Salary negotiation

A good overview of salary negotiation, not necessary for salary only, but for other parts of life as well

Write good code? or write poor code with endless fix?

This is a good question, for me, I try to prefer the first option if possible, but my final answer is it depends

The story in the link is a conversation in an ancient Chinese story [魏文侯问扁鹊] b/w a king and the best doctor at that time. One day, the king asked the doctor: among your brothers, who is the best doctor? The doctor replied: My oldest brother is the best, followed by the second oldest, and I am the worst.

The king was confused, isn’t you the best one?

The doctor said: My oldest brother can notice & fix the problems perfectly BEFORE it becomes a problem, the patients think he does nothing and only our family know he is the best.

My second oldest brother can notice & fix the problems WHEN it first appears, the patients think he can only fix the small cases.

I can only notice the problems AFTER they become serious, the patients are impressed by my fancy treatments and risky operations, therefore, everyone knows me

Which one do you prefer :-)?

Resources: Made with ML

This material is super awesome! Contains all the things I wish to learn connecting ML and engineering! Other than concepts, the example codes are great, no BS, I can just reference, copy, use them right away and it WORKS!

机器学习：软件工程方法与实现

The direct translation is “Machine Learning: software engineering and implementation”. Surprisingly good and practical book that introduces the ML work in reality [including the software engineering side with main focus on non-deep learning areas] and offers the high level picture of different aspects in reality, strongly recommend if you can read Chinese

Designing Data-Intensive Applications

Maybe one of the best tech books in recent years, must read for serious players in data science / ML engineering, period

Sapiens: A Brief History of Humankind

I have met many people said they have read this, maybe it is too famous lol I have read it and finished the Coursera course from the same author.

The single view that impacts me most is: humans are living in both subjective and objective world. In fact, money, nation, values etc are all constructed by our minds. They are meaningless if no one believe that

Resources: Sebastian Raschka

In short, this is the first place I will go if I need to review the topics about ML/DL. I particularly enjoy the clear & easy-to-follow logic. Plus, there are many code examples!

Resources: Full Stack Deep Learning

Cover almost everything I want to learn about ML workflow, strongly recommend if you need to worry engineering part of the ML as well

Coursera: 史記

「從歷史來看，決定人生成敗的不過就是兩樣東西，第一是運氣，第二是自我要求。什麼是運氣？往大的方面來說，生在什麼時代是運氣，生在什麼國家是運氣，生在什麼家庭也是運氣。往小的說，出門左轉或右轉，接不接一個電話人生可能就不一樣，運氣對每個人就是這麼重要。但是運氣不可測也不能控制，我們所能控制的只有自我要求，如果我們真心想成為什麼樣的人，是真心不是發夢想想，就必須不斷奮鬥努力讓自己具有足夠條件，等待偶然的機會」

這段課文感受良多，沒有人會希望運氣不好，人品好做好準備也是會有倒楣的時候，但個人只能控制自我要求，想成為什麼境界的人就做什麼事

How to Fail at Almost Everything and Still Win Big

I knew Scott Adams because of his cartoons. This book shares many practical “principles” that impact me a lot, eg:

Goals are for losers; systems are for winners [the world is so random, it is hard to set the goal to become millionaire next year but we can do what millionaire will do]
Passion is useless [sad but true, no one is interested in how passionate Steve Jobs is before his success]
A combination of mediocre skills can make you surprisingly valuable [few top 10% skills maybe better than a single top 1%]

Atomic Habits

I found this after I read “systems are for winners” from Scott Adams’s book. If I need to summarize the whole book, below are my takes:

Find the habits you want to build and decompose until it is almost impossible to fail
- eg: fitness: instead of going to Gym room 4 times a week, just do at least 3 push up per day
- eg: reading: instead of reading certain pages, just read at least one paragraph per day
- eg: programming: instead of doing 10 Leetcode during weekend, just do the daily challenge
After you identify the habits, do them for 21+ days
Don’t cultivate too many at a time

Video: DO NOT SPLIT 不割席

As Hong Konger, it doesn’t make sense if I tell you I have no views on what happened over the last few years :( It is sad that the conflicts will probably last for decades b/w the people and you can do nth about that.

Over these 2 years, I always have these questions in my mind:

The conflict could be resolved easily from the beginning, how is it possible that the government picked the worst possible options and escalated the situation to irreversible?
At what cost people are willing to pay for what they disagree on? It is naive to deny/justify the violence that happened. The issue is after you have tried all the possible peaceful means and got no response, it is encouraging people to go to extremes, especially if they are not the minority. People always talk about the events since 2019 July but they don’t discuss why the government gave no response after a peaceful protest with 2 millions+ [aka around 1/3 of the city population].
In my view, this is removing the brake from the movement because the peaceful supporters cannot propose another way but sit back. When did the stupid government ask why do you not condemn the violence? The answer is simple: it was you who told us peaceful marches did not work, it was you who removed the brake when we asked you not to.
Who should be responsible? Well……it is like……someone turns on the lighter and burns the house, will you blame why the house is flammable when everyone asks you not to? It is sad that hundreds if not thousands of protestors may be in jail at last for the disaster that could be avoided from the beginning.
I don’t hold any particular “political beliefs”, I don’t like the bundle package because the narrative makes us give up our rational thinking. As long as we are solving the problems but not solving the people who raise the problems, any idea should be evaluated to considered and evaluated.