Cink

[!NOTE]
未来并不是线性展开的。
所以，焦虑并不能真正帮助我们接近未来。更重要的是，在你当下所能看到的边界之内，做一个对得起自己的选择；至于剩下的部分，就交给时间。

[!TIP] Source link: 2016 年，我做过一次 AI 写代码创业

[!NOTE]
Now information spreads instantly. Being fast no longer creates leverage. It simply keeps you from falling behind.

Once everyone has access, the advantage disappears. This raised the floor for everyone at the same time. Not gradually. All at once.

Competence Became Infrastructure, AI did not make work easier. It made basic competence cheaper. And when competence becomes cheap, it stops being rewarded. That is why yesterday’s excellence feels invisible today.

When one person can do the work of three, volume stops signaling value. It only signals access to tools.

AI can scale messages, but it cannot build reputation at the same depth. AI cannot speed up indecision, humans can.

A fast no is often better than a slow yes.

People who keep trying to outperform machines at machine work lose. People who redesign the work around human judgment create separation again.

[!TIP] Source link: You’re Not Getting Worse. Everyone Else Just Got AI

[!NOTE]
所谓幸福，不过是寻得一个人，肯听你说些无关紧要的废话，也肯把自己的废话，细细说与你听。

你见了海上的潮声，第一时间拍与我看，那片海便不只是海，是你想与我共赏的人间；我夜里抬头见了星月，也急急拍下来发给你，那月光与星光，便成了我藏不住的欢喜。不必刻意修饰，不必字字斟酌，只是自然而然地，想把眼里所见、心里所念，第一时间说与你听。

碎碎念只说给有回应的人听，一腔热忱，也只交付给懂得接住的人。你说一句，我应一声；你分享日常琐碎，我认真接住情绪。事事有回音，件件有着落，开心便会成倍地漾开，那些藏在心底的委屈与不安，也会在一来一往的温柔里，慢慢消散。

我们这一生，会遇见很多人，有的人只适合擦肩而过，有的人只适合浅尝辄止，唯有那个愿意听你说废话，也愿意对你讲废话的人，才是真正把你放在心尖上的人。他不会觉得你的日常琐碎无聊，不会嫌你的碎碎念聒噪，更不会对你的分享视而不见。他会把你的每一句话，都放在心上，会把你的每一次分享，都当作一份礼物。

你愿意说，我愿意听；我愿意分享，你愿意回应。

有人把你的废话当情话，把你的琐碎当珍贵，把你的日常，当作他生命里最重要的部分。

[!TIP] Source link: 人民日报：双向的分享欲真的好浪漫

[!NOTE]
我们之间有一种“留白”——不必每日问候，不必刻意维护，但知道彼此都在。就像中国山水画中的留白，那空出的部分，反而是意境所在。

我们太习惯解决问题了，以至于忘记了，有时候人需要的不是方案，而是被理解的感觉。

当朋友倾诉时，那句“我懂”比“你应该”珍贵得多；

当亲人抱怨时，安静的陪伴比理智的分析更有温度。

真正的倾听不是被动地听，而是主动地理解 。是放下自己的预设，走进对方的情感世界，哪怕那里阴雨连绵。这种倾听不评判、不打断、不急着拯救，只是温柔地承接：“我在这里，你的感受很重要。”

最舒服的关系里，你不必永远正确，不必永远阳光，不必永远游刃有余。 你可以展露困惑、恐惧、不完美，而对方不会因此看低你，反而会因为你的真实而更靠近你。

[!TIP] Source link: 相处时有这三种感觉，说明你遇到了值得珍惜的人

[!NOTE]
Build for people who don’t know what an API or AI wrapper is.

Service is valuable. Software as a Service (software is just way to make it possible, faster, easier and more profitable)

[!TIP] Source link: IndieHackers are in a Bubble. Step out of it.

[!NOTE]
You’re too busy trying to keep your current job to redesign your future one.

Nobody could define what higher-value work actually looked like in practice. Nobody could describe the new role. So they defaulted to the only thing they could measure: cost reduction.

Not “I’m more efficient now.” But “We can now do this specific thing we couldn’t do before, which creates this specific business value.”

[!TIP] Source link: Your Job Isn’t Disappearing. It’s Shrinking Around You in Real Time

[!NOTE]
The process of explaining concepts to others forces me to deepen my own understanding and uncover gaps in my knowledge.

Remember that communication is a two-way street. Listen actively to others and seek to understand their perspectives before responding.

Whether it’s spending time with loved ones, pursuing hobbies, or simply taking a walk in nature, make space for the things that nourish your soul.

[!TIP] Source link: 10 lessons from 12 years at Google

[!NOTE]
ご飯ってさあ、一人で食べてももちろん美味しいけど、
誰かと一緒に食べる事って、それだけで特別なことだと思うんだよね。

誰かを愛することって、失う時の痛みも引き受けるっていうことだと思うんだよ。
この痛みは味わえて、僕は幸せなんだ…

別れの時には、一緒にいたかろうぜ。

[!TIP] Source link: 承受失去之痛，是我爱的证明 | 我们的餐桌

[!NOTE]
Agents don’t lack intelligence. They’re forced to operate in unstable, unpredictable environments.

[!TIP] Source link: What We Talk About When We Talk About Agent Infra - me.0xffff.me

[!NOTE]
血圧が下がると、まぶたが少し腫れるので、それを軽減するためです。

[!TIP] Source link: 我们犯困时揉眼睛的意义与理由是什么？

[!NOTE]
The fastest way to get good at it is to learn the same skills that make someone effective at leading engineers: clarity, delegation, feedback loops, and async communication.

[!TIP] Source link: Your AI coding agents need a manager

[!NOTE]
吃真实的食物：完整的或最小加工的、可被识别为食物的食品。用很少的原料制备，不含添加糖、工业油脂、人工香料或防腐剂。

[!TIP] Source link: 吃真实的食物：聊聊最新的美国膳食指南（2025-2030 年版） - 少数派, 吃真实的食物：聊聊最新的美国膳食指南（2025-2030 年版） - 少数派

[!NOTE]
The five W questions are important because they help establish the framework of any project.

Once you have established the answers to the ‘who, what, when, where and why’, you’ll then have a clear picture of the tasks in front of you.

[!TIP] Source link: 5 Ws (and 1 H) To Be Asked of Every Project | Adobe Workfront

[!NOTE]
Who? What? Where? When? Why? / How?

[!TIP] Source link: 5 Ws and H

[!NOTE]
Focus on skills AI can’t easily replace: communication, problem decomposition, domain knowledge.

AI handles the routine 80%, humans focus on the hardest 20%. Architecture, tricky integrations, creative design, edge cases: the problems machines alone can’t solve.

Critical skills become software architecture, system design, performance tuning, and security analysis.

Stay current on new tools and best practices. Double down on what makes a human developer indispensable: sound judgment, system-level thinking, and mentorship.

Companies no longer want developers who are either too shallow or too narrowly focused; they want a strong core competency plus ability to work across the stack. Part of the reason is efficiency: a T-shaped engineer can often solve problems end-to-end without waiting on handoffs.

[!TIP] Source link: The Next Two Years of Software Engineering

[!NOTE]
A computer can never be held accountable. That’s your job as the human in the loop.

Code review is also how teams share system context. If AI writes the code and nobody can explain it, on-call becomes expensive.

When a developer submits AI-generated code they don’t fully understand, they’re breaking the knowledge transfer mechanism that makes teams resilient. If the original author can’t explain why the code works, how will the on-call engineer debug it at 2 AM?

[!TIP] Source link: AI writes code faster. Your job is still to prove it works.

[!NOTE]
Memory is almost like a “re-living,” an embodiment of a past state. The story becomes a part of your lived experience.

Because it turns out that you don’t really know something well until you can explain it to someone.

And the first time you try to explain something you learned to someone will not be the best way to explain it. You need to try many times. So you actually want people to ask you about it and you want to help people learn so that you can learn as well!

[!TIP] Source link: Agentultra - How I Learned Everything I Know About Programming

[!NOTE]
People increasingly read LLM summaries rather than clicking on search engine links or the accompanying ads.

LLMs potentially also make software easier to write, which is a double-edged sword. It might help you code features faster, but it also lowers the barrier, so that more people can compete.

[!TIP] Source link: Is the golden age of Indie software over?

[!NOTE]
不用太在意短期的得失，喜欢什么就去做。因为热爱一件事（或者爱一个人），其实是非常难得的，错过了多可惜啊。

做困难的事。别人害怕的事，交给我来做。别人抢着做的事情，让他们去做。

[!TIP] Source link: 技术Leader如何创造业务价值？

[!NOTE]
If we want to avoid the great content collapse, we need a model of GenAI usage that creates opportunities for all parties (user, GenAI company, content creator) to capture value.

[!TIP] Source link: GenAI, The Snake Eating Its Own Tail: How tools like ChatGPT and Claude are destroying the ecosystems they rely on, and what to do about it

[!NOTE]
Visualizing MCP as a bridge makes it clear: MCP doesn’t handle heavy logic itself; it simply coordinates the flow of data and instructions between AI models and tools.

Just as USB-C simplifies how you connect different devices to your computer, MCP simplifies how AI models interact with your data, tools, and services

[!TIP] Source link: 什么是模型上下文协议 (MCP)？ - DemoChen’s Clip

[!NOTE]
Writing forces clarity. When I explain a concept to others - in a doc, a talk, a code review comment, even just chatting with AI - I discover the gaps in my own understanding.

Senior engineers who say “I don’t know” aren’t showing weakness - they’re creating permission.

Your job isn’t forever, but your network is

Deleting unnecessary work is almost always more impactful than doing necessary work faster. The fastest code is code that never runs.

Before you optimize, question whether the work should exist at all.

[!TIP] Source link: 21 Lessons From 14 Years at Google

[!NOTE]
自分の人生と向き合うために戻ってきたんだと思う。
私は逃げてきた。

[!TIP] Source link: 逃回此处闲静 | 小森林

[!NOTE]
Relax and have some fun with the language, everything will work out eventually.

[!TIP] Source link: 为什么很多中国学生学不好英语？ - RiddleView 的回答

[!NOTE]
English is a language and a tool for the real world and real people.
The language itself never meant to be studied, but to be used.

[!TIP] Source link: 为什么中国的英语教育这么失败? - RiddleView 的回答

[!NOTE]
A great one would be to have a panel of designers rank screenshots of product outputs for a prompt - but instead we get endless math, science and SWE benchmarks that don’t really cover this.

[!TIP] Source link: Are we in a GPT-4-style leap that evals can’t see?

[!NOTE]
お元気ですか——
私は元気です！

[!TIP] Source link: 你好吗？我很好 | 情书

[!NOTE]
When you type a sentence into a search box, it’s easy to imagine the search engine seeing the same thing you do. In reality, search engines (or search databases) don’t store blobs of text, and they don’t store sentences. They don’t even store words in the way we think of them.

[!TIP] Source link: From Text to Token: How Tokenization Pipelines Work

[!NOTE]
Natural language is also ambiguous — one of the reasons that programming languages exist.

[!TIP] Source link: Don’t Talk English to Your LLM

[!NOTE]
“You can move all your eggs to one basket,” he says, “but in the end, there’s usually a password protecting that basket.”

[!TIP] Source link: Avoiding Vulnerable Passwords—and Rules, Too