What's this about?
A Reuters investigation into an internal Meta push to replace substantial numbers of staff with AI systems, and how it fell apart.
What exactly tends to go wrong with these plans?
The failure is usually not capability but decomposition. "Replace this role with AI" assumes the role is a list of tasks, when much of it is context, judgement and knowing which of the written rules are real. Organisations discover the undocumented parts only after removing the people who held them — which is why these initiatives tend to fail in the integration rather than in the model.
What's the community saying?
HN 51 points, 57 comments
- A cynical structural read: some Meta roles exist because of the founder's generosity and image management, and cutting them plays badly.
- The management-fashion comparison — a cargo-cult adoption of a trend, likened to companies adopting Spotify's squads-and-tribes structure and changing nothing.
- Fatalism that this is coming regardless, with the argument being about timing.
- Attention on the tense: the headline says "had" a plan, which several read as the real story.