In the age of AI, the highest skill is knowing what not to do.

The one thing AI is not short on is tools and features.

AI 时代最不缺的就是工具和功能。

Every day, new models, new frameworks, new concepts arrive. Everyone announces what they will do with AI. But lately I keep coming back to a simpler idea. The most effective people are not the ones who do everything. They are the ones who know what not to do.

每天都有新模型、新框架、新概念冒出来。每个人都在喊”我要用 AI 做这个、做那个”。但我最近越来越觉得：真正厉害的人，不是什么都做的人，而是知道什么不该做的人。

We just made a decision that went against instinct: we turned off an AI feature.

我们最近做了一个违反直觉的决定：停掉了一个 AI 功能。

Not because we couldn’t build it. Because after we added it, the product got worse — more power-hungry, more complex, more confusing for users.

不是因为做不出来，而是因为它加上去之后，产品反而变差了——更耗电、更复杂、用户更困惑。

After we removed it, the battery lasted longer, the flow felt right, and users said, “finally.”

去掉之后呢？续航好了，体验顺了，用户说”终于好用了”。

Here is the pattern. Technology is not about “what can be done.” It is about “doing the right thing precisely.”

这让我想到一个道理：技术的价值不在于”能做什么”，而在于”精准地做对的事”。

A Swiss Army knife can do anything. But when you cut a steak, you still want a good knife.

一把瑞士军刀什么都能干，但你切牛排的时候，还是想要一把好刀。

The same logic holds for building a company. It is not “more features, better.” Not “bigger team, better.” Not “more funding, better.” Find the one point that matters most, and do it until it is undeniable.

同样的道理放在创业上也成立：不是功能越多越好，不是团队越大越好，不是融资越多越好。找到那个最关键的点，把它做到极致，就够了。

In a time when everyone is adding, subtraction is the real edge.

在这个所有人都在做加法的时代，做减法才是真正的竞争力。

What are you holding on to that you already know you should cut?

你最近有没有”明知该砍，但不舍得砍”的东西？