微軟和 Meta Shift AI 策略、Nvidia 過熱的晶片和強勁的獲利

微軟（納斯達克股票代碼：MSFT）最近發表了一篇關於其新人工智慧代理的部落格文章，概述了它們是什麼、它們如何運作以及它們的潛在應用。

AI Agents, Microsoft (NASDAQ:MSFT), Meta (NASDAQ:META), Nvidia (NASDAQ:NVDA)

AI Agents、微軟（納斯達克股票代碼：MSFT）、Meta（納斯達克股票代碼：META）、Nvidia（納斯達克股票代碼：NVDA）

Microsoft’s AI Agents: An evolution of Copilot

微軟的人工智慧代理：Copilot 的演變

Last week, Microsoft (NASDAQ:MSFT) unveiled its new AI Agents, generative artificial intelligence tools that can be customized to perform specific tasks. These Agents are designed to leverage data from the Microsoft Office suite to execute multi-step processes autonomously.

上週，微軟（納斯達克股票代碼：MSFT）推出了新的人工智慧代理，這是一種可以自訂以執行特定任務的生成人工智慧工具。這些代理程式旨在利用 Microsoft Office 套件中的資料自主執行多步驟流程。

While Microsoft’s Copilot product already provides many of these capabilities, the Agents expand on them by improving memory, integrating with external systems, and enabling more complex workflows.

雖然 Microsoft 的 Copilot 產品已經提供了其中許多功能，但代理商透過改進記憶體、與外部系統整合以及支援更複雜的工作流程來擴展這些功能。

The shift from Copilot to AI Agents seems to be more than just a rebranding effort. It feels like Microsoft is addressing a gap in how its customers use AI. Many users haven’t fully tapped into Copilot’s potential, and the blog post reads like an early step in a broader re-education initiative.

從 Copilot 到 AI Agents 的轉變似乎不僅僅是品牌重塑工作。微軟似乎正在解決客戶使用人工智慧的方式上的差距。許多用戶尚未充分發揮 Copilot 的潛力，這篇部落格文章讀起來像是更廣泛的再教育計畫的早期步驟。

Microsoft appears to be laying the groundwork for helping users understand how to maximize their AI tools—not just within Office but as part of their daily workflows.

微軟似乎正在為幫助用戶了解如何最大限度地利用人工智慧工具奠定基礎——不僅在 Office 中，而且將其作為日常工作流程的一部分。

At nearly every opportunity I get, I mention how most people use AI as a glorified search engine rather than using it in ways to get the most out of the system. AI tools can have a significant impact on how we work and live, streamline many processes, save resources, and increase our overall efficiency.

幾乎每次有機會，我都會提到大多數人如何將人工智慧用作美化的搜尋引擎，而不是以充分利用系統的方式使用它。人工智慧工具可以對我們的工作和生活方式產生重大影響，簡化許多流程，節省資源，並提高我們的整體效率。

However, it seems like many users just haven’t dug deeper with AI or figured that out yet. From what I see, it does seem like many users need more guidance if they are looking to get the most out of their AI tools of choice. Microsoft’s move could be the start of a broader industry trend where companies invest more heavily in user education, providing step-by-step guides that allow their users to better integrate their AI into their personal and professional lives.

然而，似乎許多用戶還沒有更深入地挖掘人工智慧或弄清楚這一點。從我看來，如果許多用戶希望充分利用他們選擇的人工智慧工具，他們似乎確實需要更多指導。微軟的舉動可能是更廣泛行業趨勢的開始，即公司加大對用戶教育的投資，提供逐步指南，使用戶能夠更好地將人工智慧融入個人和職業生活。

Meta’s push to monetize AI

Meta 推動人工智慧貨幣化

Meta (NASDAQ:META), on the other hand, has appointed former Salesforce (NASDAQ:CRM) CEO Clara Shih as its new “Head of Business AI.”

另一方面，Meta（納斯達克股票代碼：META）已任命 Salesforce（納斯達克股票代碼：CRM）前執行長 Clara Shih 為其新的「商業人工智慧主管」。

According to Shih, “Our vision for this new product group is to make cutting-edge AI accessible to every business, empowering all to find success and own their future in the AI era… Meta’s global reach and leadership in AI represent a generational opportunity for businesses, and I couldn’t be more excited and grateful to help take this from zero to one to scale.”

Shih 表示：「我們對這個新產品組的願景是讓每家企業都能接觸到尖端人工智慧，讓所有人都能在人工智慧時代取得成功並擁有自己的未來……Meta 在人工智慧領域的全球影響力和領導地位代表著一代人的機會。

The move signals Meta’s growing focus on profitability in AI. It’s no secret that many artificial intelligence companies struggle to profit. The infrastructure required to train and run advanced models comes at a massive cost.

此舉標誌著 Meta 越來越關注人工智慧的獲利能力。許多人工智慧公司難以獲利已不是什麼秘密。訓練和運行高級模型所需的基礎設施成本高。

In contrast, the revenue they collect through subscription models that offer customers access to “better” AI tools recoups only a fraction of the companies’ expenses.

相較之下，他們透過訂閱模式為客戶提供「更好」的人工智慧工具而獲得的收入只能收回公司開支的一小部分。

The purpose of Meta’s Business AI group seems to be to turn this around by productizing their AI offerings—packaging them into marketable products that attract paying customers and drive significant revenue growth for the business.

Meta 商業人工智慧團隊的目的似乎是透過將其人工智慧產品產品化來扭轉這一局面，將其打包成適銷對路的產品，吸引付費客戶並推動企業收入大幅成長。

This shift in strategy isn’t unique to Meta; it reflects a broader trend in the AI industry. It has become clear that many companies’ AI divisions are increasingly finding themselves under pressure to figure out monetization as investors start asking when they will see returns.

這種策略的轉變並不是 Meta 所獨有的。它反映了人工智慧產業更廣泛的趨勢。很明顯，隨著投資人開始詢問何時能看到回報，許多公司的人工智慧部門越來越面臨貨幣化的壓力。

The narrative around AI seems to be shifting from “AI is revolutionary” to “Who’s making money from AI, and how soon can we expect returns on our investments?”

關於人工智慧的敘述似乎正在從“人工智慧是革命性的”轉變為“誰從人工智慧中賺錢，我們多久才能期望獲得投資回報？”

It’s an important question because the industry could face a wave of business closures and mergers without a clear path to profitability. If companies can’t find a financially viable strategy for their AI divisions, we may soon see a consolidation of players in the AI space.

這是一個重要的問題，因為如果沒有明確的獲利途徑，該產業可能會面臨一波企業倒閉和合併浪潮。如果公司無法為其人工智慧部門找到財務上可行的策略，我們可能很快就會看到人工智慧領域的參與者進行整合。

Nvidia’s overheating chips and strong earnings

英偉達過熱的晶片和強勁的獲利

A recent report said that Nvidia’s (NASDAQ:NVDA) Blackwell chips are prone to overheating when added to servers, creating significant challenges for data centers. To address the issue, service providers must redesign their racks, an expensive and time-consuming process.

最近的一份報告稱，Nvidia（納斯達克股票代碼：NVDA）的 Blackwell 晶片在添加到伺服器時容易過熱，為資料中心帶來重大挑戰。為了解決這個問題，服務提供者必須重新設計他們的機架，這是一個昂貴且耗時的過程。

Despite that negative press, Nvidia’s Q3 earnings report exceeded analyst expectations. The company posted adjusted earnings per share of $0.81, representing a net income of $19.3 billion, compared to predictions of $0.75 EPS and $17.4 billion net income.

儘管有負面新聞，英偉達第三季財報還是超出了分析師的預期。該公司公佈調整後每股收益為 0.81 美元，淨利潤為 193 億美元，而預期每股收益為 0.75 美元，淨利潤為 174 億美元。

However, even with these substantial numbers, Nvidia’s stock dropped by roughly 2% when the market opened the next day.

然而，即使有這些可觀的數字，Nvidia 的股價在第二天開盤時仍下跌了約 2%。

This trend—where companies beat earnings expectations but experience a decline in share prices—has become increasingly common. This stems from the market’s perception that a company with strong current performance has less room for growth in the upcoming quarters.

這種公司超出獲利預期但股價下跌的趨勢已經變得越來越普遍。這源自於市場認為當前業績強勁的公司在未來幾季的成長空間較小。

In Nvidia’s case, their Q4 guidance of $37.5 billion in revenue, only slightly above Wall Street’s $37 billion projection, reinforces this idea.

就 Nvidia 而言，他們第四季營收指引為 375 億美元，僅略高於華爾街 370 億美元的預測，強化了這個想法。

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system

為了讓人工智慧 (AI) 在法律範圍內正常運作並在面臨日益增長的挑戰時蓬勃發展，它需要整合企業區塊鏈系統