The Bottleneck Moved
When answers become cheap, the hard part is knowing what to ask.
After using the ultra-deep research system for a while — the one I wrote about in previous posts — I realized my bottleneck is no longer answers. It is questions. What should I even be asking, and how should I be asking it?
So I built a council. Elon, Charlie Munger, Naval, and a few others. I throw a direction question at them, and they argue with each other, poke holes in my thinking, and point out what I am missing. They don’t decide for me, but they crack open my thinking so I know what to ask next.
This is not a new idea. People were talking about AI advisory councils on Twitter maybe a year ago. I saw those posts and did not think much of it. But now that I actually have more information than I know what to do with, I get it. I genuinely need something like this.
Paul Graham wrote recently: “When anyone can make anything, the big differentiator is what you choose to make.” Greg Brockman called taste “a new core skill.” They are talking about the same thing I ran into — when execution gets cheap, the bottleneck moves up a level — from “how to do it” to “what to do,” from “finding answers” to “asking the right questions.” The council is the tool I use to keep pushing myself to iterate at a higher level.
The council is open source: github.com/mkmkkkkk/wisdom-council
瓶颈转移了
当答案变得廉价，难的是知道该问什么。
用了一段时间超深度研究系统 — 就是之前博客写的那个 — 我发现我的瓶颈不再是答案了，而是变成了该问什么问题、该怎么问问题。
所以我搭了一个顾问团。Elon、查理芒格、纳瓦尔，还有几个人。我把方向性的问题丢进去，他们互相吵，戳我思维的漏洞，指出我没看到的东西。他们不替我做决定，但会把我的思路撕开一个口子，让我知道下一步该往哪问。
这不是个新概念，大概一年前 Twitter 上就有人在讨论了。我当时看了没当回事。但现在手里的信息多到不知道该拿它们怎么办的时候，我才理解了。我是真的需要这个东西。
Paul Graham 最近写到：“当任何人都能做出任何东西的时候，真正的差异化是你选择做什么。” Greg Brockman 说品味是“一项新的核心技能”。他们说的和我撞上的是同一件事 — 当执行变得廉价，瓶颈就往上移了一层——从「怎么做」变成「做什么」，从「找答案」变成「问对问题」。顾问团是我推动自己持续往更高层迭代的工具。