Concise Summary简洁概述
After two weeks meeting China's AI founders, VCs, and listed-company executives, the author's priors flip: he leaves more bullish on hardware and more bearish on software than when he arrived.
The sharpest surprise is that the founder archetype he relies on to spot great teams in the West — spiky, contrarian, obsessive — doesn't map cleanly onto the founders he met in China.
走访中国 AI 创始人、VC 和上市公司高管两周后，作者的先验判断发生反转：离开时他对硬件更看好，对软件更看衰，与出发前的预期完全相反。
最让他意外的是，他在西方用来识别伟大创始人的那套“特质模型”——特立独行、爱较真、近乎偏执——在中国创始人身上并不适用。
Infographic信息图
The founder-genotype heuristic
“创始人基因”启发式
The author's proven filter for great founders is spikiness: independent thinking, rebelliousness, obsessive focus, decisions that look crazy from outside but are internally coherent. This heuristic, honed investing in Western/crypto founders, is his default lens walking into China.
作者筛选伟大创始人的经验公式是“锋芒”：独立思考、反骨、极致执着，做出旁人看不懂但自己逻辑自洽的决定。这套在西方与加密圈投资中验证过的直觉，是他带着走进中国生态圈的默认标尺。
Hardware surprise
硬件超预期
Contrary to the Western narrative of chip-export chokepoints crippling Chinese AI, the author comes away more convinced of China's hardware capability and resilience after seeing it firsthand.
与西方“芯片出口管制卡住中国 AI 脖子”的主流叙事相反，作者实地考察后反而更相信中国硬件的能力与韧性。
Software skepticism
软件端存疑
On the software/application layer, the trip leaves him more bearish — implying overcrowding, thin differentiation, or a gap between hype and defensible product, though the piece is truncated before the specifics.
在软件与应用层，这趟行程反而让他更看衰——暗示赛道拥挤、差异化不足，或炒作与真实产品壁垒之间存在落差，可惜原文在展开细节前被截断。
A different founder type
另一种创始人类型
The founders he met don't fit the rebellious-outsider mold; they represent an alternative path to building great companies, forcing him to question whether his pattern-matching is culturally bound rather than universal.
他遇到的创始人并不符合“叛逆局外人”的模板，代表着另一条通往优秀公司的路径，这迫使他反思自己那套识人直觉究竟是普适规律，还是被西方语境限定的偏见。
Detailed Summary详细解读
The piece opens by establishing the author's credibility as a pattern-matcher: as a partner deploying capital across hundreds of portfolio companies, he has a battle-tested filter for identifying exceptional founders. That filter is built on legible traits — independence of thought, a refusal to defer to convention, decisions that read as irrational to outsiders but follow an internal logic, and an obsessive, almost monomaniacal history in their domain. This is not abstract theorizing; it's the operational heuristic a high-volume investor uses daily to separate signal from noise in a sea of high-IQ founders.
The core narrative move is a two-week immersion trip that inverts his priors along two separate axes. First, an overall-ecosystem axis: he arrives bullish (expecting cheap access to world-class AI talent) and leaves with a more 'complex and nuanced' view — signaling the piece will resist a simple bull/bear verdict. Second, a hardware-vs-software axis: he becomes more bullish on hardware and more bearish on software, a split that runs directly against the popular Western narrative that chip export controls are the decisive constraint on China's AI trajectory.
The essay then pivots to its most distinctive claim: founders in China frequently don't match the 'spiky' archetype he relies on. This matters methodologically, not just anecdotally — if his heuristic was calibrated on a Western/crypto sample, its failure to generalize to China raises the question of whether great-founder traits are culturally contingent rather than universal laws of company-building, and whether VCs risk systematically underrating founders who succeed through different psychological wiring (e.g., discipline, execution consistency, or state/ecosystem-aligned ambition rather than lone-wolf rebellion).
Structurally, the piece is built as a three-act reveal — founder question, hardware surprise, software bearishness — each act functioning as a separate 'update' to the author's model of China AI. The truncated portion (per the summary) promises a fourth beat: an 'ignored asymmetry,' likely referring to a structural gap between how fast hardware capability compounds versus how fast software/application moats erode or fail to form in China's hyper-competitive, fast-copying startup environment.
What makes this more than a travelogue is the self-implicating frame: the author isn't just reporting what China's ecosystem looks like, he's reporting where his own investing instincts broke. That reflexivity — 'my model failed to predict this' — is rarer and more valuable than a straightforward bull or bear case, because it signals genuine updating rather than confirmation of an ideology he brought with him.
文章开篇先建立作者的“识人”可信度：作为一家管理超过 200 个投资组合项目的机构合伙人，他有一套久经考验的筛选伟大创始人的标准。这套标准建立在几个可辨识的特质上——独立思考、拒绝盲从惯例、做出旁人看似不理性但内在逻辑自洽的决定，以及在自己领域近乎偏执的钻研史。这不是空谈理论，而是一位高频投资人每天用来在茫茫高智商创始人中辨别真正信号的实操工具。
文章的核心叙事转折来自这趟两周的沉浸式考察，它在两条轴线上颠覆了作者的先验判断。第一条是整体生态轴线：出发前看好（期待以低价接触世界级 AI 人才），归来后转为“更复杂、更微妙”的判断——预示全文不会给出简单的看多或看空结论。第二条是硬件对软件轴线：作者变得更看好硬件、更看衰软件，这一判断与西方主流叙事——即芯片出口管制是制约中国 AI 发展的决定性因素——直接相悖。
随后文章转向最独特的论点：中国创始人普遍不符合他所依赖的“锋芒型”原型。这一点在方法论层面而非仅是轶事层面很重要——如果这套启发式是在西方与加密圈样本上校准出来的，它在中国失灵就提出一个更深的问题：伟大创始人的特质究竟是文化情境下的产物，还是放之四海皆准的创业铁律？VC 是否正在系统性低估那些依靠纪律、执行一致性或与生态/体制协同的野心（而非孤狼式反骨）取得成功的创始人？
从结构上看，全文按“创始人之谜—硬件超预期—软件看衰”三段式推进，每一段都是作者对中国 AI 认知模型的一次独立更新。根据摘要，被截断的部分还应有第四层递进：一个“被忽视的不对称性”，很可能指的是硬件能力复合增长的速度，与软件/应用层护城河在中国高度内卷、快速复制的创业环境中难以形成或迅速被侵蚀之间的结构性落差。
这篇文章之所以超越普通游记，关键在于它的自我反思框架：作者不只是在报告中国生态圈是什么样，更是在报告自己的投资直觉在哪里失灵了。这种“我的模型没能预测到这一点”的反身性表达，比单纯的看多或看空更为稀缺、也更有价值，因为它意味着真实的认知更新，而不是对出发前既有立场的自我确认。
FAQ常见问答
What's the author's actual conclusion — bullish or bearish on China AI overall?作者对中国 AI 整体到底是看多还是看空？
Neither cleanly — he calls his post-trip view 'more complex and nuanced,' splitting it: more bullish on hardware, more bearish on software than his pre-trip expectations.
两者都不是简单的答案——他把行程后的判断称为“更复杂微妙”，拆成两条：硬件比预期更看好，软件比预期更看衰。
Why does the founder archetype matter if the piece is supposedly about hardware and software?文章名义上谈硬件软件，为什么创始人特质这条线反而更重要？
Because it's the mechanism behind his other judgments — if his founder-quality filter is miscalibrated for China, his sector-level bullish/bearish calls inherit that same uncertainty.
因为它是其他判断背后的机制——如果他识别创始人质量的标尺在中国本就失准，那么他对硬件、软件各赛道的看多看空判断也会继承同样的不确定性。
Is this piece representative of the wider VC view, or one person's anecdote?这篇文章代表 VC 圈的普遍看法，还是个人化的一家之言？
It reads as a single, credentialed investor's first-hand impressions from one two-week trip — valuable as data but not a survey; treat specific claims as hypotheses to check against other sources.
这更像是一位有分量的投资人在一次两周行程中的第一手观察，价值在于其亲历性，而非系统性调研；具体判断应当视为待验证的假设，而非定论。
Does the piece explain WHY Chinese founders differ from the Western 'spiky' archetype?文章解释了中国创始人为何不同于西方“锋芒型”原型的原因吗？
The provided excerpt states the observation and the author's surprise but is truncated before mechanisms (education system, capital environment, cultural norms around dissent) are explored.
目前提供的节选只陈述了现象和作者的意外，尚未展开具体机制（教育体制、资本环境、对“出格”行为的文化容忍度等），原文在此处被截断。
What should a reader do with the 'ignored asymmetry' the summary mentions but the text doesn't reach?摘要提到但正文未展开的“被忽视的不对称性”，读者该如何对待？
Treat it as an open thread — the summary flags it as a key takeaway, so a reader wanting the full argument should follow the original link rather than rely on this excerpt alone.
应把它当作一条待完成的线索——摘要将其列为关键结论，若想获得完整论证，读者应点击原文链接而非仅依赖这段节选。
In-depth Analysis · Pros & Cons深入解读 · 优缺点
A Western crypto-native VC spends two weeks touring China's AI ecosystem — founders, VCs, listed-company CEOs — and returns with a thesis reversal: bullish on hardware, bearish on software, and unsettled by a founder archetype that doesn't match his usual pattern-matching.
一位深耕加密货币领域的西方 VC 用两周时间走访中国 AI 生态圈的创始人、投资人与上市公司 CEO，带回一个颠覆预期的判断：硬件比想象中更强，软件比想象中更弱，而他熟悉的那套“识别伟大创始人”的直觉在中国频频失灵。
- First-hand, high-access observation第一手、高准入度的观察Two weeks of direct meetings with founders, VCs, and listed-company CEOs gives this more evidentiary weight than desk research or secondhand takes on China AI.两周内直接会见创始人、VC 与上市公司 CEO，比坐在办公室里做二手研究或转述观点更具证据分量。
- Intellectually honest reversal诚实的认知反转The author reports his priors being falsified rather than confirmed, a rarer and more credible move than typical trip-report confirmation bias.作者报告的是先验判断被证伪而非被证实，这比典型游记式的确认偏误式写作更为可信、也更少见。
- Counter-narrative on hardware对硬件叙事的反向修正Pushing back on the dominant 'export controls cripple China AI' story with direct observation adds a useful corrective data point to a heavily politicized debate.用实地观察反驳“出口管制卡住中国 AI 脖子”的主流叙事，为一个高度政治化的争论提供了有价值的纠偏样本。
- Names a transferable methodological question提出可迁移的方法论问题By questioning whether his founder heuristic generalizes across cultures, the piece raises a point useful to any investor operating cross-border, not just China specialists.文章质疑其创始人识别启发式能否跨文化通用，这对任何跨境投资人都有参考价值，而不仅限于中国专项投资者。
- Single-observer, short window单一观察者、时间窗口短Two weeks and one investor's network cannot represent an ecosystem as large and heterogeneous as China's AI sector; selection bias in who he met is unaddressed.两周和一位投资人的人脉网络，不足以代表中国 AI 这样庞大且异质化的生态；他所接触对象的选择性偏差，文中并未说明或校正。
- No quantitative backing缺乏量化支撑Claims about hardware resilience and software weakness are impressionistic; no shipment data, revenue figures, or funding comparisons are cited in the available text.关于硬件韧性与软件疲软的判断多为印象式表述；现有文本中没有引用出货数据、营收数字或融资对比等量化依据。
- Truncated argument论证被截断The available excerpt ends before the founder-type discussion, hardware evidence, and the 'ignored asymmetry' are fully developed, so key claims can't be fully evaluated here.现有节选在创始人类型讨论、硬件证据以及“被忽视的不对称性”充分展开之前就已中断，导致关键论点在此处无法完整评估。
- Founder heuristic may be self-serving识人启发式可能带有自证倾向A VC's 'spiky founder' filter is itself an unverified theory of what makes companies succeed; treating its non-applicability in China as surprising assumes the filter was correct to begin with.VC 的“锋芒型创始人”筛选标准本身也是一套未经充分验证的成功理论；把它在中国的失效当作意外，其实预设了这套标准原本就是正确的。
Worth reading for anyone tracking China-US AI competition or cross-border VC theses, precisely because the author reports his own model breaking rather than confirming it — but treat the specific hardware/software calls as one well-connected investor's two-week impression, not verified data, especially since the available text cuts off before the core evidence is presented.
对关注中美 AI 竞争或跨境风投逻辑的读者而言，这篇文章值得一读，原因恰恰在于作者报告的是自己认知模型的失灵而非自我确认；但硬件、软件的具体判断应被视为一位人脉丰富的投资人两周内的印象式观察，而非经过验证的数据，尤其是现有文本在核心证据展开前就已中断。
Excerpt原文节选
This is a short excerpt, not the full piece — the complete essay belongs to its original author; please read it in full at the link above.
以下仅为节选，并非全文——完整文章版权归原作者所有，请点击上方链接阅读全文。
The English text on this side is an AI translation provided for convenience; the authoritative version is the source in the other language.
Translator's note: About the author, José Maria Macedo, founding partner of Delphi Ventures and founder and CEO of Delphi Labs. Founded in 2018, Delphi Digital is one of the most influential independent research firms in the crypto and digital assets space, and its Delphi Ventures arm manages a portfolio of over 200 investments. Macedo himself has spent many years deep in the crypto industry; he studied Philosophy, Politics and Economics (PPE) at the International School of Brussels, founded several profitable businesses while still in school, and received honors such as the Duke of York Young Entrepreneur Award. In recent years, Delphi Ventures' investments have been expanding from crypto into AI and deep tech, and this article is Macedo's summary of observations after an on-the-ground survey of China's AI ecosystem.
What I Learned from Two Weeks Inside China's AI Ecosystem
[…the source continues — read the rest at the link above]
[……原文更长，完整内容请点击上方链接阅读]
Original: What I Learned from Two Weeks Inside China's AI Ecosystem https://x.com/ZeMariaMacedo/status/2036779960598163864
译注：关于作者 José Maria Macedo（若泽·马里亚·马塞多），Delphi Ventures 创始合伙人、Delphi Labs 创始人兼 CEO。Delphi Digital 成立于 2018 年，是加密货币和数字资产领域最具影响力的独立研究机构之一，旗下 Delphi Ventures 管理着超过 200 个投资组合项目。Macedo 本人在加密货币行业深耕多年，曾就读于布鲁塞尔国际学校，学习政治、哲学与经济学（PPE），在校期间就创办了多家盈利企业，并获得过约克公爵青年企业家奖等荣誉。近年来，Delphi Ventures 的投资触角正从 Crypto 延伸到 AI 和深度科技领域，本文正是 Macedo 在中国进行 AI 生态实地考察后写的观察总结。
在中国 AI 生态圈摸底了两周，我学到了什么
最近，我在中国待了两个星期，拜访了 AI 生态圈里的各位创始人、风险投资人（VC）以及上市公司的 CEO。去之前，我对这个生态是非常看好的（bullish），满心期待能在这里以远低于西方的估值，发掘到世界级的 AI 人才。
但离开时，我的看法变得更加复杂且微妙： 我对中国硬件的看好程度超出了我的预期，但对软件却更为看衰（bearish） ；同时，我对中国创始人们的一些观察，也让我感到十分惊讶。
创始人的特质之谜 (The Founder Question)
我所投资过的那些伟大的创始人，身上都有一种极具辨识度的“基因”： 独立思考、一身反骨、激情四射且极其执着 。他们绝不乖乖听话，总是不断地问“为什么”，并且拒绝盲从现成的经验。他们做出的决定在外人看来往往匪夷所思，但对他们自己来说却顺理成章。他们骨子里有一种毫不妥协的冲劲，这往往体现在他们过往那种近乎痴迷和追求极致的经历中。在 VC 每天接触的茫茫多的高智商人群里，他们身上那种“锋芒”（spikiness）总能让你一眼就认出来。
然而，我在中国遇到的很多创始人，却属于另一种完全不同的类型——这让我挺吃惊的。
他们无疑是极其优秀的：顶尖名校毕业，在字节跳动或大疆（DJI）有过光鲜履历，在《自然》（Nature）杂志发过论…
[…the source continues — read the rest at the link above]
[……原文更长，完整内容请点击上方链接阅读]
原文：What I Learned from Two Weeks Inside China's AI Ecosystem https://x.com/ZeMariaMacedo/status/2036779960598163864