科學家從三位百歲以上的巴西姊妹身上探索長壽之道

Recently Reuters reported the following:

Scientists seek clues to longevity from three Brazilian sisters over 100 

REUTERS - Reporting by Aline Massuca in Rio de Janeiro and Victoria Pacheco in Sao Paulo; Editing by Manuela Andreoni and Bill Berkrot

June 24, 202612:32 PM PDTUpdated 17 hours ago

RIO DE JANEIRO, June 24 (Reuters) - What is the secret to a long life? Three Brazilian sisters with a combined age of 316, who were named by Guinness this month ​as the oldest living trio of siblings in the world, may help researchers find ‌out.

The DNA Longevo Project, a study led by scientist Mayana Zatz from the University of Sao Paulo, aims to investigate the biological factors behind aging.

Findings from the three sisters' case could help scientists better understand why some people remain ​physically and cognitively resilient at exceptionally advanced ages.

Researchers will compare nonagenarians and centenarians with people who ​have developed frailty, cognitive decline or chronic diseases, seeking traits linked to longevity.

"Through ⁠DNA testing, we look for protective genes, and we know there are several of them," said Zatz, ​who coordinates the university's Human Genome Research Center. "The more people we have who live past 100, especially ​families with multiple centenarians, the more accurate our research will be in identifying them."

Scientists believe inherited factors may play a larger role than environmental influences in preserving health and function later in life.

The sisters, Zulina de Deus Nunes, 103, Zoraide de ​Deus Mota, 104, and Levita de Deus Nunes, 109, who live in Rio de Janeiro, were ​identified through LongeviQuest, a global organization that verifies longevity records and partners with Guinness World Records.

"When sisters reach that age, ‌there ⁠is clearly a strong genetic component," said Ben Meyers, CEO of LongeviQuest. "But because they live near each other, they also have a support network, with family able to help when needed. There is definitely a community aspect as well."

The three sisters credit their longevity to a healthy diet and an active lifestyle.

Zulina ​recalled a childhood spent ​swimming and fishing in rivers. "Everything ⁠was fresh. We didn't have a refrigerator," she said.

"Breastfeeding is incredibly important," Zoraide added.

The sisters otherwise led fairly ordinary lives. Levita worked as a craftswoman and ​later at a television network. Zoraide worked as a nurse and raised five ​children, while ⁠Zulina, a stay-at-home mom, raised six.

Levita looks back on her life without regrets. "I had a good childhood and adolescence. I can't complain."

Researchers hope to understand how genetic factors, rather than lifestyle, help protect the heart, muscles, ⁠and cognitive ​function from the ravages of aging.

The study's goal, said researcher ​Joao Paulo Guilherme, who works with Zatz, "is to reach 500 centenarians so we can draw more definitive conclusions about longevity."

Translation

科學家從三位百歲以上的巴西姊妹身上探索長壽之道

裡約熱內盧，6月24日（路透）－長壽的秘訣是什麼？三位巴西姊妹，加起來316歲，本月被《健力士世界紀錄大全》認證為世界上最長壽的三姊妹，或許能幫助研究人員找到答案。

由聖保羅大學科學家Mayana Zatz領導的「DNA長壽計劃」旨在研究老化背後的生物學因素。

這三位姊妹的案例或許能幫助科學家更能理解，為什麼有些人能在如此高齡時依然保持良好的身心狀態。

研究人員將把九十歲以上和百歲老人，與出現體弱、認知能力下降或慢性疾病的人群去比較，以尋找與長壽相關的特徵。

該大學人類基因組研究中心協調員Zatz說:「透過DNA檢測，我們尋找保護性基因，而且我們知道存在好幾種這樣的基因」; 「我們擁有的百歲老人越多，尤其是擁有多位百歲老人的家庭，我們的研究就越能準確地識別出這些基因」。

科學家認為，在晚年可保持健康和機能方面，遺傳因素可能比環境因素發揮更大的作用。

居住在裡約熱內盧的Zulina de Deus Nunes（103歲）、Zoraide de Deus Mota（104歲）和Levita de Deus Nunes（109歲）三姐妹是透過一個叫LongeviQuest的驗證長壽紀錄並與健力士世界紀錄 合作的全球性組織所確認。

LongeviQuest 的執行長Ben Meyers說: 「姐妹有這個年紀，長壽顯然有很強的遺傳因素」; 「亦由於她們互相住得很近，也擁有一個支持網絡，家人可以在需要時提供幫助。社區因素顯然也發揮了作用」。

三姊妹將她們的長壽歸功於健康的飲食和積極的生活方式。

Zulina回憶起童年時在河裡游​​泳和釣魚的時光。 她說:「所有東西都很新鮮。我們家沒有冰箱」。

Zoraide補充道: 「母乳哺育非常重要」。

除此之外，姐妹們過著相當普通的生活。Levita曾是女手工匠，後來在電視台工作。Zoraide是一名護士，養育了五個孩子，而Zulina則是一位全職媽媽，養育了六個孩子。

Levita回顧自己的人生，毫無遺憾。 “我的童年和青少年時期都很美好。我沒有什麼可抱怨。”

研究人員希望了解遺傳因素而非生活方式是如何幫助保護心臟、肌肉和認知功能免受老化的損害。

與Zatz合作的研究員Joao Paulo Guilherme表示，這項研究的目標是“找到500位百歲老人，以便我們能夠對長壽得出更確實的結論。”

So, three Brazilian sisters with a combined age of 316 may help scientists better understand why some people remain physically and cognitively resilient at exceptionally advanced ages. Apparently, many people are looking forward to seeing a more definitive conclusions about longevity.

「好得令人難以置信」：一項關於癌症治療時機的中國研究被撤稿 (2/2)

Recently The New York Times reported the following:

‘Too Good to Be True’: A Chinese Study on Timing Cancer Therapy Is Retracted (2/2)

In a notice flagging a series of problems with a clinical trial, the journal Nature Medicine said its editors “no longer have confidence in the integrity of the results.”

By Rebecca Robbins - Rebecca Robbins is a Times reporter covering the pharmaceutical industry. She has been reporting on health and medicine since 2015.

June 25, 2026

(continue)

Dr. Yongchang Zhang, the study’s senior author, said in a statement that an internal review had “confirmed that part of the study execution and manuscript preparation might not reach the standards for publication in a high-impact journal.”

Dr. Zhang, a researcher at the Chinese hospital where the study was conducted, added: “We acknowledge these shortcomings and sincerely apologize for any inconvenience caused to the journal and its readers.” He did not provide an explanation for the problems cited by the journal.

The study had enrolled 210 patients with advanced lung cancer at Hunan Cancer Hospital in Changsha, a city in south-central China. Patients were randomly assigned to receive infusions of an immunotherapy — Merck’s blockbuster drug Keytruda or Tyvyt, which is not approved in the United States — either before or after 3 p.m.

The study reported that tumors did not progress for 11 months in patients who received the earlier infusions, compared with six months for those given the later infusions. Patients receiving the earlier infusions lived for 28 months, compared with 17 months for those infused later in the day.

Those “were numbers we usually associate with new blockbuster drugs, not scheduling decisions,” wrote Dr. Gilberto Lopes, an oncologist at the University of Miami. And rescheduling a patient “costs nothing,” he noted.

Dr. Anil Makam, an epidemiologist and health services researcher at the University of California, San Francisco, said such a drastic benefit, if real, would have prompted infusion clinics to overhaul staffing and scheduling to shift appointments to earlier in the day.

“If we believed the effects, it would be malpractice not to,” he said.

But within days after the study’s publication in Nature Medicine, online sleuths and physicians like Dr. Makam began raising concerns on social media and in blog posts. Less than three weeks after the study was published, the journal published an editor’s note saying it was investigating the issues.

In a statement on Thursday, Dr. João Monteiro, chief editor of Nature Medicine, which is published by Springer Nature, said, “We are grateful to the research community for bringing these concerns to our attention.”

Some scientists have long been intrigued by the idea of harnessing the body’s circadian rhythms to try to make drugs more effective or less toxic. For example, enzymes that the body uses to break down certain medicines are more abundant at certain times of day.

But critics of the Chinese study argued that immunotherapy works very differently from fast-acting drugs like Tylenol. The drug Keytruda lingers in the body and takes effect over several weeks. The study did not identify a biological reason for a difference in treatment timing that would produce such a significant benefit.

“There wasn’t any sound scientific rationale behind it,” said Dr. Roy Herbst, the incoming director of the Dartmouth Cancer Center in New Hampshire and Vermont.

In March, a team led by European researchers reported the findings of an analysis that looked at whether more than 3,000 cancer patients across eight studies had received immunotherapy in the morning or the afternoon. Their study, funded by the drugmaker Roche, which sells a number of cancer drugs, concluded that timing was “unlikely to be a critical determinant” of how well patients did.

Other studies looking back at outcomes have found an association between the time of day that patients receive a cancer immunotherapy and how well they fare. But the why remains unclear.

Doctors said it was possible that more energetic, healthier patients might opt for morning slots. Poorer or rural patients who live far from an infusion center — and tend to fare worse — might ask for afternoon slots because they need to spend the morning traveling to their appointment.

Translation

「好得令人難以置信」：一項關於癌症治療時機的中國研究被撤稿 (2/2)

《自然醫學》雜誌在聲明中指出一項臨床試驗存在一系列問題，並表示其編輯「不再對研究結果的可靠性有信心」

 （繼續）

該研究的資深作者Yongchang Zhang博士在聲明中表示，內部審查「證實部分研究執行和稿件準備工作可能不符合高影響力期刊的發表標準」。

Zhang博士是該研究開展所在中國醫院的研究員，他補充說：「我們承認這些不足，並對給期刊及其讀者帶來的任何不便深表歉意」。他沒有解釋期刊所指出的問題。

這項研究在湖南省長沙市湖南省腫瘤醫院 招募了210名晚期肺癌患者。患者被隨機分配接受免疫療法輸注－默克公司的重磅藥物Keytruda或Tyvyt（該藥尚未在美國被批準）－分別在下午3點前或下午3點後進行。

研究報告顯示，接受較早輸注的患者腫瘤持續11個月沒有增大或惡化，而接受較晚輸注的患者腫瘤沒增大或惡化持續時間為6個月。接受較早輸注的患者平均存活期為28個月，而接受較晚輸注的患者平均存活期為17個月。

University of Miami腫瘤學家Gilberto Lopes博士寫道：「這些數據通常與重磅新藥相關，而不是與輸注時間安排相關的」。他還指出，重新安排患者的輸注時間「無需任何成本」。

加州大學舊金山分校 的流行病學家和衛生服務研究員Anil Makam博士表示，如果這種顯著的益處真的存在，那麼輸液診所應該會徹底改革人員配備與排班制度，調整時間，將輸液提前到一天中的早些時候。

他說:「如果我們相信這些效果，那麼不去調整就是瀆職」。

然而，這項研究發表在《自然醫學》雜誌幾天後，像Makam博士這樣的網路偵探和醫生就開始在社群媒體和部落格文章中表達擔憂。研究發表不到三週後，雜誌發表了編輯說明，表示正在調查這些問題。

《自然醫學》雜誌主編João Monteiro博士週四在一份聲明中表示：「我們感謝研究界讓我們注意到這些問題」。該雜誌由Springer Nature出版。

一些科學家長期以來一直對利用人體晝夜節律來提高藥物療效或降低藥物毒性的想法很感興趣。例如，人體用於分解某些藥物的酵素在一天中的某些時段含量更高。

但批評這項中國研究的人士認為，免疫療法的作用機制與Tylenol等速效藥物截然不同。藥物Keytruda會在體內停留數週，逐漸起效。研究並未找到治療時間差異有產生如此顯著療效的生物學原因。

即將上任的新罕布什爾州和佛蒙特州Dartmouth癌症中心主任Roy Herbst博士說:「這項研究背後沒有任何合理的科學依據」。

今年3月，一個由歐洲研究人員領導的團隊發表了一項分析結果，該分析考察了八項研究中3,000多名癌症患者接受免疫療法的時間（上午或下午）。這項由Roche製藥公司資助的研究得出結論，治療時間「不太可能是決定」患者療效的關鍵因素。Roche製藥公司銷售多種抗癌藥物。

其他回顧性研究發現，患者接受癌症免疫療法的時間與治療效果之間存在關聯。但原因尚不明確。

有醫生表示，精力更充沛、身體更健康的患者可能更傾向於選擇上午的治療時段。而居住地偏遠、遠離輸液中心的貧困患者（療效對他們往往較差）則可能要求下午的治療時段，因為他們需要花費上午的時間前往治療地點。

So, recently a medical journal article has caught the attention of cancer patients and doctors worldwide because of its extraordinary conclusion. Simply changing the time of day that immunotherapy was administered appeared to produce a stunning benefit for lung cancer patients, and eventually the study was retracted. Apparently, it is possible that healthier patients may have opted for morning slots while poorer patients, usually live in the rural area and tend to fare worse, may ask for afternoon slots because they need to spend the morning traveling to receive the immunotherapy in the afternoon.

「好得令人難以置信」：一項關於癌症治療時間的中國研究被撤稿（1/2）

Recently The New York Times reported the following:

‘Too Good to Be True’: A Chinese Study on Timing Cancer Therapy Is Retracted (1/2)

In a notice flagging a series of problems with a clinical trial, the journal Nature Medicine said its editors “no longer have confidence in the integrity of the results.”

By Rebecca Robbins - Rebecca Robbins is a Times reporter covering the pharmaceutical industry. She has been reporting on health and medicine since 2015.

June 25, 2026

Early this year, a medical journal article caught the attention of cancer patients and doctors worldwide because of its extraordinary conclusion. Simply changing the time of day that immunotherapy was administered appeared to produce a stunning benefit for lung cancer patients.

Those who received IV infusions in the morning had their cancer kept at bay for twice as long as those who got it in the afternoon, according to the results from a clinical trial in China and published in the journal Nature Medicine in February. The study also reported that the patients lived nearly twice as long.

Several oncologists said that in recent months they and their hospitals had received a flurry of calls from patients inquiring about switching to morning infusions.

But on Wednesday, Nature Medicine retracted the study, citing a list of inconsistencies and irregularities in the trial’s design and results.

“It was too good to be true,” said Dr. Toni Choueiri, an oncologist at the Dana-Farber Cancer Institute in Boston who helped conduct the post-publication review that led to the retraction.

Among the issues that the journal cited in its retraction notice: Records that were supposed to be locked before the study started were changed midway through. There were discrepancies between the Chinese version of the study’s plan and the translated version. Every patient remained treated and tracked in the study’s first year, and no one dropped out because of side effects — highly unusual in an oncology study. And unusual patterns were found in the timing of follow-up scans.

“Due to the amount and nature of the problems identified, the editors no longer have confidence in the integrity of the results,” the journal said.

Most of the study’s 28 authors were in China, with several collaborators in Europe. The study was funded by the Chinese government.

China has been pumping money into its hospitals and drug companies, fueling a surge of patents, publications and new clinical trials. In just a few years, the country has rapidly transformed into a powerhouse in drug development, a shift that some U.S. officials, doctors and executives see as a threat to longstanding American dominance in the field.

China’s critics often question the reliability of its biomedical research. Experts said that, similar to studies in the United States, China’s research output spans a wide range in quality: Some Chinese scientists run their studies at the most meticulous standards. Others were said to cut corners.

(to be continued)

Translation

「好得令人難以置信」：一項關於癌症治療時間的中國研究被撤稿（1/2）

《自然醫學》雜誌在聲明中指出一項臨床試驗存在一系列問題，並表示其編輯「不再相信該研究結果的可靠性」

今年年初，一篇醫學期刊文章因其非凡的結論而引起了全球癌症患者和醫生的注意。文章指出，光是改變免疫療法的給藥時間，就能為肺癌患者帶來驚人的益處。

根據2月發表在《自然醫學》雜誌上的一項中國臨床試驗的結果，上午接受靜脈輸液的患者，其控制癌症的時間是下午接受輸液患者的兩倍。該研究還報告稱，上午接受輸液的患者存活期也延長了近一倍。

多位腫瘤科醫師表示，近幾個月來，他們和他們的醫院接到大量病患來電，詢問如何改為上午輸液。

但在周三，《自然醫學》雜誌撤回了這項研究，理由是試驗設計和結果有一系列不一致和不規範之處。

波士頓丹娜-法伯癌症研究所 的腫瘤科醫生Toni Choueiri博士說道:「這簡直好得令人難以置信」。她參與了協助進行發表後的審查，並最終導致撤稿的人仕。

該期刊在撤稿聲明中列舉的問題包括：本應在研究開始前鎖定的記錄在研究進行到一半時被更改；研究方案的中文版和翻譯版之間存在差異；所有患者在研究的第一年都接受了治療和隨訪，而且沒有人因副作用而退出 - 這在腫瘤學研究中極為罕見；此外，隨訪掃描的時間安排也存在異常模式。

期刊表示:「鑑於發現的問題數量之多、性質之複雜，編輯們對研究結果的可靠性已不再有信心」。

這項研究的28位作者大多來自中國，另有幾位合作者在歐洲。該研究由中國政府資助。

 中國一直在大力投資其醫院和製藥公司，推動了專利、論文發表和新臨床試驗數量的激增。短短幾年內，中國已迅速轉型為藥物研發強國，這一轉變被一些美國官員、醫生和企業高管視為對美國在該領域長期主導地位的威脅。

中國的批評者經常質疑其生物醫學研究的可靠性。專家表示，與美國的研究類似，中國的研究成果品質參差不齊：一些中國科學家以極為嚴謹的標準進行研究，而另一些科學家則被指走捷徑。

（待續）

中國人工智能模型縮小與 Anthropic 和 OpenAI 的差距 (2/2)

Recently The New York Times reported the following:

Chinese A.I. Models Close the Gap with Anthropic and OpenAI (2/2)

Silicon Valley engineers recently flocked to new technology from a Chinese company, Z.ai, that is almost as good as its American competitors but much cheaper.

By Cade MetzKaren Weise and Meaghan Tobin - Cade Metz reported from San Francisco, Karen Weise from Seattle and Meaghan Tobin from Taipei, Taiwan.

June 25, 2026

(continue)

Companies can still use the model without sending data back to China in violation of U.S. export rules, as long as they are careful about how they set up their systems, said Wei Chen, chief legal officer at Infoblox, a network security company.

“The Chinese models do not have the same restrictions if you host them yourself or you go through another provider,” Mr. Ahmad of alphaXiv said. “Right now, there are more restrictions on models from Anthropic.”

After DeepSeek’s release in 2025, governments around the world passed regulations limiting its use because of data security concerns. But so far, GLM-5.2 has not raised similar alarms, Ms. Chen said.

Anthropic and OpenAI have accused Chinese companies of improperly harvesting data from their A.I. systems to accelerate the development of the Chinese technology. On Wednesday, Anthropic sent a letter, viewed by The New York Times, to Senators Tim Scott, Republican of South Carolina, and Elizabeth Warren, Democrat of Massachusetts, accusing the Chinese tech giant Alibaba of “brazenly” and “illicitly” trying to copy its technology through 24,000 fraudulent accounts.

Alibaba declined to comment.

When Mustafa Suleyman, the head of Microsoft’s A.I. lab, unveiled a suite of new models this month, he emphasized that they had been built from scratch on data that the company had commercially licensed.

“That means that you can put it into production in a very trustworthy way with complete confidence,” he said.

Using data from one system to train another — a process called distillation — is common in A.I. development. But the Anthropic and OpenAI terms of service forbid anyone to surreptitiously harvest data for distillation. It is not clear whether Z.ai used distillation in the development of its technology.

But distillation alone cannot build a top A.I. system. That requires several other complex techniques as well, said Charles O’Neill, head of model training at Baseten, a company that sells access to GLM-5.2.

“This narrative that all of the capabilities of these models are coming from Anthropic is not as true as people say it is,” Mr. O’Neill said.

Chinese A.I. start-ups can offer their models as an open-source technology at far lower prices in part because the industry has benefited from years of support from the Chinese government, which views A.I. as a critical engine of economic growth.

Many executives have said U.S. companies should not open-source their technology because it could be used in harmful situations. But other experts argue that if regulators stifle open-source technology in the United States, China will gain a significant edge.

Because China produces most of the top-performing open-source systems, they say, U.S. developers will build their software atop those technologies. In the long run, that could put China at the heart of A.I. development.

Some argue that Chinese systems will always trail the top U.S. models because U.S. export controls limit the flow of the specialized computer chips needed to train A.I. technologies. Z.ai and other Chinese start-ups spend millions for access to chips in data centers outside China.

Z.ai’s filings in Hong Kong show that in the first half of 2025, the company spent more than seven times its revenue on expenses that essentially boiled down to fees for such computing services.

Still, experts estimate that China is just six months or less behind the American companies.

“There had been this speculation that the export controls would eventually bite and the gap would widen between American frontier models and their Chinese models, but GLM is pushing things in the other direction,” said Jeffrey Ding, an assistant professor at George Washington University who specializes in emerging technology and international relations.

And with Fable and Mythos sidelined, many businesses have realized the importance of having alternatives.

“There is a bit of apprehension at large organizations about loyalty,” said Justin Summerville, who runs data analytics at OpenRouter. “Who knows what the top model will be in three weeks?”

Translation

中國人工智能模型縮小與 Anthropic 和 OpenAI 的差距 (2/2)

矽谷工程師最近紛紛湧向一家中國公司 Z.ai 的新技術，它幾乎與美國競爭對手不相伯仲，但價格卻便宜得多

（繼續）

網路安全公司 Infoblox 的首席法務官Wei Chen表示，只要企業謹慎設置係統，就無需將數據發送回中國，而可以在不違反美國出口管制條例的情況下使用該模型。

alphaXiv 的 Ahmad 先生表示： “如果您在自己的伺服器上運行或通過其他提供商，則中國模型不受同樣的限制。目前，Anthropic 的模型受到的限制更多。”

DeepSeek於2025年發佈後，基於資料安全的擔憂，世界各國政府紛紛推出法規限制其使用。但Chen女士表示，到目前為止，GLM-5.2尚未引發類似的擔憂。

Anthropic和OpenAI指責中國公司不正當地從其人工智能系統中竊取數據，以加速中國技術的發展。週三，Anthropic致函南卡羅來納州共和黨參議員Tim Scott和馬薩諸塞州民主黨參議員Elizabeth Warren（《紐約時報》看過該信函），指責中國科技巨頭阿里巴巴 “明目張膽” 且 “非法” 地試圖通過2.4萬個欺詐賬戶複製其技術。

阿里巴巴拒絕置評。

本月，微軟人工智能實驗室負責人Mustafa Suleyman發佈了一系列新模型，他強調這些模型是基於微軟已獲得商業許可的數據從零開始建構的。

他說:「這意味著您可以完全放心地以非常可靠的方式將其投入製作」。

利用一個系統的資料來訓練另一個系統 - 這個過程被稱為資料蒸餾 - 在人工智能開發中很常見。但Anthropic和OpenAI的服務條款禁止任何人暗中收集資料用於資料蒸餾。目前尚不清楚Z.ai在其技術開發過程中是否使用了蒸餾方法。

但僅靠數據蒸餾是無法建構頂級的人工智能系統。 Baseten公司模型訓練負責人Charles O’Neill表示，這還需要其他幾種複雜的技術。 Baseten是一家出售GLM-5.2存取權限的公司。

O’Neill先生說: 「認為這些模型的所有功能都來自Anthropic的說法並不像人們所說的那樣真確」。

中國的人工智能初創公司能夠以遠低於其他國家的價格去提供開源模型，部分原因是該行業多年來一直受益於中國政府的支持，中國政府將人工智能視為經濟成長的關鍵引擎。

許多高階主管表示，美國公司不應開源其技術，因為這些技術可能被用於有害用途。但亦有專家則認為，如果美國監管機構扼殺開源技術，中國將獲得顯著優勢。

他們指出，由於中國生產了大多數性能卓越的開源系統，美國開發者將基於這些技術建立軟件。從長遠來看，這可能使中國成為人工智能發展的中心。

有些人認為，由於美國出口管制限制了訓練人工智能技術所需的專用電腦晶片的流通，中國系統將始終都落後於美國頂尖模型。 Z.ai 和其他中國新創公司花費數百萬美元取得中國境外數據庫的晶片的使用權。

Z.ai 在香港提交的文件顯示，該公司在 2025 年上半年的支出超過了其收入的七倍，而這些支出實際上主要用於支付此類計算服務的費用。

儘管如此，專家估計中國在人工智能領域僅落後美國公司六個月或更短的時間。

喬治·華盛頓大學 專業於新興技術與國際關係的助理教授Jeffrey Ding說:「先前一直有傳言說，出口管制最終會起到作用，美國尖端模型與中國模型之間的差距會進一步擴大，但GLM卻把情況推向相反的方向發展」。

隨著Fable和Mythos的退出，許多企業意識到擁有替代方案的重要性。

 OpenRouter的數據分析負責人Justin Summerville表示: “在信任依赖度上，大型企業有點忐忑不安” ， “誰會知道三週後顶级的模型是什麼樣呢？”

        So, recently, a Chinese start-up Z.ai has released an A.I. model that is nearly as powerful as Anthropic’s models. But Z.ai’s new technology costs much less to use. Powerful but inexpensive A.I. from China that is challenging the hurdles that OpenAI, Anthropic and Google have put on the industry. Apparently, in the field of A.I. development, the presence of China is obvious.

中國人工智能模型縮小與 Anthropic 和 OpenAI 的差距（1/2）

Recently The New York Times reported the following:

Chinese A.I. Models Close the Gap with Anthropic and OpenAI (1/2)

Silicon Valley engineers recently flocked to new technology from a Chinese company, Z.ai, that is almost as good as its American competitors but much cheaper.

The NYT - By Cade MetzKaren Weise and Meaghan Tobin - Cade Metz reported from San Francisco, Karen Weise from Seattle and Meaghan Tobin from Taipei, Taiwan.

June 25, 2026

Two weeks ago, the artificial intelligence company Anthropic shut down its two most powerful A.I. systems after an unexpected demand from the U.S. government to cut access to it.

Days later, a Chinese start-up, Z.ai, released an A.I. model that is nearly as powerful as Anthropic’s models, Fable and Mythos. But Z.ai’s new technology costs much less to use, and no one in the United States was putting restrictions on it. It quickly landed on a closely watched leaderboard of the world’s 10 most popular models.

Z.ai is on the cutting edge of a wave of powerful but inexpensive A.I. from China that is challenging the lock that OpenAI, Anthropic and Google have had on the industry. Six of the models now on the A.I. leaderboard were developed in China.

Z.ai’s new model, GLM-5.2, arrived just as U.S. businesses realized that they had to find ways to cut down on how much they were spending on A.I. It also landed when executives in Silicon Valley were becoming worried that the Trump administration was leaning toward regulating the technology.

“With Fable restricted, the gap between the U.S. and China is very slim,” said Rehaan Ahmad, a co-founder of the Silicon Valley start-up alphaXiv, who has been using Z.ai’s new model for more than a week.

The Chinese models still face two big hurdles to widespread use in the United States: concerns about their ties to the Chinese government and complaints that Chinese companies have unfairly used American technology to build these cheaper models. But their low cost is winning converts.

About 18 months ago, the Chinese start-up DeepSeek shocked Silicon Valley when it demonstrated that it could build effective A.I. far more affordably than many of its American counterparts. Z.ai is doing something similar. When performing certain tasks, GLM-5.2 costs about an eighth as much as Anthropic’s Claude Opus 4.8, which came out shortly before Fable and Mythos, according to OpenRouter, a start-up that runs the A.I. leaderboard.

Like most top-performing Chinese models, GLM-5.2 is open source software, which means anyone can use and modify it for free. That makes it much cheaper to use, even if it is not quite as powerful as what American companies have created.

“Do you need to drive a Ferrari everywhere?” asked Vivek Ramaswami, a start-up investor at Madrona Venture Group. “Probably not.”

GLM-5.2 is particularly good at generating computer code and powering A.I. agents, digital assistants that can use other software to perform tasks. Z.ai’s technology is now the third most widely used in the world for A.I. tasks, said Anastasios Angelopoulos, chief executive of ArenaAI, which tracks millions of A.I. users.

The largest cloud computing providers, including Microsoft and Amazon, already offer access to some systems from Z.ai, DeepSeek, MiniMax and other Chinese start-ups. Microsoft has also considered adding the latest DeepSeek model as an option to power one of its own products, which now runs on technology from Anthropic and OpenAI, two people familiar with the deliberations said, speaking on the condition of anonymity because they were not authorized to discuss them publicly.

The talks were reported earlier by Axios.

Microsoft, Anthropic and OpenAI declined to comment.

Some software developers are reluctant to use the A.I. system that Z.ai offers from computers in China, because they worry about sharing data with the company or with the Chinese government. They are also wary of China’s efforts to censor its A.I. systems or running afoul of U.S. export restrictions.

Z.ai was added to the Commerce Department’s trade blacklist in 2025. Corporate filings show that several of the company’s shareholders are controlled by a Chinese government agency that supervises the country’s defense industry.

(to be continued)

Translation

中國人工智能模型縮小與 Anthropic 和 OpenAI 的差距（1/2）

矽谷工程師最近紛紛湧向一家中國公司 Z.ai 的新技術，它幾乎與美國競爭對手不相伯仲，但價格卻便宜得多

兩週前，人工智能公司 Anthropic 在美國政府出人意料地要求限制其存取權限後，關閉了公司其兩個最強的人工智能系統。

幾天后，一家中國新創公司 Z.ai 發佈了一款人工智能模型，其功能幾乎與 Anthropic 的 Fable 和 Mythos 兩款模型相當。但 Z.ai 的新技術使用成本要低得多，而且美國方面也沒有對其施加任何限制。它迅速躋身備受關注的全球十大最受歡迎模型排行榜。

Z.ai 代表中國強大且低成本的人工智能浪潮，這股浪潮正在挑戰 OpenAI、Anthropic 和 Google 對產業的壟斷地位。目前人工智能排行榜上的六個模型均是由中國開發。

Z.ai 的新模型 GLM-5.2 的發佈正值美國企業意識到必須找到削減人工智能支出的方法之際。同時，矽谷的高層也開始擔憂特朗普政府可能傾向對這項技術進行監管。

矽谷新創公司 alphaXiv 的聯合創始人 Rehaan Ahmad 表示，他已經使用 Z.ai 的新模型超過一週了:「隨著 Fable 受到限制，美國和中國之間的差距已經非常小了」。

中國人工智能模型在美國的廣泛應用仍面臨兩大障礙：一是人們對其與中國政府關係的擔憂，二是有人指責中國公司不公平地利用美國技術來製造這些價格更低廉的模型。但其低成本優勢正令越來越多人把擔憂及指責放下。

大約18個月前，中國新創公司DeepSeek震驚了矽谷，該公司展示了能夠以遠低於許多美國同行的成本建立高效的人工智能。 Z.ai也在做類似的事情。根據營運人工智能排行榜的新創公司OpenRouter稱，在執行某些任務時，GLM-5.2的成本約為Anthropic公司Claude Opus 4.8的八分之一，後者發佈時間略早於Fable和Mythos。

與大多數性能卓越的中國模型一樣，GLM-5.2是開源軟件，這意味著任何人都可以免費使用和修改它。這使得它的使用成本更低，即使它的功能不如美國公司開發的產品般強大。

Madrona Venture Group的創業投資人Vivek Ramaswami問道: “你真的需要開著法拉利到處跑嗎？” ， “大概不需要吧。”

GLM-5.2尤其擅長產生電腦程式碼，並為人工智能代理（即能夠使用其他軟件執行任務的電子助理）提供支援。  ArenaAI在追蹤著數以百萬計的人工智能用戶，其執行長Anastasios Angelopoulos表示，Z.ai的技術目前是全球第三大最廣泛用於人工智能任務的技術。。

包括微軟和亞馬遜在內的大型雲端運算供應商已經提供對Z.ai、DeepSeek、MiniMax和其他中國新創公司部分系統的存取。兩位知情人士透露，微軟也考慮將最新的DeepSeek模型作為其自身產品的運行選項，該產品目前運行在Anthropic和OpenAI的技術之上。這兩位人士因未獲授權公開討論此事而要求匿名。

Axios先前報道了此次洽談。

微軟、Anthropic 和 OpenAI 均拒絕置評。

一些軟件開發人員不願使用 Z.ai 提供的、運行在中國境內的 AI 系統，因為他們擔心與該公司或中國政府共享數據。他們也擔心中國會審查其 AI 系統的企圖，或觸犯美國的出口限制。

Z.ai 已被列入美國商務部的貿易黑名單，生效日期為 2025 年。公司文件顯示，該公司的幾位股東受控於一家負責監管中國國防工業的中國政府機構。

（待續）

中國自2017年以來首次從美國手中奪得超級電腦王寶座 (2/2)

Recently The New York Times reported the following:

China Takes Supercomputer Crown from U.S. for First Time Since 2017 (2/2)

A supercomputer in Shenzhen was declared the world’s fastest. It uses only standard microprocessors and not the special-purpose chips called graphics processing units.

The NYT - By Don Clark - Reporting from San Francisco

June 23, 2026, 5:02 a.m. ET

(continue)

U.S. national labs, which are the main buyers of some of the largest supercomputers, are eager to use A.I. to accelerate aspects of their scientific work. So they are adopting more of these less precise calculations, along with the 64-bit variety.

Though U.S. companies have historically dominated the ranks of the very largest supercomputers, foreign systems have sometimes vaulted to the top. A system in Japan, for example, ranked No. 1 on the list from 2020 to 2022.

“There’s a lot of talk that America is the only country capable of these systems,” said Addison Snell, an analyst at Intersect360 Research, a firm tracking the sector. “Then you find that other companies have capabilities, too.”

Powerful systems from China and Japan have regularly spurred the Department of Energy and other U.S. agencies to push for more funding for supercomputers. In November, the Trump administration started the Genesis Mission, which aims to exploit supercomputers at U.S. national labs, along with private companies, to supercharge A.I. and scientific research.

GPUs, primarily developed by Nvidia and Advanced Micro Devices, have been a critical weapon in the recent supercomputer race. These chips excel in doing many chores simultaneously, including so-called vector calculations used in science and matrix multiplication at the heart of many A.I. tasks.

When U.S. officials limited China’s access to GPUs and other powerful chips, as well as restricting exports of some machines for manufacturing the most advanced semiconductors, that caused it “to invest in developing architectures and technology to effectively have supercomputers that are at the same level as the U.S.’s highest-performing systems,” Dr. Dongarra said.

China’s LineShine system does not separate the traditional jobs of microprocessors and GPUs, as most high-end systems do. Instead, it builds in GPU-style tasks with specialized circuitry that accelerates matrix and vector calculations. That ability is embedded in chips that have a total of nearly 14 million computing cores, or tiny electronic brains, installed in 90 hardware cabinets.

These chips are an original design based on a set of instructions licensed by Arm Holdings, a British company that is controlled by the Japanese conglomerate SoftBank. Arm’s technology is best known for powering smartphones but has lately been adapted by Nvidia, Amazon, Qualcomm and others for use in data centers.

Arm has long operated in China. “Arm operates globally, including in China, in compliance with applicable export control laws and regulations,” a company spokeswoman said.

LineShine’s designers, who are supercomputer veterans in China, have not disclosed details about which company manufactured the chips or the level of chip production technology used, Dr. Dongarra said.

He and other experts have long thought that China had systems capable of a No. 1 ranking, but laboratories there had not recently submitted test results.

“It doesn’t surprise me that there is a Chinese machine capable of being No. 1,” Mr. Snell said. “The surprise is that they wanted the acknowledgment.”

Dr. Dongarra, who wrote a detailed report on the new system, was told while visiting China that the system had been made without government funding, so the designers felt it was permissible to submit tests for the Top500 ranking, he said.

The Shenzhen scientists have also sought recognition for the new machine through 14 submissions for the Gordon Bell Prize, which promotes solving sophisticated problems in science, Dr. Dongarra said. Three systems are finalists for that award, and three for a related prize in climate science.

LineShine has been used for projects like a sophisticated simulation of Earth, including atmosphere, ocean, land and ice components, as well as a complex simulation of the human brain, according to Dr. Dongarra’s report.

Translation

中國自2017年以來首次從美國手中奪得超級電腦王寶座 (2/2)

深圳一台超級電腦被宣佈為世界上速度最快的超級電腦。它只使用標準微處理器，而不是被稱為圖形處理器（GPU）的專用晶片

（繼續）

美國國家實驗室是部分最大型超級電腦的主要買家，它們渴望利用人工智能來加速其研究工作。因此，它們正在採用更多精度較低的計算方法，亦有用64位元計算方法。

儘管美國公司歷來在最大型超級電腦領域佔據主導地位，但外國系統有時也能躍居榜首。例如，日本的一套系統在2020年至2022年期間一直位居榜首。

追踪該領域的公司Intersect360 Research的分析師Addison Snell表示: “很多人都說美國是唯一能夠擁有這類系統的國家”; “但後來你會發現，其他公司也具備這種能力。”

來自中國和日本的強大系統，不斷驅使美國能源部和其他機構爭取更多資金用於建立超級電腦。去年11月，特朗普政府啟動了「創世紀計劃」（Genesis Mission），旨在利用美國國家實驗室以及私人公司的超級電腦，大幅強化人工智能和科學研究。

GPU（圖形處理器），主要由英偉達（Nvidia）和AMD（超微半導體）公司開發，一直是近期超級電腦競賽的關鍵武器。這些晶片擅長同時處理多項任務，包括科學中使用的向量運算以及許多人工智能任務的核心 - 矩陣乘法。

Dongarra博士表示，當美國官員限制中國取得GPU和其他高性能晶片，並限制部分用於製造最先進半導體的設備出口時，中國「被迫投資開發架構和技術，以打造與美國最高性能係統性能相當的超級電腦」。

中國的LineShine系統不像大多數高階系統那樣，將微處理器和GPU的傳統功能分開。相反，它內建了GPU式任務，並配備了專門的電路來加速矩陣和向量計算。這種能力被嵌入到總共擁有近1400萬個運算核心（或稱為微型電子大腦）的晶片中，這些晶片安裝在90個硬體機櫃中。

這些晶片是基於Arm Holdings授權的一套指令而設計的。 Arm Holdings是一家英國公司，由日本軟銀集團控股。 Arm的技術最廣為人知的應用領域是智能型手機，但最近也被英偉達、亞馬遜、高通等公司應用於數據庫。

Arm公司長期以來一直在中國開展業務。 該公司一位發言人表示:「Arm在全球範圍內運營，包括在中國，並遵守所有適用的出口管制法律法規」。

Dongarra博士表示，LineShine的設計者是中國超級電腦領域的資深人士，他們沒有透露晶片的製造商或所使用的晶片生產技術水平等細節。

他和一些專家一直認為中國擁有能夠躋身世界第一的系統，但中國實驗室近期沒有提交測試結果。

Snell先生說: “中國有能夠排名第一的機器並不令我驚訝” ， “令人驚訝的是，他們竟然想要獲得認可。”

曾撰寫過一份關於這套新系統的詳細報告的Dongarra博士說，他在中國訪問期間得知，該系統並非由政府資助研發，因此設計者認為可以提交測試結果以參與Top500排名。

Dongarra博士表示深圳科學家也透過14項申請角逐Gordon Bell獎，力求讓這台新型機器獲得認可。該獎項旨在表彰解決複雜科學難題的傑出成果。目前已有三套系統入圍該獎項的最終候選名單，另有三套系統入圍氣候科學領域的相關獎項。

Dongarra博士的報告指出，LineShine已被用於多個項目，例如對地球進行複雜的模擬包括大氣、海洋、陸地和冰蓋，以及對人腦進行複雜的模擬。

So, China takes back the computing crown from the United States. It is the LineShine, a massive computing system declared to be the fastest in the world. Besides raw speed, the system stands out because it uses only standard microprocessors and not the special-purpose chips called GPU. Apparently, China can use ordinary computer chips to make high speed computers at a low cost.

中國自2017年以來首次從美國手中奪得超級電腦王寶座（1/2）

Recently The New York Times reported the following:

China Takes Supercomputer Crown From U.S. for First Time Since 2017 (1/2)

A supercomputer in Shenzhen was declared the world’s fastest. It uses only standard microprocessors and not the special-purpose chips called graphics processing units.

The NYT - By Don Clark - Reporting from San Francisco

June 23, 2026, 5:02 a.m. ET

China took back a coveted computing crown from the United States on Tuesday, ratcheting up a fierce technological competition that has implications for science, national security and geopolitics.

LineShine, a massive computing system in Shenzhen, China, was declared the world’s fastest by a group of researchers using a set of standard tests for supercomputers. Besides raw speed, the system stood out because it uses only standard microprocessors and not the special-purpose chips called graphics processing units, which most high-end supercomputers rely on for heavy number crunching.

That underlying design could point to a better way to blend artificial intelligence with traditional scientific tasks, said Jack Dongarra, an organizer of the so-called Top500 list of the world’s most powerful supercomputers.

Dr. Dongarra, a professor of computer science and electrical engineering at the University of Tennessee, recently inspected the new machine, at the Shenzhen Cloud Computing Center. LineShine’s test results were more than 20 percent faster than those of El Capitan, a system at Lawrence Livermore National Laboratory in California that has topped a twice-yearly ranking of supercomputer performance since November 2024. China had not placed a machine at the top of the list since 2017.

“It’s an impressive system,” Dr. Dongarra said of LineShine. “They upped us by developing a system that is not reliant on GPUs.”

The new supercomputer adds to the race between China and the United States for technological supremacy. U.S. tech giants like OpenAI, Anthropic and Google have developed leading A.I. models, while another American company, Nvidia, has become the world’s dominant supplier of A.I. chips. China has tried to innovate in different ways, with the Chinese start-up DeepSeek releasing a cutting-edge A.I. model last year using just a tiny fraction of specialized A.I. chips.

To prevent China from catching up, President Trump has imposed tariffs and at times placed limits on A.I. chip exports. But China’s use of standard microprocessors, which are known as CPUs, rather than GPUs to create an ultrafast supercomputer suggests a potential way to get around those roadblocks.

“The U.S. government should have stronger controls on the export and manufacturing of CPUs for the China market,” said Jimmy Goodrich, a senior fellow at the University of California Institute on Global Conflict and Cooperation. “It is a loophole in the current regulations.”

Supercomputers, a term for the largest machines dedicated to science, have been used since the 1960s for tasks like creating climate models, cracking codes and designing nuclear weapons. They typically use high-precision mathematics, expressing numbers with 64 bits of data.

Commercial A.I. systems from companies like Google and OpenAI, by contrast, can be even faster. They can use approximations for tasks such as identifying images or selecting the next word in a sentence, relying on what are known as four-bit and eight-bit numbers that allow the systems to make many simpler calculations at once.

“It is notable and impressive what China has done here, but they can’t hold a candle to these massive A.I. supercomputers that have been built by American A.I. labs” and others, Mr. Goodrich said.

(to be continued)

Translation

中國自2017年以來首次從美國手中奪得超級電腦王寶座（1/2）

深圳一台超級電腦被宣佈為世界上最快的超級電腦。它只使用標準微處理器，而不是被稱為圖形處理器（GPU）的專用晶片

週二，中國從美國手中奪回了夢寐以求的超級電腦之王的寶座，加劇了激烈的技術競爭，這場競爭對科學、國家安全和地緣政治都具有重要意義。

位於中國深圳的巨型計算系統「天翔」（LineShine）被一組研究人員使用一系列超級電腦標準測試方法宣佈為世界上最快的超級電腦。除了極快的速度外，該系統脫穎而出的另一個原因是它僅使用標準微處理器，而非大多數高階超級電腦用於繁重運算的專用晶片 - 圖形處理器（GPU）。

全球最強超級電腦Top500名單的組織者之一Jack Dongarra表示，這種底層設計或許能為人工智能與傳統科學任務的融合提供更好的途徑。

Dongarra博士是田納西大學電腦科學與電子工程系的教授，他最近在深圳雲端運算中心考察了這台新機器。 LineShine的測試結果比加州Lawrence Livermore國家實驗室的El Capitan系統快20%以上。 El Capitan自2024年11月以來一直穩居全球超級電腦效能榜首，該名單每半年更新一次。自2017年以來，中國就再也沒有超級電腦位居榜首。

Dongarra博士評價LineShine時說道: “這是一套令人印象深刻的系統” ， “他們開發了一套不依賴GPU的系統，超越了我們。”

這台新的超級電腦加劇了中美兩國在科技領域的競爭。美國科技巨擘如OpenAI、Anthropic和Google已開發出領先的人工智能模型，而另一家美國公司英偉達則成為全球人工智能晶片的主要供應商。中國則嘗試以不同的方式進行創新，例如，中國新創公司DeepSeek去年發佈了一款尖端人工智能模型，該模型僅使用了極少量的專用人工智能晶片。

為了阻止中國迎頭趕上，特朗普總統曾對人工智能晶片加徵關稅，並一度限制其出口。但中國使用標準微處理器（即CPU）而非GPU來製造超高速超級電腦，這表明中國可能找到了繞過這些障礙的方法。

加州大學全球衝突與合作研究所 高級研究員吉Jimmy Goodrich表示:“美國政府應該加強對中國市場CPU出口和製造的管控” ;“這是現有法規中的一個漏洞。”

超級電腦，即專門用於科學研究的最大型機器，自 20 世紀 60 年代以來就被用於創建氣候模型、破解密碼和設計核武器等任務。它們通常使用高精度數學運算，以 64 位元資料表示數字。

相比之下，Google和 OpenAI 等公司的商業人工智能系統速度更快。它們可以使用近似值來完成諸如識別圖像或選擇句子中的下一個單字等任務，依賴所謂的 4 位和 8 位數字，這使得系統能夠即時進行許多較簡單的計算。

Goodrich先生說: 「中國在這方面取得的成就令人矚目，但與美國人工智能實驗室和其他國家建造的這些巨型人工智能超級電腦相比，它們根本無法相提並論」。

 （待續）