# AI Won't Close the Gap Between People, and 7 Other Notes

- Author: Rory Cai (https://coiggahou2002.github.io/)
- Published: 2026-08-09 (Asia/Shanghai; 2026-08-09T06:00:35.000Z)
- Language: en
- Canonical: https://coiggahou2002.github.io/blogs/recent-thoughts/
- Chinese version: https://coiggahou2002.github.io/zh/blogs/recent-thoughts/

**#1**

AI won't narrow the gap between people. The reality is the exact opposite.

It amplifies people's creativity and scales their judgment, but it also amplifies their stupidity and bias.

**#2**

If you're just starting your career, learning to be a good cog, grab credit and dodge blame in a place riddled with big-company disease isn't the only path. Being thrown in the deep end on a small team is not necessarily a bad thing. It may be more tiring, but it builds a strong sense of ownership and a view of the whole picture.

Right after graduating I was in exactly that kind of environment. At the time it felt exhausting and hard. Looking back, it's what gave me a well-rounded skill set, and the foundation for catching better opportunities later.

You can't connect the dots looking forward; you can only connect them looking backwards. So you have to trust that the dots will somehow connect in your future.

**#3**

If an environment isn't right for your growth, leave, boldly, and find one where you can really let loose. Charlie Munger put it roughly like this: I never waste a second trying to change other people, but it's never too late to change yourself.

**#4**

There's a line from *Ashes of Time* I love: everyone goes through this stage. You see a mountain and want to know what's behind it. I'd like to tell him that once you cross it, you may find nothing special on the other side, and looking back, this side might seem better. But I know he won't believe me. With his temperament, he won't be satisfied until he's tried it himself.

Growing up is a steady process of disenchantment with the things you once worshipped and idolized. Along the way you meet a lot of things and people, and find they're nothing special after all.

**#5**

The world has always had plenty of problems. Before large language models, a lot of people spent their time looking for good answers.

Answers aren't scarce anymore. What's scarce is the ability to ask good questions.

But I've noticed that many people can't even be bothered to ask "why".

So curiosity and the urge to explore matter more in the age of AI.

**#6**

A lot has changed in software engineering. We used to stress abstraction and reuse so people would write less duplicate code, because writing code was expensive. Now that writing code is cheap, do those rules still need to be followed?

I think reducing entropy is still one of an engineer's most important jobs, but we don't have to follow those dogmas as strictly as before.

The bottleneck in software engineering today isn't implementation time. It's review time.

People with the skill and judgment to guard code quality are scarce, and review bandwidth can't keep up with how fast code is being produced.

**#7**

What is knowledge? Knowledge is information used to produce something.

The ancients answered this long ago: learning by doing is the most efficient way.

With AI, if you're willing to push through with gritted teeth, you really do get tougher.

**#8**

For someone who moves fast and never has enough time, it matters a lot to find a quick meal that's nutritionally complete, low-GI, doesn't cause a carb crash, leaves you about 60% full, and can be eaten within 15 minutes.

---

Originally published in Chinese on my WeChat account 罗里戴行思录: [read the original](https://mp.weixin.qq.com/s/Izu8XXMdc9PwLwQv_xiiDQ)