AITOPAI热报
Industry76°

FT：Meta 拟发行无投票权股份为 AI 建设融资

rohanpaul_ai@rohanpaul_ai
Post· 2 sources
View the post
Why we picked it

Meta 为了给 AI 烧钱又不想丢控制权，打算发无投票权股票，扎克伯格还拿搬家当筹码，这操作挺少见。

据 FT 报道，Meta 的 AI 基础设施扩建需要引入外部股权融资，规模超出其创始人控制结构所能承受的范围。方案是设立新的无投票权股份类别，让投资者获得经济回报但没有投票权，Mark Zuckerberg 出售股份同时保住控制权。作为交换，Zuckerberg 承诺在加州或 Meta 总部所在地居住 8 年，此前他曾向董事会表示可能离开。背景是加州 Proposition 40，一项针对亿万富翁的一次性 5% 税收提案，将于 11 月 3 日交由选民投票。

Original · rohanpaul_ai

FT: Meta’s AI build-out now needs outside equity on a scale that its founder-control structure could not absorb, so the company is creating a share class that gives investors economics with no votes.

i.e. now Meta will issue non-voting shares that let Mark Zuckerberg sell stock while keeping control of the company.

In return, Zuck has committed to live in California, or wherever Meta is headquartered, for 8 years after telling directors he might leave.

The threat was Proposition 40, a one-time 5% tax on California billionaires that goes to voters on November 3.

  • VnExpress Số hóa10-21 02:00原文
#Meta#Mark Zuckerberg#Proposition 40#行业融资

Keep reading

Recent stories with the same tag or section
ModelsBuildersMulti-sourceSelected16:36
Muse Glimmer 30B: Meta's distilled agent model runs on one consumer GPU

Distilling from a large teacher into a small model specialised for long tool-use chains may matter more than raw size for local agents, a trade-off the OSWorld gap makes visible. Builders running agents on local hardware can use it as a drop-in for multi-tool MCP workflows without cloud dependency. Compared with Qwen3.6-27B, it wins on tool orchestration but loses on general OS-level tasks.

#Muse Glimmer#Meta#open-source models#agents
Event
rohanpaul_ai@rohanpaul_ai4 sourcesSource
PapersBuildersSelected13:32
Meta and Michigan State paper finds LLM-built reasoning graphs unreliable

The non-obvious point is that most reasoning-graph pipelines assume the LLM's decomposition is correct without ever verifying it, so confidence scores inherit hidden graph errors. Builders of agent workflows and RAG-style decomposition systems should add graph-level validation before trusting computed confidence. The Qwen3-32B versus Qwen3-8B result is a concrete warning that scaling a model does not guarantee better sub-question answering.

#Qwen3#Meta#reasoning models#AI safety
rohanpaul_ai@rohanpaul_aiSource