大型科技、軍工複合體和聯邦政府的危險融合

最近有消息稱，OpenAI 執行長 Sam Altman 向拜登政府提出大規模建造人工智慧 (AI) 資料中心的計劃，這應該會引起人們的注意。

Recent news that OpenAI CEO Sam Altman pitched the Biden Administration on a massive build-out of artificial intelligence (AI) data centers should raise the alarm on a dangerous convergence: the alliance of Big Tech with the Military-Industrial Complex and the federal government. The move, reported in Zero Hedge, outlines Altman’s proposal for a network of advanced AI centers ostensibly to bolster American technological leadership. But this proposed marriage of Silicon Valley’s emerging AI technologies with federal power—and particularly with the Pentagon and intelligence agencies—poses critical threats to individual privacy and freedom of thought.

最近有消息稱，OpenAI 執行長薩姆·奧爾特曼(Sam Altman) 向拜登政府提出大規模建設人工智慧(AI) 資料中心的建議，這應該對危險的融合敲響警鐘：大型科技公司與軍工複合體和聯邦政府結盟。 《零對沖》報導的這項舉措概述了奧特曼關於建立先進人工智慧中心網路的提議，表面上是為了增強美國的技術領先地位。但矽谷新興人工智慧技術與聯邦權力（尤其是五角大廈和情報機構）的聯姻對個人隱私和思想自由構成了嚴重威脅。

For years, we’ve witnessed a steady erosion of privacy at the hands of both corporate giants and government institutions. From Facebook’s—now Meta (NASDAQ: META)’’s—data scandals to intelligence agency overreach through programs like PRISM or Project Mockingbird.

多年來，我們目睹了企業巨頭和政府機構對隱私的不斷侵蝕。從 Facebook（現為 Meta（納斯達克股票代碼：META））的數據醜聞到情報機構透過 PRISM 或 Project Mockingbird 等計畫進行的越權行為。

Just imagine what goes on at the Central Intelligence Agency’s (CIA) venture capital arm at “In-Q-Tel” designed to invest in start-ups with strategic goals that align with the Agency’s. The combined power of federal authorities and tech behemoths has continually demonstrated a desire to sidestep individual rights in pursuit of their own interests.

想像一下中央情報局 (CIA) 的風險投資部門「In-Q-Tel」正在發生什麼，該部門旨在投資與該機構戰略目標一致的新創公司。聯邦當局和科技巨頭的聯合力量不斷表現出為了追求自身利益而迴避個人權利的願望。

In a similar vein to classic Big Tech consolidation, the Crypto Open Patent Alliance (COPA), which includes Silicon Valley giants like Block, Meta and (Sam Altman’s) WorldCoin, has ambitions in the blockchain, fintech and AI spaces with lots of overlap in sectors of particular interest to the surveillance state.

與經典的大型科技整合類似，加密貨幣開放專利聯盟（COPA）包括Block、Meta 和（薩姆·奧爾特曼的）WorldCoin 等矽谷巨頭，它們在區塊鏈、金融科技和人工智慧領域有著雄心勃勃的野心，這些領域有很多重疊之處。

When we talk about the convergence of tech giants and federal power, we are not talking about speculative risk. There is a long history of partnerships that have enabled overreach, surveillance, and censorship. Amazon’s (NASDAQ: AMZN) AWS hosting CIA databases, Palantir’s data analysis contracts for Immigration and Customs Enforcement (ICE), and Google’s (NASDAQ: GOOGL) scrapped Project Maven AI contract with the Pentagon are just a few examples of the increasingly blurry line between the public and private sectors.

當我們談論科技巨頭和聯邦權力的融合時，我們並不是談論投機風險。長期以來，合作夥伴關係導致了過度擴張、監視和審查。亞馬遜（納斯達克股票代碼：AMZN）託管CIA 資料庫的AWS、Palantir 為移民和海關執法局（ICE）提供的數據分析合約以及谷歌（納斯達克股票代碼：GOOGL）與五角大樓取消的Project Maven AI 合約只是其中幾個例子，這表明兩者之間的界線日益模糊。

As these companies further align their technologies and policies with those of federal authorities and intelligence agencies, the combination of corporate influence and government power could result in an environment where free expression, data privacy and open access to information become luxuries of a bygone era.

隨著這些公司進一步將其技術和政策與聯邦當局和情報機構的技術和政策保持一致，企業影響力和政府權力的結合可能會導致言論自由、資料隱私和資訊開放取用成為過去時代的奢侈品。

AI has the capacity to democratize information and revolutionize productivity, yes. But it can also become a weapon for censorship and manipulation on an unprecedented scale. The curation of data used to train AI models becomes a choke point through which the flow of information can be filtered, edited and outright deleted before it even reaches public consciousness. As a result, powerful actors—whether corporate or governmental—can use this convergence to effectively rewrite social and historical narratives without public debate. It’s a digital “Ministry of Truth,” capable of silencing dissent or inconvenient ideas simply by controlling what AI deems “true.”

是的，人工智慧有能力使資訊民主化並徹底改變生產力。但它也可能成為前所未有的審查和操縱武器。用於訓練人工智慧模型的資料管理成為一個阻塞點，資訊流在到達公眾意識之前就可以透過該阻塞點進行過濾、編輯和徹底刪除。因此，強大的參與者——無論是企業還是政府——可以利用這種融合來有效地改寫社會和歷史敘事，而無需公開辯論。它是一個數位“真理部”，能夠透過控制人工智慧認為“真實”的內容來壓制異議或不方便的想法。

“OpenAI is actively working to strengthen AI infrastructure in the US, which we believe is critical to keeping America at the forefront of global innovation, boosting reindustrialization across the country, and making AI’s benefits accessible to everyone,” OpenAI recently told Bloomberg.

OpenAI 最近告訴彭博社：「OpenAI 正在積極努力加強美國的人工智慧基礎設施，我們認為這對於讓美國保持全球創新的前沿、促進全國範圍內的再工業化以及讓每個人都能享受到人工智慧的好處至關重要。

However, consider the implications of Altman’s proposal. The deployment of government-funded AI data centers is a step forward in technological competitiveness, and I’m not debating that, but it also effectively hands the federal government a new apparatus to shape the development and use of AI technologies. By financing and guiding these technologies, the government gains a say in how they evolve, which means the frameworks for privacy, ethical standards and acceptable uses of AI are likely to align closely with federal interests.

然而，請考慮一下奧特曼所提議的含義。政府資助的人工智慧資料中心的部署是技術競爭力的進步，我對此並不爭論，但它也有效地為聯邦政府提供了一個新的機構來塑造人工智慧技術的開發和使用。透過資助和指導這些技術，政府在它們如何發展方面獲得了發言權，這意味著人工智慧的隱私、道德標準和可接受的用途的框架可能與聯邦利益密切相關。

When AI systems become the primary tools for information dissemination, news aggregation and even research, the question of who controls the curation process becomes a matter of extreme importance. These are tools that could make fact-checking impossible, enabling narratives to be cemented without recourse to alternative viewpoints. Governments and Big Tech have already shown their propensity to censor inconvenient ideas—whether through de-platforming, content moderation, or subtle tweaks to search algorithms. We have seen Mark Zuckerberg and Jack Dorsey admit, for example, that they allowed government agencies to censor criticisms of how the pandemic was handled and also suppressed the truth about Hunter Biden’s laptop!

當人工智慧系統成為資訊傳播、新聞聚合甚至研究的主要工具時，誰控制策展過程的問題就變得極為重要。這些工具可能使事實查核變得不可能，使敘述能夠在不訴諸其他觀點的情況下得到鞏固。政府和大型科技公司已經表現出了審查不便想法的傾向——無論是通過去平台化、內容審核還是對搜尋演算法的微妙調整。例如，我們看到馬克·祖克柏和傑克·多爾西承認，他們允許政府機構審查對流行病處理方式的批評，並壓制有關亨特·拜登筆記型電腦的真相！

The result of Altman’s proposal becoming reality is much worse because it will enter into an atmosphere ripe for the perpetuation of disinformation campaigns that average people will soon be completely unable to decipher on their own.

奧特曼的提議成為現實的結果要糟糕得多，因為它將進入一種適合長期進行虛假信息宣傳的氛圍，而普通人很快將完全無法自己破譯。

Why, you ask?

你問為什麼？

Because we are entering an era where AI systems could be leveraged to automate censorship. Imagine an AI that flags or removes “misinformation” in real-time, using criteria set by corporate interests and government agencies. What counts as “harmful” or “false” becomes a question of political ideology rather than objective truth. And when these decisions are embedded in AI models—especially ones that become more sophisticated and integrated into our lives—it becomes nearly impossible to challenge them because we won’t even be able to research to find the truth. The speed at which AI moves and adapts

因為我們正在進入一個可以利用人工智慧系統實現自動化審查的時代。想像一下，人工智慧可以使用企業利益和政府機構設定的標準即時標記或刪除「錯誤訊息」。什麼算是「有害」或「錯誤」成為政治意識形態問題，而不是客觀真理問題。當這些決策嵌入到人工智慧模型中時——尤其是那些變得更加複雜並融入我們生活的模型——挑戰它們幾乎是不可能的，因為我們甚至無法透過研究找到真相。人工智慧移動和適應的速度