The seed
I wrote three lines in my notebook this morning. They came out fast, the way things do when you’ve been chewing on something for months without noticing.
Don’t chase hype blindly. Don’t ignore trends willfully. Hold your ground, embrace change, and cherish everything you have and everyone you meet.
I work in AI infrastructure. Every morning I open my laptop to a new “revolution.” Vibe coding. AI agent frameworks. Another startup claiming it’ll change everything. Last week’s trending repo, already forgotten.
But something fundamental is moving underneath all that noise. GPU compute is reshaping the entire computing stack. The infrastructure work I do every day didn’t exist three years ago. Ignoring real change is exactly as dangerous as chasing fake change.
So the question is what deserves attention, and what’s just noise dressed up as signal.
There’s no formula. But I’ve found three anchors.
I. Don’t chase hype
In January 2019, WeWork was worth $47 billion. Nine months later, it couldn’t even go public. The valuation crashed 78%, the CEO cashed out $700 million before being shown the door, and the company filed for bankruptcy in 2023.
In May 2022, Terra/Luna evaporated $45 billion in a single week. FTX followed. Celsius. Three Arrows Capital. The entire crypto world fell like dominoes. The people who jumped in didn’t truly understand blockchain. They just saw everyone else making money.
The pattern is always the same. A real innovation (shared workspaces, blockchain) gets inflated by capital and FOMO until the price tag has nothing to do with the underlying value. The people who get hurt are the followers, not the builders.
Buffett has a phrase for this — “circle of competence”:
“The size of that circle is not very important; knowing its boundaries, however, is vital.”
Chasing hype is stepping outside your circle of competence while still believing you’re inside it. You don’t understand the underlying logic. But you see prices rising, others profiting, the feed buzzing, so you follow. That’s imitation.
Naval puts it more directly: “Escape competition through authenticity.” Not chasing hype still takes work. You have to ask yourself, honestly: Do I actually understand this, or am I just afraid of missing out?
AI is in the same cycle right now. YC’s Jared Friedman said a quarter of the Winter 2025 batch had codebases that were 95% AI-generated. Sounds cool. But when CodeRabbit went through 470 open-source pull requests, the AI-co-authored ones had 1.57× the security findings of the human-only ones (2.74× for XSS). And in METR’s early-2025 randomized trial, 16 experienced open-source developers took 19% longer with AI tools, on repos they’d contributed to for years. Karpathy coined “vibe coding,” but his definition was to “forget that the code even exists.” That reads like a warning.
The SEC has started fining companies for “AI washing” — slapping “AI-powered” on products with no meaningful AI inside. The “.com” suffix of our decade.
Laozi said: The highest good is like water. Water benefits all things and does not compete.
Water doesn’t rush to be where everyone else is. It finds the low places nobody wants. And precisely because of that, nothing is stronger.
Not competing still involves moving. You just don’t get swept along by other people’s movement.
II. Don’t ignore trends
In 1975, Kodak engineer Steven Sasson built the world’s first digital camera. 10,000 pixels.
He brought it to management. Management said: Nice, but our film business makes billions a year. Drop it.
In 2012, Kodak filed for bankruptcy.
Kodak’s tragedy wasn’t a technology blind spot. They invented the technology. The tragedy was that they saw the trend and chose not to look.
Blockbuster’s story is even more absurd. In 2000, Netflix came knocking and offered to sell itself for $50 million. Blockbuster laughed and said no. At the time, Blockbuster had nearly 7,700 stores and about 96,000 employees. A decade later, bankruptcy. Netflix is now worth hundreds of billions.
A former Blockbuster executive eventually admitted: “Digital would have changed Blockbuster’s business… but it wasn’t its killer. That credit belongs to Blockbuster itself.”
What kills you is avoiding change.
Marcus Aurelius:
“The universe is change; our life is what our thoughts make it.”
He also said: “Observe constantly that all things take place by change, and accustom thyself to consider that the nature of the Universe loves nothing so much as to change the things which are.” This isn’t self-help. A Roman emperor wrote it on battlefields and during plagues, as practical philosophy.
Laozi used water to make the same point: Nothing in the world is softer or weaker than water, yet for attacking the hard and strong, nothing surpasses it.
Water is change. It doesn’t need to become a rock to fight a rock. It flows around and seeps in. Given enough time, the rock splits open.
Trends are like water. You may not like them, but you can’t pretend they don’t exist. GPU compute is changing the computing paradigm — that’s the reality of my workweek, not hype. Same goes for LLMs: they’re rewriting the basic assumptions of software engineering.
The difference between a trend and hype is durability. A trend reshapes everything even after the excitement fades.
The test for anything new, then: when the money pulls out, is it still there?
III. Hold your ground, embrace change, cherish everything
If “don’t chase hype” plus “don’t ignore trends” were enough, everyone would already be doing it. The hard part is the third line. How do you hold your ground and embrace change at the same time?
NVIDIA gave us an answer.
In 1993, three people started a company at a Denny’s, $200 each. Jensen Huang later said that if he’d known the road would be “a million times harder,” he might not have started. In 1997, the company was a month from going out of business. His running line in internal meetings was: “We are thirty days from going out of business.”
In 2006, NVIDIA released CUDA, a platform for general-purpose GPU computing. Nobody cared. Gaming companies didn’t care. Scientific computing didn’t care. Wall Street didn’t care.
Huang tried mobile chips when those got hot, and pulled back when Tegra didn’t stick in phones. Didn’t pivot to social. Didn’t rebrand as a crypto company when miners were buying every GPU on the planet. He stayed on GPU compute. For thirty years.
Then the AI wave hit. Suddenly the whole world realized only NVIDIA had the infrastructure. By October 2025, NVIDIA was the first company to clear $5 trillion in market cap.
Chinese has a four-character phrase for this, distilled from Sun Tzu’s Art of War: 守正出奇. Defend the orthodox, attack with the extraordinary. The “orthodox” is holding your ground: your foundation doesn’t move. The “extraordinary” is embracing change: on top of that foundation, you adapt and explore. NVIDIA’s “orthodox” was GPU computing. Its “extraordinary” was that every time a new wave came, it was already prepared.
Zeng Guofan, the Qing statesman, put it more bluntly:
“Shouting opinions from the sidelines is useless. You must step into the arena, shoulder the responsibility, before success is even possible.”
守正出奇 comes from the arena. You have to be on the field before you can talk about defense and offense.
There’s a Buddhist image I keep coming back to. Dhammapada, verse 401:
“Like water on a lotus leaf, or a mustard seed on the point of a needle — he who does not cling to sensual pleasures, him do I call a holy man.”
The lotus leaf doesn’t reject the water. The water falls on the leaf, fully in contact. But the water doesn’t stick. The leaf doesn’t absorb it. They meet, then each goes its own way.
That’s the stance to take toward hype, the status quo, and outcomes: fully in contact with the world, but not stuck to it.
My work puts me at an intersection: AI meets infrastructure. Every day two forces collide on my desk — a real computing paradigm shift, and all the noise around it. My anchor sits underneath any technology or trend: the conviction that value lives in the gaps between layers. Other people see systems as stacks. I see the dark spaces between them, and I build until information flows through. That doesn’t expire with any hype cycle, because weaving connections doesn’t belong to any particular layer.
Closing
Seneca: “You should not copy the bad simply because they are many, nor should you hate the many because they are unlike you.”
Probably the best social-media usage guide from 2,000 years ago.
Naval: “Desire is a contract that you make with yourself to be unhappy until you get what you want.”
Chasing hype is a form of desire. Fearing the trend is also a form of desire. Both keep you living in the future instead of the present.
Each piece of that third line is easy to mistake for something cheaper. Holding your ground looks like stubbornness, embracing change like chasing the wind, cherishing what you have like a platitude. Taken together, they’re water on a lotus leaf — in full contact with the world, bound by none of it.
Zeng Guofan again:
“A troubled mind and anxious thoughts — these are precisely the forge that tempers heroes and refines you to completion.”
So don’t fear the confusion. Feeling torn between hype and trends is the forging. You don’t need to eliminate the tension. Find the anchor underneath it, and then walk into the arena.
Written on April 1, 2026 — a morning when, predictably, three new AI frameworks were announced.
缘起
今天早上我在笔记里写下三句话：
对热点，不要趋之若鹜。 对趋势，不要熟视无睹。 要固守本心，拥抱变化，珍惜所拥有的和遇到的一切。
写得很快。有些事你心里已经琢磨了好几个月，自己却没意识到，到点了它就自己冒出来了。
我做 AI 基础设施相关的工作。每天打开电脑，都有一个新的“革命”在等我。Vibe coding。AI agent 框架。又一个号称要改变一切的 startup。上周 trending 的 repo，这周已经没人提了。
但底下确实在发生根本性的变化。GPU 算力正在重塑整个计算栈。我每天做的基础设施工作，三年前根本不存在。忽略真变化，和追逐每一个假变化，是一样危险的。
所以要问的是：什么值得关注，什么只是披着信号外衣的噪音。
这件事没公式。但我找到了三个锚点。
一、对热点，不要趋之若鹜
2019 年 1 月，WeWork 估值 470 亿美元。九个月后，连 IPO 都没上成。估值暴跌 78%，CEO 套现 7 亿美元后被扫地出门，公司最终在 2023 年申请破产。
2022 年 5 月，Terra/Luna 一周蒸发了 450 亿美元。FTX 紧随其后。Celsius。Three Arrows Capital。整个加密世界像多米诺骨牌一样倒下。追进去的人并没真弄懂区块链，只是看见别人都在赚钱。
模式每次都一样。一个真正的创新（共享办公、区块链）被资本和 FOMO 一路推高，直到价格跟底层价值毫无关系。受伤的不是建造者，是跟风的人。
巴菲特那个老概念叫“能力圈”：
“圈的大小不重要，重要的是你知道边界在哪。”
追热点，就是你已经走出了能力圈，却还以为自己在里面。你不懂底下的逻辑，但看见价格在涨、别人在赚、朋友圈在刷，于是也跟了。说白了就是模仿。
Naval 说得更直接：“Escape competition through authenticity.” 用真实摆脱竞争。不追热点，不代表躺平。你得老老实实问自己一句：这件事，我是真懂，还是只是怕错过？
AI 现在就在跑同一个循环。YC 合伙人 Jared Friedman 说，2025 冬季批次里有四分之一的公司，代码库 95% 是 AI 生成的。听着挺酷。可 CodeRabbit 翻了 470 个开源 PR，AI 参与写的那些，安全问题是纯人工 PR 的 1.57 倍（XSS 漏洞是 2.74 倍）。METR 在 2025 年初做过一个随机对照实验：16 位资深开源开发者，在自己参与了好几年的仓库里干活，用上 AI 工具反而多花了 19% 的时间。“Vibe coding”这词是 Karpathy 造的，但他的定义是“连代码的存在都忘掉”。这话听着更像个 warning。
SEC 已经开始罚“AI washing”了——给没什么实质 AI 的产品贴“AI-powered”标签。这就是我们这年代的“.com”后缀。
老子说：上善若水。水善利万物而不争。
水不争。水不急着去所有人都挤的地方。水找的是别人不要的低处。正因为这样，没什么比它更强。
不争，也得动。只是不被别人裹挟着走。
二、对趋势，不要熟视无睹
1975 年，柯达工程师 Steven Sasson 造出了世界上第一台数码相机。10000 像素。
他拿去给管理层看。管理层说：好东西，但我们的胶卷业务每年进账几十亿美元。别搞了。
2012 年，柯达申请破产。
柯达的悲剧不是技术盲区。那技术是他们发明的。悲剧是他们看到了趋势，却选择不看。
Blockbuster 的故事更荒诞。2000 年，Netflix 主动上门，想把自己卖给对方，开价 5000 万美元。Blockbuster 笑着拒了。那时候 Blockbuster 有近 7700 家门店、约 9.6 万名员工。十年后，破产。Netflix 现在市值几千亿美元。
后来一位前 Blockbuster 高管承认：“数字化会改变 Blockbuster 的业务……但不是杀死它的原因。真正杀死它的是 Blockbuster 自己。”
所以说，要命的是躲着变化走。
Marcus Aurelius 在《沉思录》里写过：
“宇宙即变化，我们的生命由思想塑造。”
他还有一句：“留意万物皆是变化的结果，习惯这样的认知：自然最喜欢做的事，就是改变现有的形态。”这是一个罗马皇帝在战场上、在瘟疫里写下的实操哲学，可不是鸡汤。
老子用水说明了同一件事：天下莫柔弱于水，而攻坚强者莫之能胜。
水本身就是变化。它不需要变成石头才能对抗石头。它绕过去，渗进去，年深日久，石头就裂开了。
趋势就是水。你可以不喜欢，但没法假装它不存在。GPU 算力正在改写计算范式——这是我每天工作的现实，不是炒作。LLM 也一样，它在重写软件工程的基本假设。
热点和趋势的区别就两个字：持久。趋势在兴奋退潮之后，还在重塑整个地形。
所以看什么新东西，都可以问一句：钱撤走之后，它还在不在？
三、固守本心，拥抱变化，珍惜一切
如果“不追热点”加上“不忽略趋势”就够了，所有人早就在做了。难就难在第三句。怎么同时做到固守和拥抱？
NVIDIA 给了答案。
1993 年，三个人在 Denny’s 餐厅里创业，每人出 200 美元。Jensen Huang 后来说，要是当时知道这条路“难一百万倍”，他可能就不会开始了。1997 年公司只剩一个月的现金流。他内部演讲的口头禅是：“我们离倒闭只有 30 天。”
2006 年，NVIDIA 发布了 CUDA，一个让 GPU 做通用计算的平台。当时没人在乎。游戏公司不在乎，科学计算社区不在乎，华尔街不在乎。
移动芯片火的时候，Huang 也下场做过 Tegra，手机没做起来，就撤了。没转向社交媒体。没在矿工买光所有 GPU 的时候给公司贴个加密货币的标签。他守着 GPU 计算，守了三十年。
然后 AI 浪潮来了。全世界突然发现，只有 NVIDIA 有这个基础设施。2025 年 10 月，NVIDIA 成了第一家市值过 5 万亿美元的公司。
《孙子兵法》说“以正合，以奇胜”，后人浓缩成四个字：守正出奇。守正就是固守本心，根基不动。出奇就是拥抱变化，在根基之上不停适应、不停探索。NVIDIA 守的“正”是 GPU 计算。出的“奇”是每一次新浪潮来，它都已经准备好了。
其实 NBA 里也有这条规律。马刺 1997 年抽到邓肯，然后波波维奇就守了二十年的“正”：低位、传切、纪律。期间小球革命来了、抱团时代来了、三分时代来了，别的球队跟着潮流换教练换体系，马刺动都不动。可马刺每年都在场上打得跟时代不一样：他们的 Beautiful Game 就是 2014 年的“奇”，三传两递把热火三巨头打到没脾气。守了二十年那个“正”，等到机会来，“奇”出得行云流水。
库里 2015 年改写了 NBA，是另一种“出奇”，但他底下也有一个“正”——从小练投篮，他爸 Dell Curry 三分球时代的功底。所有的“奇”都长在某个“正”上。单出奇没用，没根基那不叫奇兵，叫散兵。
曾国藩说得更实在：
“凡天下事，在局外呐喊议论，总是无益；必须躬身入局，挺膺负责，方有成事之可冀。”
守正出奇是躬身入局之后才有的智慧，坐在家里想不出来。你得先站在场上，才有资格谈守和攻。
佛教里有一个意象我很喜欢。《法句经》第 401 偈：
“如同荷叶上的水珠，如同针尖上的芥子——不执着于感官愉悦的人，我称他为圣者。”
荷叶不拒绝水。水落在叶上，完全接触。但水不沾叶。叶不吸水。它们相遇，然后各自自在。
对热点、对现状、对结果，都该是这个态度：跟世界充分接触，但不被黏住。
我的工作正好卡在一个特殊的位置：AI 和基础设施的交叉点。每天眼前都是两股力在撞——一个真正的计算范式转移，还有围绕它的所有噪音。我的锚点压在技术和趋势底下，是一个更老的信念：价值住在层和层之间的缝隙里。别人看到栈，我看到栈之间的暗处，然后一直搭，直到信息流得过去。这件事不会随哪一波热点过期，因为“织连接”本身不属于任何一层。
收束
Seneca：“不要因为他人众多就去模仿，也不要因为他人与你不同就去厌恶。”
大概是 2000 年前最好的社交媒体使用指南了。
Naval：“Desire is a contract that you make with yourself to be unhappy until you get what you want.” 欲望是你跟自己签的一份合同：拿到想要的之前，你同意一直不快乐。
追热点是一种欲望。对趋势的恐惧，也是一种欲望。两者都把你拽进未来，让你不在当下。
第三句拆开看，每一样都容易被看偏：固守本心看着像固执，拥抱变化像追风，珍惜一切听着又像鸡汤。三样合在一起，就是荷叶上的水珠：跟世界完全接触，谁也不黏住。
曾国藩还有一句：
“困心衡虑，正是磨练英雄，玉汝于成。”
所以别怕困惑。在热点和趋势之间撕裂的那个状态，本身就是锻造过程。张力不用消除。要做的是找到底下那个锚点，然后躬身入局。
也就这么回事。
写于 2026 年 4 月 1 日，那天早上又冒出了三个新 AI 框架。
Comments留言