跳到主要內容

AI投資回報評測：Apollo經濟學家說多數企業零回報 | AI ROI Review: Apollo Economist Says Most Firms See Zero Payoff

By Kit 小克 | AI Tool Observer | 2026-11-03

🇹🇼 AI投資回報評測：Apollo經濟學家說多數企業零回報

AI投資回報到底有沒有兌現，最近被華爾街狠狠戳了一刀。Apollo 全球管理公司首席經濟學家 Torsten Slok 在 10 月底發布的分析指出，除了少數科技巨頭之外，多數企業從 AI 投資中得到的財務回報接近於零，這份報告迅速在 Hacker News、Reddit 和財經媒體圈引爆討論，因為它第一次用具體數字，而不是情緒，去質疑這波 AI 投資熱潮。

Apollo怎麼算出「零財務回報」

Slok 把 S&P 500 拆成兩組來比較：一組是「科技七巨頭」（Magnificent Seven，包含 Nvidia、微軟、Meta、Amazon、Alphabet、蘋果、特斯拉），另一組是其餘的 493 家公司（他稱之為 S&P 493）。結果很直白：七巨頭的營收成長和每人產值確實在往上走，但其餘 493 家公司的利潤率和每人營收幾乎是持平，看不出 AI 投資帶來的生產力提升。

為什麼大部分公司賺不到錢

  • Token 支出變成成本中心：Slok 指出，許多公司現在把 AI 模型的 API 用量當成單純的費用項目，而不是成長引擎在管理。
  • 優勢被同業抵銷：如果你的競爭對手也在用同樣的 AI 工具提升效率，省下來的成本很快會被市場競爭打平，無法反映在利潤率上。
  • 只有賣AI的人在賺錢：真正受益的其實是賣 GPU、賣雲端算力、賣模型 API 的那幾家公司，買家端（也就是其他 493 家公司）目前多半還在「投入期」。

這對一般公司和開發者意味著什麼

如果你的公司正在評估要不要加碼 AI 預算，這份報告給了一個務實的提醒：先問「省下的時間有沒有換成利潤」，而不是只問「有沒有在用 AI」。對中小企業和個人開發者來說，反而是個機會——當大公司還在為 AI ROI 找不到答案而猶豫時，用對工具、用在對的場景（像是客服自動化、內容生產、程式碼審查）確實能拉開差距，但前提是要有明確的衡量指標，不是為了跟風而導入。

Goldman Sachs 同期也示警美國經濟的 AI 紅利可能已接近高峰，顯示整個市場開始從「狂熱期」轉向「驗收期」。接下來幾個季度，各家公司的財報會是檢驗這份報告最直接的證據。

好不好用，試了才知道。

🇺🇸 AI ROI Review: Apollo Economist Says Most Firms See Zero Payoff

The debate over AI ROI just got a cold dose of reality. In a late-October analysis, Torsten Slok, chief economist at Apollo Global Management, argued that outside a handful of tech giants, most companies are seeing essentially zero financial payoff from their AI investments. The report quickly spread across Hacker News, Reddit, and financial media, not because it is hype, but because it is one of the first attempts to put hard numbers behind growing skepticism about the AI spending boom.

How Apollo Measured Zero Payoff

Slok split the S&P 500 into two groups: the Magnificent Seven (Nvidia, Microsoft, Meta, Amazon, Alphabet, Apple, Tesla) and the remaining 493 companies, what he calls the S&P 493. The finding is blunt: revenue per worker and profit margins are climbing for the Magnificent Seven, but for the other 493 companies, both metrics are essentially flat, with no visible productivity boost from AI adoption.

Why Most Companies Are Not Seeing Returns

  • Token spend is becoming a cost center: Slok notes many companies now book AI API usage as a straightforward expense line, not a growth investment with tracked returns.
  • Competitive advantages get erased: If your competitors adopt the same AI tools, any efficiency gain gets competed away before it shows up in margins.
  • The sellers win, not the buyers: The real winners are companies selling GPUs, cloud compute, and model APIs. The buyers, the other 493 companies, are mostly still in the spending phase.

What This Means for Businesses and Developers

If your company is deciding whether to expand its AI budget, this report offers a practical gut-check: ask whether time saved is actually converting into profit, not just whether you are using AI. For smaller businesses and independent developers, this is actually an opening: while large companies struggle to prove AI ROI, applying the right tool to the right use case, like customer support automation, content production, or code review, can create a real edge, as long as you track concrete metrics instead of adopting AI just to follow the trend.

Goldman Sachs issued a similar warning around the same time, suggesting the U.S. economy's AI-driven boost may be nearing its peak. The market appears to be shifting from a hype phase into a verification phase, and the next few quarters of corporate earnings will be the real test of whether Apollo's numbers hold up.

You won't know until you try it.

Sources / 資料來源

延伸閱讀 / Related Articles

AI 工具觀察站 — 每日精選 AI Agent 與工具趨勢
AI Tool Observer — Daily curated AI Agent & tool trends

留言

張貼留言