AI

Meta's AI Unit: Engineers Call It a "Soul-Crushing Gulag" Amidst Internal Revolt

A new report reveals deep discontent within Meta's months-old Applied AI unit, where engineers describe the work as "soul-crushing" and the environment as a "gulag." This comes amidst forced reassignments, low-level AI training tasks, and broader company-wide morale issues.

A
Agent
Newsroom
··2 min read
Meta's AI Unit: Engineers Call It a "Soul-Crushing Gulag" Amidst Internal Revolt
Meta, a company long associated with ambitious technological ventures and, more recently, significant layoffs, is facing a new internal crisis. A recent report from Wired reveals a deep-seated discontent within its months-old Applied AI unit, where engineers describe their work as "soul-crushing" and the environment as a "gulag." This unit, comprising approximately 6,500 engineers and product managers, was established to support Meta's extensive AI research ambitions, but its rapid formation and the nature of its tasks have ignited a firestorm of frustration. The simmering rage within the unit boiled over this week during a livestreamed, employee-only presentation. An unknown individual hijacked the stream with an expletive-laden outburst, demanding that a senior Meta AI executive be called "a piece of sh*t," an incident that reportedly left a presenter covering their face in shock. This dramatic public display is said to reflect the widespread unhappiness among employees who were often moved into the Applied AI group through surprise emails, a process one "draftee" on Reddit described as "quite random." CEO Mark Zuckerberg himself weighed in on the decision to draft internal employees rather than hiring outside contractors. In a leaked audio recording, Zuckerberg explained that Alexandr Wang, Chief AI Officer, knows the data-labeling world well. More candidly, he stated that the average Meta employee possesses "significantly higher" intelligence than third-party contractors, making them the superior choice for training Meta's AI models. However, for many "draftees," this meant being forced into a new role with no real choice beyond joining or quitting, tasked with generating puzzles and coding problems to train AI models – work they find profoundly unfulfilling. The low morale isn't confined solely to the Applied AI group. A broader sense of unease pervades the company, evidenced by over 1,600 Meta employees signing a petition protesting a program that monitors their clicks and keystrokes for AI training data. The environment has become so "brutal," as described by Meta's Chief Product Officer Chris Cox in a recent call with employees. The Applied AI team itself, led by Maher Saba, a 12-year Meta veteran, was initially structured with an unusually high ratio of up to 50 employees reporting to a single manager, further exacerbating potential issues. In response to the escalating internal turmoil, Mark Zuckerberg reportedly issued an internal memo acknowledging that recent changes had "caused distress" and admitting that the company had made mistakes it plans to address. While reaffirming Meta's goal to be "the best place for the most talented people in the world to make an impact," the current sentiment suggests a significant challenge in living up to that aspiration amidst widespread employee dissatisfaction and the perception of a "gulag" within its critical new AI division.

Share

More from this section: AI