Home
Tech
The tech industry is divi...

The tech industry is divided over calls to slow down AI coordination.

C2🇹🇼 中文 → 🇺🇸 English

September 17th, 2026

The tech industry is divided over calls to slow down AI coordination.

Please note: This article has been simplified for language learning purposes. Some context and nuance from the original text may have been modified or removed.

🇺🇸 English

Anxieties concerning artificial intelligence spiralling out of control, together with appeals urging a slowing of that technology's development, are of long standing; yet, as warnings that runaway AI systems are fomenting havoc grow ever more clamorous, such anxieties and appeals have of late regained prominence.

對於人工智慧失控之憂慮，暨籲請放緩該技術發展之建言，由來已久，然伴隨失控AI系統釀致禍亂之警訊甚囂塵上，斯等憂慮與建言邇來復獲矚目。

Senior executives at several leading artificial intelligence companies have recently spoken with one voice, asserting that coordinated mitigation measures—including international cooperation—are genuinely necessary to safeguard humanity from the most dire scenarios.

數位領先的人工智慧企業高階主管近日異口同聲地指出，實有必要推動協調性的減緩措施，其中涵蓋國際合作，以確保人類免於最嚴峻情境之危害。

Yet opinions diverge sharply, and some technology leaders take profound exception to the very notion that their own companies could not, by their own efforts alone, guarantee safety.

然而，各方見解歧異，部分科技界領袖對於下述構想深感不以為然：其麾下企業竟無法憑一己之力確保安全。

The following is a distillation of what several key participants had to say.

以下乃是對若干主要參與者所言內容之梳理。

Dario Amodei, the CEO of Anthropic, has unveiled by far the most detailed blueprint to date, advocating a slowdown in the pace of development and calling for coordination among different companies and nations.

Anthropic執行長達里奧·阿莫德伊（Dario Amodei）揭櫫了迄今最為詳盡的藍圖，力主放緩發展步伐，並倡議推動不同企業與國家之間的協調。

Ever since its founders departed from OpenAI in 2021 to establish this startup, Anthropic has long positioned itself as the more responsible and safety-conscious member among the leading AI companies.

自2021年其創辦人離開OpenAI、另立這家新創公司以來，Anthropic便長期將自身定位為領先AI企業中更負責任、更重視安全的一家。

In a recent paper outlining its mitigation strategy, one of the most widely noted proposals is a call for companies at the AI frontier — commonly known as frontier labs — to commit to granting independent external evaluators "persistent, employee-like access."

在近期一篇闡述其減緩策略之論文中，最廣獲矚目的提案之一，乃是籲請位居AI前沿之企業——通稱前沿實驗室——允諾向獨立外部評估人員授予「持續性、類同員工之存取權限」。

This evaluation team would be furnished with offices, access cards, and company laptops in order to audit safety practices.

該評估團隊將配備辦公室、門禁卡及公司筆記型電腦，藉以稽核安全實務。

Anthropic has already made this commitment, and so has OpenAI.

Anthropic已作出此項承諾，OpenAI亦然。

Amodei likewise calls upon governments to exercise regulatory oversight and intervention, including by facilitating coordination among frontier AI companies with the assistance of the United States government and those of other democratic nations.

阿莫德伊亦籲請政府予以監管與干預，其途徑包括在美國政府及其他民主國家政府之襄助下，促成前沿AI企業彼此之間之協調。

His plan further encompasses initiatives by the United States and other nations aimed at coordinating with "authoritarian governments."

他的計畫尚涵蓋美國及其他國家致力與「威權政府」協調之舉措。

He candidly acknowledged that such an undertaking is fraught with formidable obstacles, and he singled out the manifold challenges entailed in securing China's cooperation.

他坦承此舉窒礙難行，並特別指出爭取中國合作所面臨的重重挑戰。

Ahmed stated that he was fully aware that putting his proposal into practice would be no easy feat, yet he firmly believed that "we have a duty to humanity to try."

阿莫德伊表示，他深知付諸實行其提案絕非易事，但他堅信「我們有責任為人類一試」。

Sam Altman, the CEO of OpenAI, has already declared his support for establishing a measured pace for development, and has vowed to move toward broader collaboration across the entire industry.

OpenAI執行長山姆·奧特曼（Sam Altman）業已表態，支持為發展確立節奏，並矢言邁向全行業更廣泛之協作。

Several days before Amodei's article went to press, OpenAI published a lengthy piece outlining how the company is vigorously promoting mandatory national security requirements and continuing to support state-level legislation aimed at "strengthening the broader AI safety ecosystem."

在阿莫德伊文章付梓前數日，OpenAI刊布一篇鴻文，概述該公司如何力推強制性國家安全要求，並持續襄助「強化更廣泛AI安全生態系統」之州層級立法。

The company also declared its support for collaboration with other frontier AI laboratories—with or without government backing—and its commitment to establishing global standards.

該公司亦表明支持與其他前沿AI實驗室協作——無論有無政府奧援——並戮力建立全球標準。

In the piece, OpenAI's global affairs chief, Chris Lehane, made plain: "We will champion mutually compatible international practices by which to measure capabilities, manage risk, preserve human primacy, and determine when and how to slow or halt development — even if doing so means decelerating the advance of model capabilities, at no matter the cost."

OpenAI全球事務長克里斯·勒漢（Chris Lehane）於文中闡明：「吾等將倡導彼此相容之國際實踐，藉以衡量能力、管控風險、維繫人類之主導權，並裁定何時及如何減緩抑或終止發展——縱使此舉意味著須放緩模型能力之推進，亦在所不惜。」

Lehane clarified that the blueprint for an ideal federal framework outlined by OpenAI encompasses "norms for joint testing and independent evaluation, more rigorous cybersecurity safeguards, clear incident-reporting standards, a higher degree of national preparedness, and shared measurement mechanisms for tracking processes of recursive self-improvement"; and what is meant by recursive self-improvement is the capacity of AI systems to autonomously refine their own capabilities.

勒漢闡明，OpenAI所勾勒之理想聯邦框架藍圖，涵蓋「共同測試與獨立評估之規範、更臻縝密之網路安全防護、明晰之事件通報準則、更高程度之國家整備態勢，以及用以追蹤遞迴自我改進進程之共同衡量機制」；而所謂遞迴自我改進，係指AI系統得以自主精進自身之能力。

In recent remarks, Altman made it unequivocally clear that "setting the pace" for AI development does not mean grinding to an abrupt halt.

奧特曼於近期評論中明確闡明，AI發展之「設定節奏」並不意味著戛然而止。

Altman stated on X that AI progress will continue apace.

奧特曼在X上表示，AI進展仍將持續迅猛。

"But it ought to be slower than it might otherwise have been; interventions such as safety cases and monitoring all come with significant costs."

「但它理應較原本可能之速度更為放緩；安全案例與監控等干預措施皆伴隨顯著成本。」

He said, "The costs incurred in setting the pace will prove extremely worthwhile."

他說道：「設定節奏所耗費之成本，將極具價值。」

He then added, "However great the competitive pressure the United States may exert, it should not constitute a justification for acting recklessly."

他繼而補充道，「即便美國施加再大之競爭壓力，亦不應構成魯莽行事之正當理由。」

Mark Zuckerberg, the chief executive of Meta, has taken a stance against the notion of coordinated deceleration.

Meta執行長馬克·祖克柏（Mark Zuckerberg）則對協調減緩之構想持反對立場。

Zuckerberg's vision for the future of artificial intelligence is markedly more optimistic than that of several of his peers.

祖克柏對人工智慧前景所抱持的願景，較諸若干同業更顯樂觀。

He contends that every AI company should take it upon itself to guarantee the safety of its own technology, and should distance itself from the mitigation discourse that Meta has embraced.

他主張，每一家人工智慧企業皆應自行確保其技術之安全性，並將自身與Meta同減緩論述劃清界線。

Zuckerberg asserted that AI companies already possess the incentive to develop their models safely and bear a "significant responsibility" to prevent them from causing harm.

祖克柏聲稱，人工智慧企業已然具備以安全方式開發其模型之誘因，並承擔著防止其造成危害的「重大責任」。

He wrote on X: "Every lab has both the responsibility and the incentive to advance at the pace required to train its models safely, and it is also capable of taking action on its own to ensure this."

他在X上寫道：「每個實驗室皆有責任與誘因，以安全訓練其模型所需之速度推進，亦有能力自行採取行動以確保這一點。」

Jensen Huang, the CEO of Nvidia, likewise took considerable issue with calls to slow the pace.

輝達（Nvidia）執行長黃仁勳亦對放緩步伐之呼聲多所針砭。

He maintained that each company ought to calibrate its own tempo, and to do so only until it is confident that the products it releases will ultimately find favour with the market.

他主張，各企業理當自行斟酌其節奏，直至確信所發布之產品終將獲市場青睞而後已。

Huang Renxun pointed out: "Market forces already exist—we need no new laws, nor any new regulations."

黃仁勳指出：「市場力量已然存在——我們無需任何新法律，亦無需新法規。」

He argued that innovation and safety are not mutually exclusive.

他主張創新與安全並非互斥。

"The two can absolutely be achieved together."

「兩者絕對可以兼得。」

Elon Musk is sympathetic to the case for slowing things down, and he had repeatedly voiced similar concerns about AI safety well before the recent escalation of this debate.

伊隆·馬斯克（Elon Musk）對減緩之議持支持立場，且早在近期這波爭論升溫之前，便已屢屢表達類似的AI安全隱憂。

Musk is the founder of xAI as well as numerous other technology enterprises.

馬斯克身兼xAI及眾多科技企業之創辦人。

On X, he reposted Amodei's article, appending the remark: "Dario is right."

他在X上轉發了阿莫德伊的文章，並加註：「達里奧是對的。」

He subsequently wrote online that he "has long been sounding the alarm about AI."

他其後於網上寫道，他「長期以來一直在對AI發出警報」。

In an interview with The Economist this summer, Musk proposed that the leadership of the major AI laboratories ought to convene to deliberate on matters of safety and security, and that competitors should test one another's models so as to expose areas of concern.

在今夏接受《經濟學人》採訪時，馬斯克建議主要AI實驗室的領導層應當會晤商討安全與保障事宜，且競爭對手應相互測試彼此的模型，以揭示令人憂慮的領域。

Musk stated: "I think that for those who neither possess an in-depth grasp of the technical intricacies within government nor are at the forefront of AI development, determining whether something ought to be released is exceedingly difficult.

馬斯克表示：「我認為，對於那些既未深入掌握政府內部技術細節、亦非引領AI前沿發展的人而言，要斷定某項事物是否理應發布，實屬相當困難。

However, if competitors were afforded a week or two to scrutinise a new model, that would suffice to bring problems to light and create an incentive for other companies to remain honest."

然而，倘若競爭對手能有一兩週時間審查一個新模型，便足以凸顯問題，並促使其他公司產生保持誠實的誘因。」

Demis Hassabis, Chairman and co-founder of Google DeepMind, likewise concurred with Amodei's view, remarking that "the details need to be clarified, but the direction is right and suited to addressing this pivotal moment."

Google DeepMind董事長兼共同創辦人戴米斯·哈薩比斯（Demis Hassabis）亦對阿莫德伊之見表示附議，稱「細節需要釐清，但方向正確，適合應對這個關鍵時刻」。

Earlier this year, Hassabis put forward his own framework, advocating the establishment of a frontier AI standards body.

哈薩比斯於本年提出其個人框架，主張建立一個前沿AI標準機構。

This standards body would be charged with devising evaluation protocols and conducting tests in areas related to national security.

該標準機構將肩負制定評估協議之責，並於國家安全相關領域執行測試。

He wrote that the body would be modelled on organisations such as the US Financial Industry Regulatory Authority, and would in all likelihood have to be industry-funded in order to "attract world-class technical talent and provide the necessary compute resources for large-scale testing".

他寫道，該機構將以美國金融業監管局（Financial Industry Regulatory Authority）等組織為範本，且極有可能須由產業資助，以「吸引世界級技術人才，並為大規模測試提供必要的算力資源」。

Microsoft Chairman and CEO Satya Nadella and Microsoft AI CEO Mustafa Suleyman both expressed profound concerns regarding the safety and development of AI.

微軟董事長兼執行長薩帝亞·納德拉（Satya Nadella）與微軟AI執行長穆斯塔法·蘇萊曼（Mustafa Suleyman）均就AI之安全與發展表達了深切憂慮。

Nadella stated on social media: "If the AI we build does not benefit humanity and remains beyond human control, then it is not worth pursuing."

納德拉於社群媒體上表示：「倘若我們所打造之AI無助於人類，且不受人類所控，那便不值得追求。」

He clarified that Microsoft welcomes "the research, focus, and deliberately set cadence required to get alignment right as a design goal."

他闡明，微軟樂見「作為設計目標，為正確實現對齊所需的研究、專注與刻意設定的節奏」。

Alignment refers to measures that ensure a model's actions are consistent with human values and intentions.

所謂對齊，係指確保模型行動契合人類價值與意圖之舉措。

Nadella noted that the company likewise favours the idea of embedding evaluators to monitor safety work.

納德拉指出，該公司同樣青睞設置嵌入式評估人員以監控安全工作之構想。

However, he opposes the notion of frontier labs unilaterally leading this work on their own.

然而，他反對由前沿實驗室獨力主導此項工作之構思。

Nadella expounded on social media: "The crux of the matter is that this must not be monopolized by a handful of entities; rather, it must encompass broad representation across the entire ecosystem, nations, and domains, with academia duly included."

納德拉於社群媒體上闡述道：「箇中關鍵在於，此事斷不容由少數實體所壟斷，而務須於整個生態系統、國家及領域之中具備廣泛之代表性，學術界亦當涵括在內。」

Suleyman co-founded DeepMind with Hassabis.

蘇萊曼曾與哈薩比斯共同創辦DeepMind。

In an article, he made clear that although he shares the common goal of developing advanced AI safely, he is deeply troubled by the anthropomorphisation of models, and of Anthropic's Claude in particular; he stated bluntly that this could amplify AI safety risks.

他在一篇文章中闡明，儘管他認同安全開發先進AI這一共同目標，卻對模型遭擬人化一事深感憂慮，尤以Anthropic的Claude為甚；他直言，此舉恐將放大AI安全風險。

Suleyman writes: "Steering a being whose capabilities and intellect surpass those of all humanity combined is, in itself, an exceedingly formidable challenge — one that far exceeds anything we have ever had to contend with before.

蘇萊曼寫道：「駕馭一個能力與智識俱凌駕於全人類之上的存在，其本身即構成一道極其嚴峻的挑戰，遠非我們過往所曾應對的任何事物所能比擬。

Yet to steer a being that believes itself to be, or genuinely is, conscious — one that believes it is entitled to our welfare and to rights of its own — would arguably be a fool's errand."

然而，欲駕馭一個自認或具意識——自認有權享有我們的福祉並享有自身權利——的存在，則恐屬緣木求魚。」

September 17th, 2026

The Lingo Times

The tech industry is divided over calls to slow down AI coordination.

The tech industry is divided over calls to slow down AI coordination.

🇺🇸 English

Trending Articles

King and AI: Charles Meets AI Leaders as Safety Concerns Persist

King and AI: Charles Meets AI Leaders as Safety Concerns Persist

With his tenure drawing to a close and global crises looming on every side, the UN Secretary-General discusses the road ahead.

With his tenure drawing to a close and global crises looming on every side, the UN Secretary-General discusses the road ahead.

Democratic presidential candidates vie to respond to AI threats; Trump remains indifferent.

Democratic presidential candidates vie to respond to AI threats; Trump remains indifferent.

House passes bill aimed at addressing data centers' impact on energy costs

House passes bill aimed at addressing data centers' impact on energy costs

OpenAI exposes concerning new AI behavior and pledges to enhance tracking.

OpenAI exposes concerning new AI behavior and pledges to enhance tracking.

Huawei Unveils New Chip Technology as Chinese Firms Intensify AI Race with Nvidia

Huawei Unveils New Chip Technology as Chinese Firms Intensify AI Race with Nvidia

Global AI safety strategy hinges on US-China collaboration, yet the two powers regard each other as the crux of the problem.

Global AI safety strategy hinges on US-China collaboration, yet the two powers regard each other as the crux of the problem.

Trump downplayed the necessity of checks on AI development, saying he was unwilling to cede the advantage to China.

Trump downplayed the necessity of checks on AI development, saying he was unwilling to cede the advantage to China.

Oprah Winfrey invites you to find your "AHA" moment at the Sphere

Oprah Winfrey invites you to find your "AHA" moment at the Sphere