Meta's Superintelligence Lab Runs 200+ Direct Reports—A Lesson for Flat Multi-Agent Systems
peterjliu · x · 2026-10-16
Alexandr Wang says more than 200 people report directly to him at Meta's superintelligence lab—by design. The TBD group is "anti-bureaucracy," using a pod structure with technical leads for tie-breaks instead of middle-management chains. Peter Liu extends the analogy to AI: deep hierarchies of agents relaying text suffer the same message-corruption as human management chains, but agents' large effective context windows mean you don't need hierarchies at all—a flat multi-agent organization with at most one thin coordinator layer is likely optimal.
Related event: Meta Superintelligence Labs runs flat with 200+ direct reports(2 posts)→
More from AGI Musings
- Minerva brings AI to a 3,000-person village and acquires accounting firms across the US — ycombinator · 2026-10-16
- Interpretability is too costly for everything, but great for post-hoc forensics — jessi_cata · 2026-10-16
- Neuron labels are like code comments: debate on interpretability vs CoT — jessi_cata · 2026-10-16
- Jensen Huang: UK Must Fix Energy Prices to Seize Its AI 'Goldilocks Moment' — KeeganMcB · 2026-10-16
- SF Vibes Are Electric as Engineers Move From Automation Anxiety to Awe — zack_overflow · 2026-10-16
- Neel Nanda: interpretability is useful but nowhere near reliable enough to ensure AI safety — aryaman2020 · 2026-10-16