckerberg outlined several initiatives aimed at improving employee experience. These include limiting the number of employees per manager, a ratio that had ballooned to 50-to-one on some teams like Applied AI. He also pledged increased budgets for team events, a large hackathon next month to foster unity, and the reintroduction of assigned desks in many locations by year-end.
Addressing the situation in Applied AI directly, Zuckerberg characterized the team as a “waypoint, not a destination.” He stressed that work like that done by AAI is “critical to advancing our models” and allows talented individuals to contribute while other roles are being created across Meta. This perspective aims to reassure those feeling stuck that their contributions are valued and that new opportunities will emerge.
Meta’s ambitious AI strategy, despite mixed results from previous open-weight model releases, involves significant investment in initiatives like Applied AI to compete in the burgeoning AI services market. Zuckerberg clarified that Meta’s primary focus isn’t simply automating work. Instead, their “north star is to be the best place for the most talented people in the world to make an impact,” building products from personalized experiences on Instagram and Facebook to glasses and “personal superintelligence agents.”
Source: Wired – AI
