Notes

One of the fastest ways to recognize and correct our own mistakes is to see the same mistakes in someone else. We often don’t realize how wrong we are. If we can't see ourselves clearly, look for someone who is similar to us in this area. After observing a Silicon Valley VC’s daily life from an outsider’s perspective and reflecting on it seriously, I felt I could see the career more objectively.

When evaluating startup opportunities, don’t assume Big Tech will always be slow, but don’t assume it can reinvent itself and outperform startups either. Core researchers from Meta FAIR and Google Deepmind said AI is moving too fast for large companies to keep up, largely because of top-down decision-making and organizational complexity. That creates opportunities for startups. But a TikTok HR leader said the company is redesigning its organization around AI agents and project-based pods to make it much more agile. Whether that transformation will actually work is still an open question.

In the tech industry, the closer our roles are to the core technology, the greater the potential payoff tends to be, whether we want to become senior executives or found startups. This is because we should join companies whose success is heavily driven by the function we perform, and also because technology increasingly drives companies’ growth, value creation, and competitive advantage. Rough ranking: Researcher > Engineer > PM > other roles like Corp Dev, Strat, and Finance. Interested in investment or finance? Join an investment firm or a finance-driven business, unless we’re just using the role to build experience for something else.

Also, whenever possible: choose the most advanced, fastest-growing sectors; join the leading companies; work on the teams most critical to companies’ future growth.

One of my former managers just raised a ~$300M Series B for his AI company, and they'll probably go public next year. Here are a few things I learned from this: 1. People who understand technology benefit the most in this era, disproportionately so. Even people without a technical background should learn technology as much as we can. 2. But success on this scale is hard to replicate. In tech, you can spend years working in a tiny niche. If that area suddenly gets a breakthrough, all the knowledge and work you've built up can turn into massive value. If the breakthrough never comes, the outcome can look completely different. 3. Never quit. If a startup fails, go get a job, learn more, and start another company. This is already about the fifth company my former manager has built.

Someone who initially likes us might later change their mind. But in my experience, the reverse is much rarer. If someone seems to dislike us from the start, even a great deal of effort is almost unlikely to make them genuinely like us. The chances of turning someone who clearly dislikes us into someone who likes us are extremely low. So, if we have shown warmth and enthusiasm but the other person consistently shows little interest in getting to know us, it may not be worth spending too much time trying to win them over. Doing so can leave us emotionally and physically exhausted. The exception is when the person plays an important role in our lives or careers. In those cases, it may be worth putting in extra effort.

We must accept being disliked or ignored, and not feel too disappointed or hurt. It does not mean there is anything wrong with us. Our focus should be on finding the right people rather than trying to change how others feel about us. We should meet more people and find those who value us and genuinely want to be our friends.

Silicon Valley might end up being my favorite place outside of my hometown. First, it's a great place to build a career. You're surrounded by talented people who want to achieve something meaningful. Compared to people working in tech in Beijing, people in tech here seem to work only about two thirds as many hours, or even less, while earning three times as much. Second, it's a great place to live. I love the climate here, the sunshine, the dry weather, and all the beautiful plants. There are no mosquitoes. It's also perfect for outdoor activities like hiking, skiing, and golfing.

我需要改进的4个地方：1. 过度准备，以致于错失机遇。不要完美主义。速度有时比质量重要。一个粗糙的开始就是最好的开始。但也不要止步于现在不好的版本，先出一个粗糙版，然后不断迭代。其他人更会关注之后的成果，而且会被你的进步impressed。2.没有足够自信和勇敢去把真正想得到、但是竞争最激烈的事情做为target。3.做事情没有计划，无论是大的步骤还是日常每天的进度，做到哪里算哪里 4.没有清晰的目标，当做每一件事情时，不要忘了这个目标，我们把全部努力都要围绕着这些目标，比如做的事情、认识的人，能帮助我们实现这个目标的不要错过。

Treat everyone with respect and kindness, regardless of their title or social status. Doing so earns us more friendship, goodwill, and support. Of course, I don't think most people treat others with respect because we expect something in return. We do it because we're naturally kind and don't want to cause hurt through disrespect or rudeness.

Stay honest. It's the easiest option for me. I'm really not very good at lying. I'll make something up one minute, forget about it the next, and end up giving myself away.

Taste matters. It shows up in the career they choose, the hobbies they pursue, the life they build, the people they choose to spend their time with, including friends, colleagues, and partners, the things they pay attention to, and the way they think, speak, and carry themselves.

The Three Sources of Investor Edge

1. Information Edge: You have access to better, earlier, or more private information than others.
2. Analytical Edge: You are better at analyzing information and making judgments.
3. Behavioral Edge: You are better at controlling your emotions and making independent decisions.

1. Be selective about whose advice we take. When we're making a big life decision, only take advice from people we actually want to be like and from the very best in the field. Don't take advice from people we don't want to be like or from people who haven't achieved the results we want.
2. Once we've made our decision, ignore the noise. Don't let anyone's doubts stop us, no matter how successful or respected they are. When Elon Musk started SpaceX, even some of the rocket scientists he respected most thought it couldn't be done.

我一直没有意识到我用LLM翻译的英文有多奇怪，直到我用它生成中文。LLM翻译出的文字总是不是我想要写的。LLM做不到“信达雅”，只能做到“信”。一个人遣词造句反应了一个人的性格，而LLM错误的反应了我的性格。

Strategy + Execution

Concensus: AI. Non concensus: who gonna win in AI?

I wish there was a way to know you're in the good old days before you've actually left them.

One of the most important life skills is learning how to judge people. It determines who to work for or with, who to invest in, and who to marry.

At the end of the day, whether it’s choosing a career or a city to live in, the most important thing is the people we surround ourselves with.