[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-embodied-ai-capital-flows-to-brains-and-bubbles-zh":3,"article-related-embodied-ai-capital-flows-to-brains-and-bubbles-zh":30,"series-industry-d1c1656d-c842-49c2-a777-19ce87ab0dc8":72},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":22,"views":26,"created_at":27,"published_at":28,"topic_cluster_id":29},"d1c1656d-c842-49c2-a777-19ce87ab0dc8","embodied-ai-capital-flows-to-brains-and-bubbles-zh","具身智能的钱正在往大脑集中，泡沫也在集中","\u003Cp data-speakable=\"summary\">具身智能的資本正在往大腦、世界模型和數據基礎設施集中。\u003C\u002Fp>\u003Cp>我認為具身智能當前最值得押注的不是本體，而是大腦與數據基礎設施。量子位統計顯示，2026年上半年國內具身智能賽道融資約438億元，超過一半流向「大腦派」公司；同期本體派只占12.8%，甚至低於零部件公司。這不是短期噪音，\u003Ca href=\"\u002Fnews\u002Fbaya-openchip-bet-ai-silicon-data-movement-zh\">而是資\u003C\u002Fa>本對行業價值分配的重新定價：硬體決定能不能做，大腦決定能不能規模化複製。\u003C\u002Fp>\u003Ch2>第一個論點：資本已經把「上限」押在大腦上了\u003C\u002Fh2>\u003Cp>投資人那句「本體兜住下限，大腦決定上限」，不是口號，是資金流向的真實寫照。Pre-A輪平均7億元，B輪平均22.5億元，這種量級放在多數行業裡已經接近C輪、D輪。它石智航Pre-A輪拿到4.55億美元，千尋智能在2月至6月連融四輪、累計近50億元，說明市場不再按傳統里程碑定價，而是按「誰更像未來的\u003Ca href=\"\u002Fnews\u002Fcoordex-humanoid-loco-manipulation-priors-zh\">機器人\u003C\u002Fa>作業系統」定價。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782207175134-576j.png\" alt=\"具身智能的钱正在往大脑集中，泡沫也在集中\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>背後邏輯也很直接。機器人本體當然重要，但它更像製造業問題，考驗供應鏈、成本和量產；大腦則更像AI問題，邊際成本低、複用性強、遷移性高。中國在製造端有天然優勢，意味著本體會快速進化，但真正能拉開估值差距的，還是誰先把感知、推理和行動串成閉環。資本追逐的是可複製的能力，而不是一台機器的單點性能。\u003C\u002Fp>\u003Ch2>第二個論點：世界模型正在取代VLA成為新敘事中心\u003C\u002Fh2>\u003Cp>今年上半年有融資動態的35家「大腦派」公司裡，27家在研發世界模型，占比接近八成，這個數字說明賽道共識已經發生遷移。2024年大家還在比誰的VLA更強、誰的真機數據更多；到了今天，不做世界模型似乎就像沒跟上時代。它不只是技術路線變化，更是融資語言變化，因為「世界模型」比「VLA」更能承接想像空間，也更容易讓投資人相信公司在做下一代平台。\u003C\u002Fp>\u003Cp>但真正重要的不是名詞，而是底層實現。蒋子元的判斷很有代表性：很多所謂世界模型，本質上只是把影片生成模型作為backbone，而VLA更多依賴語言模型作為backbone。換句話說，路線之爭沒有外界想得那麼硬，創業公司會不斷切換更順手的架構。就像影片生成從U-Net轉向DiT，行業不會為一種範式守身如玉，誰能更快把效果做出來，誰就會成為「正確路線」。\u003C\u002Fp>\u003Ch2>第三個論點：決定勝負的不是路線潔癖，而是數據基礎設施\u003C\u002Fh2>\u003Cp>自變量機器人算法負責人甘如饴說得更到位：比起世界模型還是VLA，底層的數據基礎設施才是核心競爭力。這套基礎設施覆蓋數采、訓練和評測全流程，目標不是做一次漂亮演示，而是形成可規模化運轉的工業級體系。這個判斷比任何路線宣言都更接近現實，因為機器人真正難的地方，從來不是寫出一個demo，而是\u003Ca href=\"\u002Fnews\u002Flifescibench-tests-biotech-models-zh\">讓模型\u003C\u002Fa>在不同場景裡持續穩定地變強。\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782207172429-q1fi.png\" alt=\"具身智能的钱正在往大脑集中，泡沫也在集中\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>這也是為什麼「能不能快速趕上」不取決於今天站在哪條路線，而取決於有沒有把數據管道、評測體系和訓練閉環先搭好。技術架構會變，今天的世界模型明天可能被更新範式替代，但數據資產不會憑空出現。誰先把真實世界的數據採集、標註、回放、訓練和評測做成流水線，誰就擁有跨架構遷移的能力。對創業公司來說，這比押注某個名詞更重要。\u003C\u002Fp>\u003Ch2>反方可能怎麼說\u003C\u002Fh2>\u003Cp>最強的反對意見很簡單：具身智能現在估值太高、融資太快、路線太散，泡沫遲早會破。這個擔心不是空穴來風。行業自己也承認，90%以上的公司可能會消失，而且泛化性如果不能兌現，真正可落地的場景會非常有限。與大模型相比，具身智能的商業化確定性更低，今天砸進去的錢，確實有相當一部分可能打水漂。\u003C\u002Fp>\u003Cp>但我不接受「因此現在不該投」的結論。早期產業發展需要泡沫來聚集資本、人才和注意力，這一點在大模型時代已經被驗證過。不同的是，具身智能還沒收斂到只剩幾條清晰賽道，所以最優策略不是回避，而是承認它是高波動、高淘汰率的窗口期。問題不在於有沒有泡沫，而在於你是否把泡沫當成確定性，把故事當成結果。\u003C\u002Fp>\u003Ch2>你能做什麼\u003C\u002Fh2>\u003Cp>如果你是工程師，就別被「世界模型」三個字綁架，先把數據管道、評測集和閉環訓練做扎實；如果你是PM，就盯住可複用場景和失敗邊界，不要把演示當產品；如果你是創辦人，就把融資當作搶時間而不是講故事，盡快建立能跨架構遷移的數據資產和人才密度。這個賽道最後贏的，不是最會喊口號的人，而是最快把不確定性工程化的人。\u003C\u002Fp>","我認為具身智能現在最值得押注的不是本體，而是大腦、世界模型與數據基礎設施；資本正在往這裡集中，泡沫也一起集中。","zhuanlan.zhihu.com","https:\u002F\u002Fzhuanlan.zhihu.com\u002Fp\u002F2052091107127906984",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782207175134-576j.png","industry","zh","c996e251-6cca-452c-b87d-a88a6c9ada28",[17,18,19,20,21],"具身智能","世界模型","大腦派","數據基礎設施","融資泡沫",[23,24,25],"資本正在從本體轉向大腦與世界模型，這是行業定價邏輯的重排。","真正的護城河不是某個模型名詞，而是數據基礎設施與可複用閉環。","泡沫確實存在，但早期賽道的關鍵不是躲開泡沫，而是辨識誰能把泡沫轉成能力。",0,"2026-06-23T09:32:27.187862+00:00","2026-06-23T09:32:27.179+00:00","caa87b65-9bbc-46fe-bba8-4f4158dd2d8b",{"tags":31,"relatedLang":32,"relatedPosts":35},[],{"id":15,"slug":33,"title":6,"language":34},"embodied-ai-capital-flows-to-brains-and-bubbles-en","en",[36,42,48,54,60,66],{"id":37,"slug":38,"title":39,"cover_image":40,"image_url":40,"created_at":41,"category":13},"11467534-dfd5-4ba4-8816-15c6421e0263","microsoft-june-2026-partner-center-changes-zh","6 月 Partner Center 變更先看這 5 項","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782216176004-w0p7.png","2026-06-23T12:02:29.913535+00:00",{"id":43,"slug":44,"title":45,"cover_image":46,"image_url":46,"created_at":47,"category":13},"4c9747cd-4eb2-4e08-ab87-207690b85526","sia-openai-partner-network-enterprise-ai-services-game-zh","Sia 與 OpenAI 結盟，證明企業 AI 已經變成服務戰","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782208977450-zj5e.png","2026-06-23T10:02:23.060282+00:00",{"id":49,"slug":50,"title":51,"cover_image":52,"image_url":52,"created_at":53,"category":13},"2166c4ec-e128-4281-931e-30b896b6c0ca","baya-openchip-bet-ai-silicon-data-movement-zh","Baya 與 Openchip 押注 AI 晶片未來：關鍵不是算力，而是資料搬運","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782193668588-hvs5.png","2026-06-23T05:47:23.424804+00:00",{"id":55,"slug":56,"title":57,"cover_image":58,"image_url":58,"created_at":59,"category":13},"031360c4-4d0a-46bf-bbd1-ec2db9392537","citigroup-sees-tokenized-assets-hitting-8-2t-zh","Citi估8.2兆美元資產將被代幣化","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782185577271-2wur.png","2026-06-23T03:32:33.388896+00:00",{"id":61,"slug":62,"title":63,"cover_image":64,"image_url":64,"created_at":65,"category":13},"9bff47f9-faba-408a-b0f9-83d050ed3ab6","rwa-tokenization-turns-assets-into-on-chain-rails-zh","RWA 代幣化把資產變上鏈通道","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782184696459-t7qe.png","2026-06-23T03:17:50.080174+00:00",{"id":67,"slug":68,"title":69,"cover_image":70,"image_url":70,"created_at":71,"category":13},"0bcd7d43-04e9-455e-b0c0-a9d9679a50a2","micron-anthropic-deal-earnings-signal-zh","Micron 合約把財報變訊號","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1782183794288-6y4p.png","2026-06-23T03:02:49.715904+00:00",[73,78,83,88,93,98,103,108,113,118],{"id":74,"slug":75,"title":76,"created_at":77},"ee073da7-28b3-4752-a319-5a501459fb87","ai-in-2026-what-actually-matters-now-zh","2026 AI 真正重要的事","2026-03-26T07:09:12.008134+00:00",{"id":79,"slug":80,"title":81,"created_at":82},"83bd1795-8548-44c9-9a7e-de50a0923f71","trump-ai-framework-power-speech-state-preemption-zh","川普 AI 框架瞄準電力、言論與州權","2026-03-26T07:12:18.695466+00:00",{"id":84,"slug":85,"title":86,"created_at":87},"ea6be18b-c903-4e54-97b7-5f7447a612e0","nvidia-gtc-2026-big-ai-announcements-zh","NVIDIA GTC 2026 重點拆解","2026-03-26T07:14:26.62638+00:00",{"id":89,"slug":90,"title":91,"created_at":92},"4bcec76f-4c36-4daa-909f-54cd702f7c93","claude-users-spreading-out-and-getting-better-zh","Claude 用戶更分散，也更會用","2026-03-26T07:22:52.325888+00:00",{"id":94,"slug":95,"title":96,"created_at":97},"bd903b15-2473-4178-9789-b7557816e535","openclaw-raises-hard-question-for-ai-models-zh","OpenClaw 逼問 AI 模型價值","2026-03-26T07:24:54.707486+00:00",{"id":99,"slug":100,"title":101,"created_at":102},"eeac6b9e-ad9d-4831-8eec-8bba3f9bca6a","gap-google-gemini-checkout-fashion-search-zh","Gap 把結帳搬進 Gemini","2026-03-26T07:28:23.937768+00:00",{"id":104,"slug":105,"title":106,"created_at":107},"0740e53f-605d-4d57-8601-c10beb126f3c","google-pushes-gemini-transition-to-march-2026-zh","Google 把 Gemini 轉換延到 2026 年 3…","2026-03-26T07:30:12.825269+00:00",{"id":109,"slug":110,"title":111,"created_at":112},"e660d801-2421-4529-8fa9-86b82b066990","metas-llama-4-benchmark-scandal-gets-worse-zh","Meta Llama 4 分數風波又擴大","2026-03-26T07:34:21.156421+00:00",{"id":114,"slug":115,"title":116,"created_at":117},"183f9e7c-e143-40bb-a6d5-67ba84a3a8bc","accenture-mistral-ai-sovereign-enterprise-deal-zh","Accenture 攜手 Mistral AI 賣主權 AI","2026-03-26T07:38:14.818906+00:00",{"id":119,"slug":120,"title":121,"created_at":122},"191d9b1b-768a-478c-978c-dd7431a38149","mistral-ai-faces-its-hardest-year-yet-zh","Mistral AI 迎來最硬的一年","2026-03-26T07:40:23.716374+00:00"]