Meta Changes Its AI Plans
Meta Changes Its AI Plans
Meta 改變其 AI 計劃
Introduction
Meta is changing how it makes AI. It is also selling a company from China.
Meta 正在改變其開發 AI 的方式，同時也將出售一家中國公司。
Main Body
Meta now makes its own AI tools. It spent a lot of money on Scale AI. It made a new model called Muse Spark. Now, Meta keeps its AI secret. Some developers are unhappy because they cannot see how the AI works.
Meta 現在開發自己的 AI 工具。它在 Scale AI 上投入了大量資金。它開發了一個名為 Muse Spark 的新模型。現在，Meta 將其 AI 祕密化。部分開發者對此感到不滿，因為他們無法得知該 AI 的運作方式。
Meta needs more money. Most of its money comes from ads. The company is losing money on virtual reality. In May, Meta fired 8,000 workers. Some leaders in the company are fighting.
Meta 需要更多資金。其大部分收入來自廣告。該公司在虛擬實境（VR）領域遭受虧損。五月份，Meta 解雇了 8,000 名員工。公司內部部分領導層正處於權力鬥爭中。
Meta is selling a company called Manus. Manus cost 2 billion dollars. The government in China said this was a security risk. Now, Meta must stop using Manus tools.
Meta 正在出售一家名為 Manus 的公司。Manus 的價值為 20 億美元。中國政府表示這存在安全風險。因此，Meta 現在必須停止使用 Manus 的工具。
Conclusion
Meta wants to make money from AI. But it has many problems with laws and trust.
Meta 希望透過 AI 獲利，但在法律與信任方面面臨許多問題。
Vocabulary Learning
💡 The 'Something + Verb' Pattern
Look at how the text describes actions. It uses a simple pattern: Person/Company Action.
- Meta changes
- Meta needs
- Meta fired
The Rule: When talking about a company or a person in the present, we often add an -s to the action word.
Examples from the text:
- Meta makes (Not: Meta make)
- Meta wants (Not: Meta want)
- It spent (This is for the past!)
💰 Money Words
To reach A2, you need to describe business simply. Use these three words from the article:
- Selling Giving something to get money.
- Losing When money goes away.
- Cost The price of something.
Quick View:
- Selling a company
- Losing money
Vocabulary Learning
Meta Platforms Inc. Changes AI Strategy and Sells Chinese Asset
Meta Platforms Inc. 改變 AI 策略並出售中國資產
Introduction
Meta Platforms Inc. is currently changing its approach to artificial intelligence. The company is moving toward developing its own private models and is being forced to sell a company it acquired in China.
Meta Platforms Inc. 目前正在改變其對人工智慧的處理方式。該公司正轉向開發自有私有模型，並被迫出售一家在中國收購的公司。
Main Body
Meta decided to move toward private AI development because the Llama 4 release was not successful enough. To support this change, Meta invested $14.3 billion in Scale AI and created the Meta Superintelligence Labs. In April, the company released the Muse Spark model, which is designed to work specifically with Meta's own apps and hardware, such as the Ray-Ban Meta glasses. However, this shift has caused problems with developers, who feel that Meta is becoming less transparent and open than it was in the past.
Meta 決定轉向私有 AI 開發，是因為 Llama 4 的發布成效不足。為了支持這項改變，Meta 向 Scale AI 投資了 143 億美元，並成立了 Meta 超級智能實驗室（Meta Superintelligence Labs）。四月份，該公司發布了 Muse Spark 模型，旨在專門與 Meta 自家的應用程式和硬體（如 Ray-Ban Meta 眼鏡）配合運作。然而，這一轉變引起了開發者的不滿，他們認為 Meta 變得不像過去那樣透明且開放。
At the same time, Meta is facing financial challenges. The company relies on advertising for 98% of its income and needs to find new ways to make money. Although revenue grew by 33% in the first quarter, the company's total market value dropped by 18% over the last year. Furthermore, internal stability is an issue, as Meta laid off about 8,000 employees in May and reports suggest tension between old managers and new AI experts. Investors are also worried because the company has lost over $80 billion on virtual reality projects.
與此同時，Meta 正面臨財務挑戰。該公司 98% 的收入依賴廣告，因此需要尋找新的獲利方式。儘管第一季營收增長了 33%，但公司總市值在過去一年下降了 18%。此外，內部穩定也是一個問題，Meta 在五月份裁員約 8,000 人，且有報導指出舊經理與新 AI 專家之間存在緊張關係。投資者也感到擔憂，因為該公司在虛擬實境項目上損失超過 800 億美元。
Finally, Meta has started the process of separating from Manus, a company it bought for $2 billion. This decision follows an order from the Chinese government, which cited national security concerns and rules about exporting technology. As a result, Meta must stop sharing data and remove Manus tools from its systems. This situation shows how the Chinese government is increasing its control over AI companies and foreign investments.
最後，Meta 已開始與其以 20 億美元收購的公司 Manus 進行分拆程序。此決定是根據中國政府的命令，理由是國家安全考量及技術出口規定。因此，Meta 必須停止共享數據，並將 Manus 的工具從其系統中移除。這一情況顯示了中國政府如何加強對 AI 公司及外國投資的控制。
Conclusion
Meta is now in a difficult position as it tries to make money from private AI while following government rules and trying to win back the trust of developers.
Meta 目前處於困難的境地，因為它在嘗試從私有 AI 獲利的同時，還必須遵守政府規定，並試圖贏回開發者的信任。
Vocabulary Learning
⚡ The 'Nuance' Leap: Moving from Basic to Professional
An A2 learner says: "Meta has problems." A B2 learner says: "Meta is facing financial challenges."
Notice the difference? B2 fluency isn't about knowing 'bigger' words; it's about using Collocations (words that naturally live together).
🔍 The Power Duo: Verb + Noun
In the text, we see patterns that make English sound 'expensive' and professional. Instead of using simple verbs like have or get, try these:
- Facing challenges / issues (Not "having problems")
- Citing concerns / rules (Not "saying why")
- Winning back trust / confidence (Not "getting trust again")
- Relying on income / support (Not "using for money")
🛠️ The "Connective Tissue"
To stop sounding like a list of sentences, B2 students use Transitions to show the relationship between ideas. Look at how the article moves from one thought to another:
- Adding more bad news: "Furthermore..."
- Showing a contradiction: "Although [X] happened, [Y] occurred..."
- Showing a result: "As a result..."
Pro Tip: If you use 'And' or 'But' at the start of every sentence, you are staying at A2. Swap 'And' for 'Furthermore' and 'But' for 'However', and you instantly bridge the gap to B2.
💡 Quick Upgrade Guide
| A2 Level (Simple) | B2 Level (Professional) | Context from Text |
|---|---|---|
| Big change | Shift | "...this shift has caused problems" |
| Give a reason | Cite | "...which cited national security concerns" |
| Split/Break | Separating from | "...process of separating from Manus" |
Vocabulary Learning
Meta Platforms Inc. Strategic Realignment of Artificial Intelligence Assets and Regulatory Divestiture
Meta Platforms Inc. 人工智能資產策略調整與監管剝離
Introduction
Meta Platforms Inc. is currently undergoing a structural transition in its artificial intelligence strategy, characterized by a shift toward proprietary model development and the forced divestiture of a Chinese-linked acquisition.
Meta Platforms Inc. 目前正經歷人工智能策略的結構性轉型，其特點是轉向專有模型的開發，以及被強制剝離一項與中國相關的收購項目。
Main Body
The institutional pivot toward proprietary AI development was precipitated by the perceived inadequacy of the Llama 4 release and a subsequent strategic departure from a strictly open-weight framework. This transition was facilitated by a $14.3 billion investment in Scale AI, resulting in the integration of Alexandr Wang and the establishment of Meta Superintelligence Labs. The subsequent deployment of the Muse Spark model in April represents a transition toward a 'walled-garden' ecosystem, designed for seamless integration into Meta's internal application suite and hardware, such as the Ray-Ban Meta glasses. However, this shift has engendered friction with the developer community, who perceive a decline in transparency and accessibility compared to previous open-source initiatives.
機構向專有 AI 開發的轉向是由於 Llama 4 的發佈被認為不足，以及隨後在策略上脫離純粹的開放權重框架。這次轉型得益於對 Scale AI 143 億美元的投資，促成了 Alexandr Wang 的加入以及 Meta 超級智能實驗室 (Meta Superintelligence Labs) 的成立。隨後於 4 月部署的 Muse Spark 模型代表了向「圍牆花園」生態系統的過渡，旨在無縫整合至 Meta 的內部應用軟體套件與硬體（如 Ray-Ban Meta 眼鏡）中。然而，這一轉向引起了開發者社群的摩擦，他們認為與之前的開源計畫相比，透明度與可近接性有所下降。
Financial and operational pressures persist, as the organization seeks to diversify revenue streams beyond its current 98% reliance on advertising. Despite a 33% increase in first-quarter revenue, the company's market valuation has declined by 18% over the preceding twelve months. This fiscal volatility is compounded by internal instability, including the termination of approximately 8,000 employees in May and reported tensions between established leadership and new AI recruits. Furthermore, the company's historical losses exceeding $80 billion in virtual reality ventures have diminished investor confidence in executive credibility.
財務與營運壓力持續存在，因為組織尋求將收入來源多元化，以擺脫目前 98% 對廣告的依賴。儘管第一季營收增長 33%，但公司市值在過去 12 個月內下降了 18%。這種財政波動因內部不穩定而加劇，包括 5 月份解僱約 8,000 名員工，以及據報的既有領導層與新 AI 招募人才之間的緊張關係。此外，公司在虛擬實境 (VR) venture 中超過 800 億美元的歷史虧損，削弱了投資者對執行層信譽的信心。
Concurrent with these internal shifts, Meta has commenced the operational separation of Manus, a $2 billion acquisition. This divestiture is a direct response to a mandate from Beijing citing national security concerns and violations of technology export controls. The process involves the cessation of data sharing and the removal of Manus tools from internal systems. This action reflects a broader geopolitical trend wherein the Chinese government is intensifying oversight of AI firms, implementing travel restrictions for executives and requiring state approval for foreign capital infusions.
與這些內部轉變同步，Meta 已開始對收購價 20 億美元的 Manus 進行營運分離。此次剝離是直接回應北京方面以國家安全為由，指其違反技術出口管制的指令。該過程涉及停止數據共享，並將 Manus 工具從內部系統中移除。此舉反映了更廣泛的地緣政治趨勢，即中國政府正強化對 AI 企業的監管，對高管實施旅遊限制，並要求外資注入需經國家批准。
Conclusion
Meta remains in a precarious position, attempting to monetize proprietary AI while navigating regulatory mandates and restoring developer trust.
Meta 仍處於不穩定的位置，試圖將專有 AI 貨幣化，同時需應對監管指令並恢復開發者的信任。
Vocabulary Learning
The Architecture of Nominalization and Lexical Density
To bridge the gap from B2 to C2, one must move beyond describing actions and begin conceptualizing states. The provided text is a masterclass in Nominalization—the process of turning verbs (actions) and adjectives (qualities) into nouns. This is the hallmark of high-level academic and corporate discourse, as it allows for a higher density of information and a more objective, detached tone.
◈ Deconstructing the 'Action-to-Concept' Pivot
Observe how the author avoids simple subject-verb-object constructions in favor of complex noun phrases. Compare these two versions of the same fact:
- B2 Level: Meta shifted its strategy because the Llama 4 release was not good enough. (Simple, narrative, focused on the actor)
- C2 Level: "The institutional pivot toward proprietary AI development was precipitated by the perceived inadequacy of the Llama 4 release..."
Analysis of the C2 Shift:
- The Pivot (Noun): Instead of saying "Meta pivoted" (verb), the author uses "The institutional pivot" (noun phrase). This transforms the action into an entity that can be analyzed.
- Precipitated (High-Level Verb): Instead of "caused," the author uses a word that suggests a chemical or sudden reaction, adding a layer of precision.
- Perceived Inadequacy (Abstract Noun): Instead of "was not good enough," the author uses a nominalized phrase. This shifts the focus from the fact of the failure to the perception of it, which is a crucial nuance in strategic reporting.
◈ Syntactic Patterns for Mastery
To replicate this, focus on the [Noun] + [Prepositional Phrase] + [Passive Verb] structure:
"This fiscal volatility [Noun] is compounded by [Passive Verb] internal instability [Noun]..."
Why this works at C2: It removes the need for a human agent (e.g., "Investors are worried about the volatility"), creating an aura of institutional authority and systemic inevitability.
◈ Lexical Precision: The 'Corporate-Geopolitical' Register
Note the use of Collocations that anchor the text in a professional register:
- Forced divestiture (Legal/Financial)
- Walled-garden ecosystem (Tech/Strategic)
- Foreign capital infusions (Economic/Political)
- Regulatory mandates (Legal)
C2 Takeaway: Stop searching for 'big words' and start searching for 'precise pairings.' A B2 student says "selling a part of the company"; a C2 student describes a "divestiture."