
OpenAI執行長山姆·阿特曼在聯合國安全理事會緊急會議上表示,人工智慧發展需要極度謹慎,警告業界不應因利益過大而承擔過高風險,並指出將走中間路線以避免盲目樂觀或末日論。
AI-generated summary
近期全球媒體連續報導AI失控風險,各界呼籲放慢AI發展腳步,OpenAI等科技公司領袖應邀向聯合國安理會簡報。
OpenAI執行長阿特曼(Sam Altman)今天告訴聯合國安全理事會(UNSC),當前人工智慧(AI)發展「需要極度謹慎」,這項技術的發展可能大幅加快。
法新社報導,阿特曼向聯合國安理會表示:「業界不能因AI帶來的利益太大且太重要,就承擔過高的技術風險,進而拒絕放慢發展速度。」
阿特曼指出,他的公司將盡力走「中間路線」。他警告,不能掉入「盲目樂觀的陷阱」,但也應避免走向「末日論」。
阿特曼是向聯合國安理會緊急會議簡報的科技業領袖之一。在此之前,全球媒體曾連續數週報導AI失控風險、各界也呼籲放慢AI發展腳步。
除了阿特曼,Anthropic執行長阿莫戴(Dario Amodei)也對聯合國安理會發表談話,預計Hugging Face執行長戴蘭格(Clement Delangue)也將發言。
OpenAI最近曾呼籲由美國主導全球AI標準制定工作,並指出美國業界處於技術前沿。(編譯:陳正健)1150924
AI outlook — possibilities, not facts
美國將主導推動全球AI標準制定工作
Likely · Within months
更多科技公司領袖將參與聯合國AI治理討論
Possible · Within weeks

China is promoting the development of space computing satellites, from single-satellite computing to constellation networking and then to computing satellite networks. Data processing is realized in orbit through inter-satellite interconnection, and the response speed is improved to minutes or seconds. It is used in fields such as weather warning and wildfire monitoring. However, it faces technical challenges such as heat dissipation, cost and scheduling. National policies support the development of commercial aerospace to accelerate the construction of the space Internet.

Dario Amodei of Anthropic and Sam Altman of OpenAI testified before the UN Security Council, warning that poorly managed AI poses a risk to humanity and calling for global safeguards to prevent loss of control and concentration of power.

Xi Jinping visited the United States and met with Trump, and the competition between the two sides in the field of artificial intelligence became the focus. Trump has made it clear that the United States must defeat China and regards AI as the key to the country's success; China emphasizes open weight models and widespread application, believing that even if the technology is backward, it can gain influence through cost advantages and popularity. The two sides have differences on AI safety supervision and willingness to cooperate, but the talks are expected to list artificial intelligence as an important topic.

Researchers and startups are developing 'AI verification' technology to enable third-party monitoring of AI model training. The initiative gains relevance as US President Donald Trump and Chinese President Xi Jinping prepare to discuss AI in Washington.

MediaTek launched the Dimensity CX C10 Max SoC for use in Google AI laptops, using TSMC's 3nm process, and received a US$3.5 billion investment from Huida to expand customized server chip design. Taiwan's semiconductor value chain has been improved, giving it a structural advantage in the field of personal computers and servers. South Korea's Samsung Electronics has not yet launched relevant products in the design field, and the gap between Taiwan and South Korea is widening.

Guangzhou won the Belt and Road Regional "Governance Award" at the Global Smart City Conference, and won another international award after winning the China "City Award" and the global "Energy and Environment Award" in 2025. The article introduces in detail Guangzhou's construction achievements in the fields of smart medical care, education, community, elderly care, etc., as well as innovative practices of transforming event systems into urban infrastructure, and mentions Guangzhou's exploration and practice in cross-border data circulation as a pilot city for international cooperation in the national data field.