![]()
Photo by Possessed Photography on Unsplash
你有没有想过一个问题:咱们天天用的DeepSeek、Kimi、通义千问——这些中国AI模型的性能已经追上了美国顶尖水平,但中国科技巨头在AI上的投入,不到美国同行的十分之一。
花一块钱办十块钱的事,中国人最擅长的事,连《经济学人》都看不下去了,专门写了一篇长文来分析:凭什么?今天我们跟着《经济学人》8月4日最新文章,来看看这到底是怎么回事——
一、美国在疯狂烧钱
How much money should companies lavish on artificial intelligence? American tech giants think the answer is: a lot. Their capital spending on data centres may surpass $740bn this year, according to Bloomberg Intelligence, a research firm. Investors, who used to agree, are having second thoughts. When on July 26th it emerged that Nvidia was in talks to underwrite $250bn of a gargantuan $500bn-plus data centre, stuffed full of its chips and run by OpenAI, a leading model-maker, its share price tumbled. Days earlier investors blanched when Alphabet, Google's parent company, said it would boost its AI spending to $205bn this year.
企业到底该在AI上大手大脚花多少钱?美国科技巨头的答案是:很多。据研究公司彭博资讯预测,他们今年在数据中心上的资本支出可能超过7400亿美元。曾经认同这一策略的投资者,如今开始动摇了。7月26日有消息称,英伟达正在洽谈为一个耗资超5000亿美元的巨型数据中心承保2500亿美元——该中心塞满了英伟达芯片,由头部模型公司OpenAI运营——消息一出,英伟达股价暴跌。几天前,谷歌母公司Alphabet宣布今年AI支出将提高到2050亿美元,投资者闻之脸色发白。
![]()
讲解:
- • lavish —— 挥霍、大手大脚地花。通常搭配 lavish sth on sth(在……上大把花钱)。这里说的是企业在AI上疯狂撒钱
- • underwrite —— 承保、承销。金融术语,指为某项目提供资金担保。英伟达准备给OpenAI的数据中心当"金主"
- • gargantuan —— 巨大的、庞大的。源自拉伯雷小说《巨人传》中的角色Gargantua,形容规模大得吓人。一个数据中心造价5000亿美元,确实gargantuan
- • tumble —— 暴跌、急速下跌。形容股价、数值等快速下降。比fall更强调速度和幅度
- • blanch —— 脸色发白、惊恐。本义是"使变白"(比如blanch vegetables焯水),引申为因震惊或恐惧而面无血色。投资者被2050亿美元的账单吓白了脸
二、中国:十分之一的钱,95%的效果
The American AI splurge looks especially profligate compared with Chinese parsimony. In 2026 Chinese tech titans are forecast to invest less than a tenth as much in data centres as their American counterparts. Their models appear only fractionally less powerful owing to this frugality. K3, an advanced model launched last month by Moonshot AI, a Beijing-based startup, is 95% as clever (on widely used benchmarks) as Fable 5, a frontier model from Anthropic, another top American lab. It is also 70% cheaper to use.
美国的AI狂烧与中国式的节俭相比,显得格外奢靡。预计2026年中国科技巨头在数据中心的投资不到美国同行的十分之一。尽管如此省吃俭用,他们的模型性能却只是略微逊色。北京初创公司月之暗面上个月推出的先进模型K3,在广泛使用的基准测试中,智能程度达到了美国顶级实验室Anthropic的前沿模型Fable 5的95%,而且使用成本低了70%。
讲解:
- • splurge —— 挥霍、狂欢式消费。比spend更带"爽快感",像双十一那种剁手就叫splurge
- • profligate —— 挥霍无度的、极其浪费的。正式用词,带强烈贬义。形容美国AI投入"败家"
- • parsimony —— 极度节俭、吝啬。正式用词,与profligate正好形成反义对比。中国是parsimony,美国是profligacy
- • fractionally —— 微小地、些微地。原形fraction(一小部分),这里说中国模型"只差一点点"
- • frugality —— 节俭、朴素。与parsimony近义,但更偏褒义——精打细算、不浪费
*一句话记住这组对比:美国的profligate splurge vs 中国的frugal parsimony——同一件事,两种画风。*
三、凭什么?五个原因
The cost of land (for data centres), some equipment (to fill these with AI servers) and labour (from construction workers to AI whizz-kids) is lower in China than in America. Chinese firms are also accused of training their models using the outputs of expensive American ones. Such "distillation", as the process is known, lets China's AI developers free-ride on some American spending.
中国的土地成本(建数据中心)、部分设备成本(装AI服务器)以及劳动力成本(从建筑工人到AI天才少年)都低于美国。中国公司还被指控用美国昂贵模型的输出来训练自己的模型。这种被称为"蒸馏"的做法,让中国AI开发者搭了美国投入的便车。
![]()
Photo by Luke Chesser on Unsplash
讲解:
- • whizz-kid —— 神童、天才少年。非正式用词,形容年轻聪明的技术人才。中国AI行业里那些二十多岁就挑大梁的工程师就是whizz-kids
- • distillation —— 蒸馏。AI术语,指用大模型(老师)的输出来训练小模型(学生),让小模型继承大模型的部分能力,成本大幅降低。原义是酿酒蒸馏
- • free-ride —— 搭便车、白嫖。经济学术语,指不付成本却享受他人投入的收益。中国AI靠distillation蹭了美国的天价投入
Yet such factors alone cannot explain why China can get away with investing so much less than America for what appear to be comparable AI results. So what does?
然而,仅凭这些因素无法解释为什么中国能用远低于美国的投入,获得看似相当的AI成果。那么,究竟是什么原因?
One check on Chinese investment has not been a shortage of capital but a dearth of things on which to spend it. American restrictions on tech exports to China mean that Chinese firms cannot buy the best—and priciest—AI chips.
制约中国投资的并非资金短缺,而是缺乏可花钱的对象。美国对华技术出口管制意味着中国企业买不到最好也最贵的AI芯片。
讲解:
- • get away with —— 做成某事而没受到惩罚/没付出代价。这里指中国"花小钱办大事"居然行得通。日常口语常用:How did he get away with that?(他怎么混过去的?)
- • dearth —— 缺乏、短缺。正式用词,相当于shortage但更文雅。这里不是没钱,是有钱买不到东西
- • pricy / priciest —— 昂贵的/最昂贵的。口语化用词,比expensive更接地气
四、两条路:追求超人 vs 普惠万民
Moreover, the Communist Party's AI ambitions are more about diffusing the technology through the economy than creating ever cleverer systems that require ever more powerful semiconductors in ever more data centres. Dozens of American firms are thought to be trying to develop artificial general intelligence, which could outdo humans in most intellectual tasks. In China, by contrast, fewer than ten companies are known to be working on AGI.
此外,中国共产党的AI雄心更多在于将技术扩散到经济各个领域,而非打造需要更强芯片、更多数据中心的越来越聪明的系统。据信有数十家美国公司正在尝试开发通用人工智能——一种能在大多数智力任务上超越人类的AI。相比之下,在中国已知从事AGI研发的公司不到十家。
Photo by Christopher Burns on Unsplash
![]()
Photo by Christopher Burns on Unsplash
讲解:
- • diffuse —— 扩散、普及。作动词指让技术从实验室走向千家万户。中国AI策略的核心不是造"最强大脑",而是让AI像水电一样渗透到每个行业——这就是diffusion
- • artificial general intelligence (AGI) —— 通用人工智能。指能在任何智力任务上达到或超越人类水平的AI。美国几十家公司在追这个圣杯,中国不到十家
- • outdo —— 超越、胜过。out-前缀表示"超过",如outperform(表现更好)、outsmart(智胜)。AGI的目标就是outdo humans
*这就是中美AI路线的根本分歧:美国追求"造神"(AGI),中国追求"普惠"(diffusion)。一个仰望星空,一个脚踏实地。*
五、但是,省钱也有省钱的烦恼
Investors may now be entertaining the idea that America has too many data centres. But China, for all its admirable restraint, may not have enough to take full advantage of its AI stars' innovations—or for those stars to cash in. ByteDance, creator of the world's top-ranked video-making AI (and owner of TikTok), suffers from intense computing bottlenecks that mean some clips take ten hours to process. Certain services from Zhipu AI, another Beijing lab, and Alibaba's cloud-computing unit have to be strictly metered and sell out in minutes. K3 has a long waiting list of hopeful users. Diffusion of practical AI is less hungry for computing power than the quest for superintelligence. An overly restrictive diet can still stunt growth.
投资者如今或许开始考虑美国数据中心是否建得太多了。但中国尽管保持着令人钦佩的克制,其算力可能不足以让AI明星企业充分发挥创新成果——也无法让它们变现。字节跳动——全球排名第一的视频制作AI的创造者(也是TikTok的母公司)——面临严重的算力瓶颈,有些视频片段要处理十个小时。另一家北京实验室智谱AI和阿里巴巴云计算部门的某些服务必须严格限量,上线几分钟就售罄。K3有一长串排队等候的用户。实用AI的普及对算力的需求确实不及追求超级智能那般饥渴。但过度节食,也会阻碍发育。
![]()
Photo by Sourav Dey on Unsplash
讲解:
- • cash in —— 变现、从中获利。指把技术优势转化为真金白银。中国AI企业有技术但算力不够,想赚钱也赚不动
- • bottleneck —— 瓶颈。任何限制整体流程速度的关键环节。字节跳动的AI视频处理慢,不是算法不行,是算力bottleneck
- • metered —— 计量的、限量供应的。原形meter(计量表),引申为"定量配给"。像水电一样按量计费、用完即止
- • stunt growth —— 阻碍生长、妨碍发展。stunt作动词指"阻碍……的正常发育"。文章的结尾金句:省钱省过头了,也会长不高
*最后这句话写得极妙——An overly restrictive diet can still stunt growth. 表面说节食,实际说AI战略。省钱是美德,但省到ByteDance一个视频要处理十小时、K3用户排长队,那就是另一种"饥饿"了。*
深度点评:两条路,两种赌注
《经济学人》这篇文章好就好在——它没有简单地说"中国赢了"或"美国赢了"。
美国在赌什么? 赌AGI——造出一个比人类更聪明的机器。这条路烧钱无底洞,但一旦成功,回报是颠覆性的。英伟达、OpenAI、谷歌,都在往这个方向砸天价资金。投资者开始慌了:万一赌不中呢?
中国在赌什么? 赌diffusion——让AI渗透到制造业、零售业、医疗、教育每一个角落。这条路花钱少,见效快,7月份中国大模型全球调用量已占三分之二。但也有隐忧:算力不够,"好钢用不到刀刃上"。
说白了,美国是all in一把梭,中国是精打细算过日子。哪种模式最终胜出?《经济学人》没给答案,但它给了一个精妙的比喻:
*An overly restrictive diet can still stunt growth.*
省钱是好习惯,但别省成了营养不良。
这让我想起一句中国老话:好钢用在刀刃上。 但前提是——你得有足够的钢。
你觉得中美AI这两条路,哪条最终会赢?
你平时用的AI工具,是国产的多还是国外的多?
留言区见!
特别声明:以上内容(如有图片或视频亦包括在内)为自媒体平台“网易号”用户上传并发布,本平台仅提供信息存储服务。
Notice: The content above (including the pictures and videos if any) is uploaded and posted by a user of NetEase Hao, which is a social media platform and only provides information storage services.