[新聞] 人形機器人產業準備好它的ChatGPT時刻嗎?

看板Stock (股票)作者TyuzuChou (潤潤潤)時間4小時前 (2025/09/15 14:26)推噓9(9推 0噓 4→)

留言13則, 11人參與討論串1/1

原文標題：Is the humanoid robot industry ready for its ChatGPT moment? 原文連結：https://tinyurl.com/6239np28 發布時間：Published Sun, Sep 14 2025 10:21 PM EDT 記者署名：Dylan Butts Victoria Yeo 原文內容： Humanoid robots, which have made significant technological advances this year, may be at the precipice of a ChatGPT-like spike in investment and popularity — or at least, that’s what many in the industry believe. 今年，人形機器人取得了重大的技術進展，或許正站在如同 ChatGPT 一般爆發式投資與人氣高漲的臨界點——至少產業內許多人是這樣認為的。 So-called humanoid robots are artificial intelligence-powered machines designed to resemble humans in appearance and movement, with expected use cases across the industrial and service sectors. 所謂的人形機器人，是由人工智慧驅動的機械裝置，設計上模仿人類的外觀與動作，預期可應用於工業與服務領域。 Makers of these robots have been working on the technology in the background for years. Now, they say they’re ready to unleash the technology into the world. 這些機器人的製造商已經默默研發多年。如今，他們宣稱已經準備好將這項技術推向全球。 “There is a consensus in our industry that the ChatGPT moment for humanoid robots has arrived,” Xiong Youjun, general manager at the Beijing Innovation Center for Humanoid Robotics, said during a panel in Singapore on Thursday, alongside other professionals from China’s robotics industry. 「我們產業內普遍認為，人形機器人的 ChatGPT 時刻已經到來。」北京人形機器人創新中心總經理熊有軍週四在新加坡的一場論壇上，與其他中國機器人專業人士同台時表示。 “This year has been defined as the first year of mass production of humanoid robots,” Xiong, chief technology officer and executive director of robotics firm UBTech, said in Mandarin translated by CNBC. He added that there had been rapid progress in both the mechanical bodies and the AI-powered “brains.” 「今年被定義為人形機器人大規模量產的元年。」同時身兼優必選科技（UBTech）首席技術長與執行董事的熊有軍，用中文（由 CNBC 翻譯）補充說，在機械本體與 AI 驅動的「大腦」方面都有快速進展。 The original “ChatGPT moment” occurred in late 2022, when OpenAI released its groundbreaking generative AI chatbot to the public, leading to mass adoption of large language models and widespread recognition of their potential. 所謂「ChatGPT 時刻」最初發生在 2022 年底，當時 OpenAI 向公眾推出突破性的生成式 AI 聊天機器人，促使大型語言模型被廣泛採用並獲得普遍認可。 Robotics players hoping to recreate that impact include Tesla’s Optimus. Meanwhile, a growing number of humanoid robot start-ups are emerging in China, with companies like Unitree, Galbot, Agibot and UBtech Robotics bringing products to market. 試圖複製這種影響力的業者包括特斯拉的 Optimus。同時，中國湧現越來越多的人形機器人新創企業，如宇樹（Unitree）、高博特（Galbot）、Agibot 與優必選（UBTech Robotics），並已開始將產品推向市場。 While humanoid robots are yet to reach a fraction of the adoption seen with generative AI, many experts do expect the technology to have a transformative impact on the global economy in a matter of years. 雖然人形機器人的普及率遠不及生成式 AI，但許多專家預期，這項技術在幾年內就能對全球經濟產生變革性的影響。 Meanwhile, robots have begun to appear everywhere, from factories to technology conferences and sporting events. 與此同時，從工廠到科技會議甚至體育賽事，機器人已經開始隨處可見。 Humanoids pick up steam 人形機器人加速推進 Tesla CEO Elon Musk has said he expects the company to produce 5,000 of its Optimus robots this year, with the technology expected to eventually make up the majority of the EV maker’s business. 特斯拉執行長伊隆·馬斯克表示，公司今年預計將生產 5,000 台 Optimus 機器人，且這項技術最終將成為這家電動車製造商業務的主要部分。 Meanwhile, humanoid robot firms in China say their products are already being used in factories and for commercial services. 與此同時，中國的人形機器人公司則表示，他們的產品已經投入工廠與商業服務使用。 Speaking on Thursday, Zhao Yuli, chief strategy officer at Galbot, said the start-up had already deployed almost 1,000 robots across different businesses. 高博特首席策略官趙雨力週四表示，該新創已在不同企業部署了近 1,000 台機器人。 Other companies, such as UBTech Robotics and Galbot, have also installed robots in local factories, according to local media reports. 據當地媒體報導，包括優必選與高博特在內的公司，也已在本地工廠安裝機器人。 According to Zhao, these deployments have come alongside a surge of investor interest and government support in the sector, as well as the maturation of both robotics and generative AI technology. 趙雨力指出，這些部署伴隨著投資人興趣的湧現與政府的支持，同時也得益於機器人技術與生成式 AI 的成熟。 Industry experts noted that this maturation in technology has been on display at a number of conferences and events this year, such as China’s World Humanoid Robotics Game, which sees robots compete in practical scenarios. 產業專家指出，技術的成熟已在今年的多場會議與活動中展現，例如中國的「世界人形機器人大賽」，讓機器人在實際場景中競技。 Galbot won a gold medal in the Robot Skills event after placing first in a pharmaceutical sorting challenge. 高博特在機器人技能賽事中獲得金牌，並在藥品分揀挑戰賽中奪得第一。 Improvements in Chinese humanoid robots’ motion control have also been on display in recent months at sporting events such as marathons and boxing matches. 近幾個月，中國人形機器人在動作控制上的提升，也在馬拉松、拳擊賽等體育活動中展現。 Guo Yandong, founder and CEO of AI2 Robotics, added that improvements in generative AI have also enabled robots to learn on the job rather than rely solely on preset commands, a shift that could expand the uses of humanoids across sectors. AI2 Robotics 創辦人兼執行長郭彥東補充說，生成式 AI 的進步也使機器人能在工作中學習，而非僅依賴預設指令，這一轉變可望拓展人形機器人在各行各業的應用。 Not so fast 別太急 Despite the hype from humanoid robotics companies, however, many experts resist the idea that mass public adoption will occur anytime soon. 然而，儘管人形機器人公司大肆宣傳，許多專家仍不認為大眾會在短期內大量採用這項技術。 “Humanoids won’t arrive all at once in a ChatGPT moment, but slowly enter more and more positions as their capabilities increase,” said Reyk Knuhtsen, analyst at SemiAnalysis, an independent research and analysis company specializing in semiconductors and AI. He added that their first uses will be in low-stakes, failure-tolerant tasks. 「人形機器人不會像 ChatGPT 那樣突然全面到來，而是會隨著能力提升逐步進入更多崗位。」專精半導體與 AI 的獨立研究公司 SemiAnalysis 分析師 Reyk Knuhtsen 說。他補充，人形機器人最初的應用會是在低風險、容錯率高的任務中。 That’s not to mention long manufacturing timelines and high costs, which will also slow adoption compared to generative AI, he added. 他還指出，漫長的製造週期與高昂成本，也會使其普及速度比生成式 AI 更慢。 Even UBTech’s Xiong conceded that some hurdles remain for the sector, such as ethical considerations, laws and regulations that need to be addressed. 就連優必選的熊有軍也承認，這個產業仍面臨一些障礙，例如必須解決的倫理問題、法律與監管挑戰。 Still, analyst Knuhtsen expects investment in the space to continue as long as the autonomy of the robots continues to improve. 儘管如此，Knuhtsen 仍預期，只要機器人的自主性持續提升，這一領域的投資將會延續。 “The market opportunity for humanoids is enormous, contingent on how well the AI performs ... If the technology works, it has the chance to transform many labor processes around the world,” he said. 「人形機器人的市場機會非常巨大，關鍵在於 AI 的表現……如果技術奏效，它將有機會改變全球許多勞動流程。」他說。 Merrill Lynch analysts recently estimated in a research note that global humanoid robot shipments will reach 18,000 units in 2025 from 2,500 units last year. It also predicts a global robot “population” of 3 billion by 2060. 美林分析師近期在研究報告中估計，全球人形機器人出貨量將從去年的 2,500 台增至 2025 年的 18,000 台，並預測到 2060 年全球「機器人族群」將達到 30 億。心得/評論：支持者（產業內人士，如熊有軍、趙雨力、馬斯克）立場：人形機器人產業已經迎來「ChatGPT 時刻」，技術成熟、量產開始，應用正快速擴展。理由：機械本體與 AI 大腦同步進步。已經在工廠與商業服務中投入使用。投資人與政府支持明顯增加。有如特斯拉、優必選等大公司帶頭，還有中國新創企業加速推進。目標：塑造產業熱潮，吸引更多投資與市場關注。質疑者（分析師、部分專家）立場：不會像 ChatGPT 一樣短期爆發，大規模普及還需很久。理由：高成本與長製造週期。初期應用侷限於低風險、容錯性高的工作。存在倫理、法律與監管障礙。觀點：人形機器人的成長會是「漸進式」而非「爆發式」。共識雖然對「速度」有分歧，但雙方都承認市場規模巨大，若 AI 自主性與實用性真正落地，對勞動力與經濟的改變將是深遠的。 -- ※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 150.117.197.145 (臺灣) ※ 文章網址: https://www.ptt.cc/bbs/Stock/M.1757917614.A.8F8.html