分享一个我从经典科幻小说《雪崩》里学习到的思维，用于对应目前AI大爆发的情况，普通人应该如何面对：
我们处理问题和信息，首先第一层是自己的大脑和思维模型。这一层有信息接收、处理和分析的能力。
它就像是一个操作系统。
每个人的操作系统都有不同，有的快，有的慢，有的有误差。
早在电脑和互联网出现前，部分人可以用他人作为自己的延伸，帮助自己处理信息，或是处理事务。他们可以雇佣专职的人员帮助他们完成事物。这种人的终级体现，就是一个帝国的皇帝。全国的人，都是他的“agent”，他控制甚至拥有这些人。
后来技术发展，普通人也都有了使用“agent”的能力。比如蒸汽机，就是一种agent，解放了人的体力。后来的计算机，个人电脑，互联网，也都是。
而现在，我们有了AI，agent的范围又被扩大了。
但大多数人，其操作系统并没有提升，所以对于使用AI，想象力也有限。比如我的twitter互关朋友 @yihong0618 分享的，他的一个联系人托他问GPT的第一个问题：双色球怎么选。我开玩笑说这是大多数人的第一想法：它能帮我赚钱么？它能帮我省钱和时间么？它能帮我泡妞/帅哥么？
扯远了，我想说的是，要想在AI爆发的时代生存，有两条路线：
提升自己的操作系统性能和带宽。
研究如何运用Agent来帮助自己解决实际问题，从而可以有更多时间和资源来提升自己的操作系统。
第一条路，是长期方向，有没有AI，都得做。 第二条路，要灵活，根据情况调整。
就我自己来说，第一条路走了几十年，经过了两种语言和文化的历练，还算有些经验和方法了。现在我会花一些时间和精力，来研究第二条路，两条路同时走。
对这个话题感兴趣的，订阅我的中文邮件 bearliu.substack.com 或是关注我的博客 beartalking.com ，之后会分享更多关于AI的东西。
Are you tired or exhausted with all the AI-related news and posts? I am! But then I realised I couldn’t avoid it, so I started to think (as a designer), how am I going to deal with it?
Then I got this thought - actually, I picked it up from the classic sci-fi novel "Snow Crash":
We process problems and information through our own brains and mental models. Think of it like an operating system.
But here's the thing, everyone's operating system is different. Some are fast, some are slow, some have errors(bias).
Even before computers and the internet came around, some people could use others to help them process information or manage affairs. They could hire dedicated personnel to help them complete tasks. The ultimate embodiment of such people is an emperor of an empire. All the people in the country are his "agents," and he controls and even owns these people.
Later technological advancements gave ordinary people the ability to use "agents" as well. For example, the steam engine was an agent that freed people from physical labor. Later, computers, personal computers, and the internet all were.
Nowadays, we have AI, and the range of agents has been expanded again.
However, most people's operating systems have not improved, so their imagination is also limited when it comes to using AI. For example, a contact of my friend asked GPT the first question: how to pick the lottery number. I joked that this is probably most people's first thought: can it help me make money? Can it help me save money and time? Can it help me pick up girls/guys?
To survive in the age of AI explosion, there are two ways:
Improve the performance and bandwidth of your own operating system.
Study how to use agents to help solve practical problems, thereby freeing up more time and resources to improve your own operating system.
The first path is a long-term direction that needs to be pursued regardless of whether there is AI or not. The second path requires flexibility and adjustment depending on the situation.
For myself, I've been working on the first path for decades, with two languages and cultures, and gaining some experience and methods. Now, I'll spend some time and effort to study the second path while continuing on the first path.
This is inevitable, so I will start to deal with it.