DAVID E. SANGER, JONATHAN SWAN, CECILIA KANG, DUSTIN VOLZ2026年9月21日去年，特朗普总统在白宫与戴维·萨克斯和马克·扎克伯格共进晚餐。 Haiyun Jiang/The New York TimesPresident Trump made clear in the past week that he views fears that A.I. could lead to mass extinction as a “HOAX,” and that he sees opponents of data centers as people who “want to end up being backwards and poor.”特朗普总统在过去一周明确表示，他认为有关人工智能可能导致大规模人类灭绝的担忧是“骗局”，并将反对建设数据中心的人视为“想要最终变得落后贫穷”的人。People who have discussed the subject with Mr. Trump in private say his outbursts are rooted in worries that if the artificial intelligence boom fueling the American economy slows down, the market could crash and recession could quickly follow.私下里曾与特朗普讨论过这个话题的人说，他这些言论的根源在于担心，如果推动美国经济的人工智能热潮放缓，市场可能会崩盘，经济衰退可能会很快接踵而至。His anti-regulatory, anti-slowdown outbursts have been encouraged by a slice of Silicon Valley’s most influential voices. Chief among them has been David Sacks, the venture capitalist who interceded to weaken an executive order in May that would have required safety reviews of new A.I. models. Mark Zuckerberg of Meta and Jensen Huang of Nvidia have also spoken to the president. All have advised Mr. Trump that the bigger risk is being overtaken by Chinese competitors.硅谷部分最具影响力的声音一直在助长他这些反对监管、反放缓的言论。其中首要的是风险投资家戴维·萨克斯，他曾出面斡旋，削弱了5月一项原本要求对新人工智能模型进行安全审查的行政命令。Meta的马克·扎克伯格和英伟达的黄仁勋也曾与总统交谈。他们都向特朗普建议，更大的风险是被中国竞争对手超越。But inside the administration, the debate is far more complex, and there are other voices in Mr. Trump’s ear, urging caution.但在政府内部，讨论要复杂得多，特朗普耳边还有其他声音在敦促他保持谨慎。The more cautious tone inside the White House reflects a growing concern that despite the strategic challenge posed by China, a summer of A.I. surprises has raised the prospect of risks that were often dismissed a year ago as the wild talk of “doomers.”白宫内部更加谨慎的基调反映出一种日益增长的担忧：尽管中国构成了战略挑战，但今年夏天以来人工智能带来的一系列意外事件增加了风险发生的可能性，而就在一年前，这些风险还经常被当作“末日论者”的疯言疯语而被置之不理。Now, even the biggest A.I. boosters in the administration concede that A.I.-fueled hacks into infrastructure, biological threats, attacks on nuclear command-and-control systems, job losses and global financial instability seem more possible than they did before A.I. agents went rogue.现在，即使是政府内部最为热衷于人工智能的人也承认，与人工智能智能体失控之前相比，由人工智能驱动的针对基础设施的黑客攻击、生物威胁、针对核指挥和控制系统的攻击、失业以及全球金融动荡现在看来都更有可能发生。The issue of A.I. safety is expected to come to a head this weekend as Treasury Secretary Scott Bessent meets in New York with his Chinese counterpart to set the agenda for Mr. Trump’s visit in Washington with President Xi Jinping of China, which starts on Wednesday. Mr. Bessent is trying to negotiate some kind of agreement between Mr. Trump and Mr. Xi that would suggest the two countries are working together on mutual A.I. safety concerns.预计人工智能安全问题将在周末成为焦点，届时财政部长贝森特将在纽约会见中国同级官员，为特朗普在华盛顿与中国国家主席习近平从周三开始的会晤设定议程。贝森特正试图在特朗普和习近平之间谈判达成某种协议，表明两国正在就共同的人工智能安全担忧开展合作。But already, officials say, some of the most expansive ideas being discussed among A.I. safety experts, including an agreement on safety standards and reviews of forthcoming Chinese and American A.I. models before their public release, are unlikely to be part of any initial accord.但官员们表示，人工智能安全专家正在讨论的一些最宏大的构想——包括就安全标准达成协议，以及在中美两国即将推出的人工智能模型公开发布前对其进行审查——目前看来已不太可能被纳入任何初步协议。本周，财政部长斯科特·贝森特在国会大厦。今年，多家银行的高管曾向贝森特警示人工智能对全球金融体系构成的风险。 Haiyun Jiang/The New York TimesOne diplomat involved in the effort said he believed “the best we can hope for” is the creation of some kind of U.S.-China working group that might build on a Biden-era agreement with China to bar the use of artificial intelligence in decisions about the employment of nuclear weapons. But that does not begin to get at the complexity of the problem of unpredictable A.I.-driven agents, which took the industry and the government by surprise this summer.一位参与此项工作的外交官表示，他认为“我们所能期盼的最好结果”是成立某种形式的美中工作组，该工作组可能会建立在拜登时代与中国达成的一项协议的基础之上，该协议禁止在核武器使用决策中运用人工智能。但这根本无法触及不可预测的人工智能智能体这一问题的复杂性，今年夏天，这个问题让业界和政府都措手不及。Nor does it get to the fundamental issues surrounding imposing safeguards on frontier models. The United States could go its own way and impose mandatory reviews of the safety of new, American-generated models, but Mr. Trump and his aides believe that would effectively drive the global market to Chinese largely open-source alternatives. And because those open-source models can be modified by users — who could remove the “guardrails” meant to limit their powers — a safety review may mean little.这项协议也没有触及对前沿模型实施安全保障措施的根本问题。美国可以单独行动，强制对美国生成的新模型进行安全审查，但特朗普及其助手认为，这将在实际上把全球市场推向主要基于开源技术的中国替代品。而且，由于这些开源模型可以被用户自行修改——用户可以移除旨在限制其能力的“护栏”——安全审查可能意义不大。Mr. Trump wants to meet with the A.I. executives next week, according to company representatives. (A White House official said on Thursday that nothing was on the schedule as of now.)据企业代表称，特朗普希望下周会见人工智能高管。（一位白宫官员周四表示，目前日程上尚无相关安排。）If that session happens, it would presumably include some of the A.I. executives who, in recent days, have been calling for the kind of slowdown in the release of new models that Mr. Trump says he opposes.如果这次会议如期举行，一些人工智能高管很可能参加，他们近日来一直在呼吁放缓新模型的发布，而特朗普表示他反对这种做法。Few expect Dario Amodei, a founder of Anthropic, to be among the invitees: His company’s products have been banned by the Pentagon, and he appeared to anger Mr. Trump by proposing last weekend a detailed plan for the industry to pace itself and bring in independent safety inspectors. Mr. Trump dismissed him, and denounced an unnamed class of “Conspiracy Theorists, Treasonists, Traitors and Leakers” who were politicizing the debate over A.I. and undercutting American leadership.几乎没有人预计Anthropic的创始人达里奥·阿莫迪会在受邀之列：他公司的产品已被五角大楼禁用，而且他在上周末提出了一项详细计划，要求该行业控制发展步伐，并引入独立的安全检查员，这似乎激怒了特朗普。特朗普对他不屑一顾，并且不点名地谴责了一批“阴谋论者、叛国者、卖国贼和泄密者”，称他们将人工智能的辩论政治化，并削弱美国的领导地位。But another Anthropic founder, Tom Brown, who raises less ire in the White House, talks to officials regularly. In recent days, Mr. Trump has also talked with Sam Altman of OpenAI and Elon Musk of SpaceX, both of whom endorsed Mr. Amodei’s plan to slow down, test new models for vulnerabilities and let in outside inspectors, at least in general terms.但Anthropic的另一位创始人汤姆·布朗在白宫引起的愤怒较少，他经常与官员沟通。最近几天，特朗普还与OpenAI的萨姆·奥特曼和SpaceX的埃隆·马斯克进行了交谈，这两人至少大体上都支持阿莫迪的计划：放缓步伐、测试新模型的漏洞并允许外部检查员介入。Mr. Bessent and the White House chief of staff, Susie Wiles, have been increasingly concerned over the past several months that A.I. could cause massive disruptions across industries, including hacks into major financial institutions that could destabilize the world economy.过去几个月里，贝森特和白宫幕僚长苏西·怀尔斯越来越担心，人工智能可能会在各行各业造成大规模破坏，包括对主要金融机构的黑客攻击，这可能会破坏世界经济的稳定。Ms. Wiles this year began organizing an informal group to think about how to deal with the risk created by new, more powerful models. Mr. Bessent is chief among the members of that informal group, an administration official said, but other experts are called in depending on the topic.怀尔斯今年开始组织一个非正式小组，思考如何应对更强大的新模型带来的风险。一位政府官员说，贝森特是这个非正式小组的主要成员之一，但根据讨论主题的不同，也会邀请其他专家参与。The White House pointed to Ms. Wiles’s social media post on May 6 as the foundational statement of how Mr. Trump balances the risks and opportunities of the new technology. “This administration has one goal: ensure the best and safest tech is deployed rapidly to defeat any and all threats,” she wrote at the time. She said that “the White House will continue to lead an America First effort that empowers America’s great innovators, not bureaucracy, to drive safe deployment of powerful technologies while keeping America safe.”白宫指出，怀尔斯在5月6日发布的社交媒体帖子是特朗普关于如何平衡这项新技术的风险与机遇的基础性表态。“本届政府只有一个目标：确保快速部署最好、最安全的技术，以击败任何及所有威胁，”她当时写道。她表示，“白宫将继续领导一项‘美国优先’的努力，赋能美国的伟大创新者——而非官僚机构——推动强大技术的安全部署，同时保障美国的安全。”Asked about Mr. Trump’s accusation that A.I. risk is overblown, Elizabeth Huston, a White House spokeswoman, said that “President Trump is saying that Democrats using A.I. concern as a way to score points is a hoax, just like their claims that climate change would kill everyone in six years.”当被问及特朗普指责人工智能风险被夸大一事时，白宫发言人伊丽莎白·休斯顿表示：“特朗普总统的意思是，民主党人利用对人工智能的担忧来捞取政治资本，这是一种骗局，就像他们声称气候变化将在六年内杀死所有人一样。”She added that Mr. Trump had reversed the “failed approach” of the Biden administration, which she said had “stifled innovation, hurt consumers and nearly cost America its competitive edge over China.”她还说，特朗普扭转了拜登政府的“失败做法”，她表示那种做法“扼杀了创新，伤害了消费者，几乎让美国丧失了对中国的竞争优势”。But at a moment when A.I. has emerged as arguably the most important long-term issue for the country’s economic health and national security, in the Trump White House there is no single person in charge of coordinating A.I. policy. Marco Rubio, in his role as national security adviser, is rarely discussed as a player on these issues. The White House eliminated the post of deputy national security adviser for cyber and emerging technologies created during the Biden administration.但在人工智能已成为对国家经济健康和国家安全最重要的长期问题之际，特朗普的白宫却没有专人负责协调人工智能政策。国家安全顾问鲁比奥很少被视为这些议题的决策者。白宫取消了拜登政府时期设立的负责网络和新兴技术的副国家安全顾问职位。And the national cyber director, a position created by Congress, has only indirect authority over artificial intelligence issues. The current director is Sean Cairncross, a former senior official at the Republican National Committee, who was light on cyber and A.I. experience before taking the post.由国会设立的国家网络总监一职对人工智能问题则只有间接权力。现任总监是共和党全国委员会前高级官员肖恩·凯恩克罗斯，他在上任前缺乏网络和人工智能方面的经验。And there has been a purging of some technical expertise, including at the Cybersecurity and Infrastructure Security Agency, part of the Department of Homeland Security. (The acting director of CISA said last week that the agency was hiring about 250 new employees after shedding around a thousand jobs — roughly a third of its work force — since Mr. Trump returned to the White House.)此外，一些技术专家已被裁撤，包括国土安全部下属的网络安全和基础设施安全局。（该局代理局长上周表示，自特朗普重返白宫以来，该机构裁减了大约一千个职位——约占其员工总数的三分之一——目前正在招聘约250名新员工。）As a result, A.I. has become the administration’s latest hot potato, with various officials dipping in and out of influence depending on the topic, or the day of the week.结果，人工智能成了该政府最新的一块烫手山芋：各位官员时而加入，时而退出，取决于讨论的是什么议题，甚至取决于当天是星期几。Mr. Trump seems driven chiefly by the sense that A.I. is the way to economic growth and nothing can be allowed to stand in its way.特朗普似乎主要是受这样一种感觉的驱使：人工智能是实现经济增长的途径，绝不允许任何事物阻挡它的发展。Anthropic公司首席执行官达里奥·阿莫迪上周在旧金山发表演讲。该公司的产品已被五角大楼禁用，他提出的关于对人工智能实施行业限制的计划也遭到总统反对。He had latched onto this idea during the late campaign season. In a private meeting with OpenAI executives in June 2024, Mr. Trump discussed with them the idea of creating a version of “Operation Warp Speed” — his all-of-government enterprise to fast-track Covid vaccines — for the building of data centers.在竞选季后期，他就抓住了这个想法。在2024年6月与OpenAI高管的一次私人会议上，特朗普与他们讨论了创建一个类似“曲速行动”（ 他为加速新冠疫苗研发而启动的政府全员行动）的计划，用于建设数据中心。Ever since, he has been determined to fuel A.I. growth at nearly any political cost, to keep alive what he has referred to as “the golden goose.” Mr. Trump, a real estate developer, is eager to keep building, dismissing the concerns of states and towns that want to slow or halt construction.从那时起，他就决心几乎不计任何政治代价来推动人工智能的发展，以保住他所说的“金鹅”。身为房地产开发商的特朗普热衷于持续推进建设，不理会希望放缓或停止建设的州和城镇的担忧。“The robots will not be taking over,” Mr. Trump insisted when he called into a large conference where Mr. Sacks and Mr. Huang were onstage this week. “The A.I. will not be taking over the rest of the world.”“机器人不会接管一切，”本周，当萨克斯和黄仁勋在一场大型会议上登台发言时，特朗普通过电话连线坚称。“人工智能不会接管世界其他地方。”Maggie Haberman对本文有报道贡献。David E. Sanger报道特朗普政府和一系列国家安全问题。他在时报任职超过40年，著有四本关于美国国家安全挑战的书。Jonathan Swan是《纽约时报》白宫记者，报道特朗普总统。Cecilia Kang是《纽约时报》驻华盛顿的科技和监管政策记者。她从事科技报道已有20余年。Dustin Volz为时报报道网络安全和情报领域，常驻华盛顿。翻译：纽约时报中文网点击查看本文英文版。