Why is Meta destroying its engineering organization?
Source
Gergely Orosz
Author
Gergely Orosz
Date
Key takeaways · AI-distilled
Meta paid $14.8B for 49% of Scale AI, whose expertise is labeled training data and RLHFReinforcement learning from human feedback — training a model to prefer answers humans rate as better, which turns a raw text predictor into a usable assistant.Full definition → pipelines. That purchase says leadership reads its model gap as a data and post-training problem, not a research one.
In late April Meta enrolled engineers into a system that logs every keystroke and click, with no opt-out, to produce training data. The engineering workforce itself became the labeling supply.
Meta is the only one of the big five without its own hardware platform or operating system. That single gap is the through-line from the mobile miss to Oculus to Meta Glasses to the current AI spend.
The culture that worked ran on unusually few processes: little automated testing, sparse documentation, rare inline comments, speed bought by trusting engineers. Top-down mandates remove exactly the ingredient it depended on.
Terms in this piece · Glossary
grounding — Tying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.
RLHF — Reinforcement learning from human feedback — training a model to prefer answers humans rate as better, which turns a raw text predictor into a usable assistant.
Why it matters
A groundingTying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.Full definition →, insider account of how top-down AI mandates and 'always use AI' pressure are reshaping a world-class engineering org — including the failure modes ('AI psychosis', a major outage) that practitioners under similar mandates should recognize and avoid.
Key quotes
“But in the past few weeks, all that has changed, as if the leadership has been following detailed blueprints on how to demolish a proven, successful engineering culture in the most ruthlessly efficient way possible.”
“In late April, Meta told engineers they were being enrolled into a system that tracks every keystroke and click, to produce training data for Meta’s new AI.”
“Meta has around 25,000 engineers, meaning that one in every 5-6 software engineers may now find themselves doing data labeling full time.”
“It’s a strange incentive where an outage caused by a failure to review code properly is not grounds for dismissal, but writing code by hand – instead of having an AI agent write it – could cost you your job”
“Based on all I’ve learned, Meta’s engineering culture is dead because leadership has made it clear that engineering at the company is a cost center .”
