2022 Annual Summary#
- Rivals come and go, game stays forever
Rivals come and go, the game remains unchanged
Annual Vocabulary: Game Theory#
- If your dreams don’t scare you, they are not big enough.
If I am a little scared, that must mean I am doing the right thing.
I feel like I read relatively few books in 2022, and I no longer consciously mark what books I've read on Douban. I feel that recording my reading list or some notes on OneDrive is sufficient. This year, due to the pandemic, I unexpectedly rewatched many movies and TV shows, most of which were consumed during the times I couldn't go out. Among them, I particularly enjoyed "Attack on Titan Final Season" and "The Godfather Part 1." This year, I consumed about 30+ movies and shows, but it seems that apart from technical books, I only read about 20 books that were particularly enriching, which is a bit embarrassing xD. I feel that reading e-books doesn't give me the same feeling as reading physical books at home, and having physical books in a rented place doesn't provide much security, yet I have accumulated many e-books, research papers, and some technical articles. A fantastic experience this year was working remotely from my hometown for half a month; using my home’s 4K setup and desk felt very secure, and I even regained the lost afternoon naps, but I still found being at home too boring 😂.
Aside from these special times, my life has changed quite a bit. I gradually kicked the habit of sleeping in on weekends to go hiking and exercise, feeling that I have much more time and am not as easily fatigued as before. Entertaining myself, programming, reading, and thinking have become part of my daily routine in 2022. Doing the same things constitutes most of the joy in my life, and chatting with friends has ranged from technology, programming styles, stock trading, game theory, major financial events to driving, movie plots, cooking styles, and food. I view the happenings around me with George Carlin's attitude, managing my emotions and calmly observing what occurs around me. I have become less fond of arguing, directly giving up on helping others and respecting their fates.
Movies & Series#
Tier 1#
If I had to choose the most rewarding movies and series I watched in 2022, I would pick "Attack on Titan" and "The Godfather Part 1." The Tier 1 rankings are "The Godfather Part 1," "Attack on Titan," "Chainsaw Man," "Scent of a Woman," and "The Great Era," while the rest can only be Tier 2.
-
"The Godfather Part 1" -- 🌟🌟🌟🌟🌟
Although I have watched "The Godfather Part 1" multiple times, it still offers new insights. Every time I watch it, I discover something new in Victor and Mike's ways of thinking. Their personalities and working styles differ, but both are cautious, calm, and have god-like crisis management skills. Victor originally did not want Mike to become the next godfather, but Santino couldn't remain calm, Tommy couldn't handle crises well, and Fredy was just a weakling, which determined that the next godfather would definitely be Mike. One of my favorite plot designs is when Mike has the baker Russ pretend to have a gun to scare off enemies and protect the old godfather; this scene perfectly illustrates that Mike is someone who knows how to utilize his limited resources to turn the tide. -
"Attack on Titan" -- 🌟🌟🌟🌟🌟
The plot design is unbeatable. Many people criticize "Attack on Titan" for having a poor ending, but I still find it very enjoyable. Why do people living in a three-dimensional world find it unacceptable to have an ending that is not perfect? It's too abstract. -
"Chainsaw Man" -- 🌟🌟🌟🌟
You can't create such a psychotic anime without ten years of mental illness experience; it was a refreshing surprise. The author managed to suppress a mental breakdown at the last moment to draw a normal ending, which is truly inspiring—I cried 😭. The most impressive aspect of this anime is that the mental illness didn't make me feel disgusted, which can be said to surpass works by authors like Anthony Burgess. Actually, I haven't been to school 😭. -
"Scent of a Woman" -- 🌟🌟🌟🌟
I just think it's good, can't explain why, but I can't agree with the colonel's values; Americans are still too direct. -
"The Great Era" -- 🌟🌟🌟🌟
"What kind of ending is worthy of all the ups and downs along the way?" What moved me is that helping Fang Zhanbo defeat Ding Xie was not luck, but timing. When Fang Zhanbo finally visits Ding Xie in the mental hospital, Ding Xie is no longer the brute who could threaten others with force; he no longer has the strength to harm others. The ending also seems to hint at Hong Kong's transition from triad rule in finance to a new era of new money in finance, represented by people like Fang Zhanbo who rely on technology. Besides this period, I also watched quite a few finance and game theory-related movies and documentaries. Now that I'm old enough to enter the gambling city of Macau, I can't go play due to the pandemic, so I can only watch movies. I feel that the reduction of movies like "The Gambler" and "The God of Gamblers" is primarily to create a better educational atmosphere; we can't let such movies influence the next generation to think that winning 37 million with 20 bucks is easy like Chen Daozai. I feel there is another reason for the reduction of such movies.
The final boss in most games plays "Chor Dai Dee," and a small portion plays Texas Hold'em. I thought about why this is the case: because these two types of cards involve the most participants in the game, making it easier for screenwriters to design plot twists and game theory. However, the techniques for winning in the plot are hard to appreciate; it's either cheating or some special ability, oh, and the ability to produce two decks of cards.
So why have these types of movies decreased? Because Hong Kong's financial market is no longer dominated by speculators, and the card-playing style has shifted from observing opponents' behaviors to a quantitative approach using mathematical modeling tools. Hong Kong's financial market has seen the emergence of quantitative finance companies, making game theory no longer simple. "The Great Era" also carries this meaning, but what I dislike is that the screenwriter's pacing is too slow; Fang Zhanbo suffered for 99 episodes, but his revenge succeeded in just one episode, and even his career improvement was compressed into one episode, making the subsequent revenge feel a bit abrupt.
Tier 2#
There isn't much to say about Tier 2; it mostly feels like mindless fun movies, just for a laugh. Watching too many of them isn't great. Although "The Batman" has a poorly designed plot, the cinematography is well done and feels good, along with the soundtrack.
Games#
This year, I didn't play many games, except for "Elden Ring." I completed "God of War 5" and "Red Dead Redemption 2" by watching videos. I didn't play "God of War 5" at all, but I played a bit of "Red Dead Redemption 2" and found the pace too slow, so I watched the video walkthrough. I didn't play "Cyberpunk 2077" for long either, as I felt the story was too slow and lacked patience, and I was too lazy to even watch the video walkthrough.
- Elden Ring
"Elden Ring" is my first truly completed Souls-like game; previously, I just breezed through "Sekiro" 😇.In terms of playtime, I spent 60 hours playing this game, but the video guides I watched outside the game far exceeded 60 hours. Boss fight strategies, how to start with max-level weapons... these were all my research directions. I feel "Elden Ring" is easier than "Sekiro"; during my first playthrough of "Sekiro," I was tortured by the red demon at the start and directly looked up a guide. However, the music and storyline aren't as excellent as "Sekiro"; the impact sounds and such feel a bit lacking compared to the satisfying sounds during combat in "Sekiro." I mostly used guides throughout "Elden Ring," and when I reached the final boss Radagon, I just knelt down and called for help. Even so, "Elden Ring" gave me many "Ah Ha Moments," reminding me of the days spent in the lab playing "Sleeping Dogs." I feel that in the GOTY competition, I stand with "Elden Ring" over "God of War 5"!
Technology#
2022 was probably the year I improved my technical skills the fastest, as the amount of data I needed to handle was greater than ever before. It was also in 2022 that I delved a bit deeper into distributed systems; before this year, I might have read a few papers and written a couple of demos, but my understanding was still shallow. Starting from March, I began to engage with distributed applications and frameworks, papers on TiDB/spark/Akka, and became a contributor to several open-source projects. Overall, I am quite satisfied with this year.
-
Go language
After using Go for two years, I feel it might really be a somewhat flawed language. Perhaps I was too inexperienced before or the business was too simple, so I didn't feel significant bottlenecks or constraints. After the release of version 1.18 with generics, I still feel they are quite useless; this generics feature is really not practical. After using it once, I found it too ugly and didn't want to use it, but I still need to find ways to contribute some generic code to a few open-source projects, piling up a mess in others' projects. This year, I read a lot of Go source code, and my biggest takeaway is that Go's design patterns are infinitely weakened in language design. In my previous company, using gRPC was essentially a tool that helped you create interface layers and compress during coding, but I didn't understand this layer before; many times, it was just for the sake of using it. Under the interdependencies of microservices, using proto files for data format exchange is also convenient, but there is another layer of reason to obtain an abstraction interface. However, many times it feels like design patterns are used just for the sake of using them; if optimization is needed, it often requires refactoring. If too much code is changed, the reviewer can only say LGTM (Let's Gamble, Try merging). -
Rust language
Rust is a new language I attempted to learn this year, and it is probably the most difficult language I have ever learned, especially with the combination of generics and lifetime declarations; I feel like I'm about to develop reading disorders. What surprised me is that the relatively unique features in Go, such as channels, goroutines, and duck-type interfaces, have similar implementations in Rust (channels & traits). Rust is much more complete than Go, but it's also much harder to write. Writing a simple HTTP server was quite challenging; most of the time, I was battling with the compiler. Compiling Rust is also a torment, with extremely long compile times and the computer's fan roaring and heating up at the start, which is really a downside, but it doesn't make me dislike it. However, I still wouldn't recommend friends without programming backgrounds to learn this notoriously difficult language with an exceptionally steep learning curve right off the bat. -
Spark & MapReduce
This year, due to business reasons, I built a Spark framework from scratch. The business shouldn't require very complex resource scheduling, so I just set up a standalone mode Spark, and it felt sufficient; it's essentially a tool that can execute very aggressive SQL. At that time, I used TiSpark to connect to TiKV & TiFlash to read data, and it felt pretty good, but later I found TiFlash to be a bit useless, solving no problems at all. The more I used it, the more I felt that TiDB was quite mediocre; although it is an industry leader, when the data volume exceeds 1 billion rows, it becomes basically unusable. You either need to add a massive number of machines to meet the demand or purchase their solutions, which is too outrageous. If there are no requirements for distributed transactions...
Reading#
This year was probably the year I read the least, having only read a dozen books? Most of them were still technical books, and I crammed through them after September. "Siddhartha," "The Mathematics of Poker," "Capitalism and Freedom," "Norwegian Wood," etc. I read "The Mathematics of Poker" because I saw friends in quantitative finance say that this book can be considered a pioneering work in modern poker theory. A deep understanding can enhance one's grasp of probability theory and game theory. They said that companies like Jane Street and Two Sigma pay great attention to making rational and scientific decisions during interviews, and poker can be a reasonable testing method, allowing you to gamble all day in a game that only counts points, not money, to see how many rounds you can win. I read "The Mathematics of Poker" while practicing, and I also finished watching "God of Gamblers" and "The Gambler."
Since childhood, I felt I was better at "Chor Dai Dee," which is completely different from Texas Hold'em; in Texas Hold'em, the winner takes all, while in "Chor Dai Dee," you look at how many cards are left to determine small losses versus big losses. Small losses can still count as positive points, and if it's a losing situation, you can try to position yourself in a not strictly dominated position. After reading this book, I downloaded a QQ gaming hall and seriously played hundreds of rounds, constantly reviewing and summarizing, striving to make every choice rational. When I opened the QQ gaming hall and saw my old non-mainstream ID, my score was -172 points, with a win rate of 53%. After finishing the book, my score was -463 points, with a win rate of 46%. (I was about to throw this book away.)
My summary is: just like the interview method at Jane Street, they generally give you a mathematical game theory problem without a standard answer. No matter how excellent your strategy is, as long as I know your strategy, I can come up with a better one. So interviewees generally need to list and summarize all situations and adjust their gameplay based on different opponents/strategies. The multi-party game results form a Nash equilibrium.
I feel that stocks/funds related to game theory are all similar; ordinary people have little information to participate in these games, but the most important thing in game theory is to know how your opponent will operate. Therefore, generally speaking, you need to strictly execute your plan, review and summarize, and then find the strategy with the highest win rate.
This year, I watched quite a few stories, and I particularly liked the story of Jensen Huang from NVIDIA.
In the early days, Huang had a hellish start; his father took the whole family to work in a chemical factory in Thailand, a politically chaotic country where children had no good childhood. Huang's father felt it was bad for the children, so he asked his American brother-in-law to help arrange schools for Huang and his brother. The brother-in-law's English was poor, and he had no connections, so he could only find a relatively bad school. Huang's first roommate was a 17-year-old drug addict, and the differences in physical development during puberty were reflected in their intimate physical interactions. After a significant memory recovery process, young Huang and his brother learned to clean the toilets for their older brother, do homework, and fetch water. They also pretended to be bad kids by climbing trees to steal candy and smoke. Years later, when Huang reflected on this period of life, he said it was the best teacher.
The Kentucky juvenile detention center could be considered Huang's first dragon arena. The insights gained during these years taught him what strength is, why people need to fight, how important survival is, etc. These became important chapters in his survival manual for the hellish mode of entrepreneurship later on. Struggle is where greatness comes from.
Another event this year was the former genius trader from Jane Street, SBF, the founder of FTX, whose assets plummeted by 94% in five days due to leveraging. Although a 94% drop is still a level that ordinary people find hard to reach, it seems unlikely he can recover. I feel that SBF, who initially appeared to be faultless, seems to have grown up in a greenhouse, overly protected, and won't have the same journey as someone like Jensen Huang.
Huang said that the pursuit of turning the impossible into the possible is the charm of this industry itself. This is why Huang is still passionate about going to work today. I often wondered why Buffett, at such an old age and with so much wealth, still works and doesn't retire. I think it might be for similar reasons: "Rivals come and go, the game stays forever." Rivals come and go, and the game with rivals is the truly interesting part.
I think this is also why the darker forest in blockchain, despite its dangers, attracts people to engage in the game.