Q&A: How do I keep from getting set up?

From Day Class: The Inventor’s Handbook

Reader Ji Xiaoshuai: This one reminds me of the website TV Tropes that Mr. Wan mentioned earlier. It feels like Europeans and Americans like this kind of systematic sharing or summarizing. Does Mr. Wan know of any other areas where Chinese people have this kind of systematic sorting or summarizing?

Mr. Wan replied -

TV Tropes is a book of movie and TV drama storylines, which is equivalent to a “library of plot functions” that you can directly call from when you write a drama or a novel. The reasoning is that the story is not only structurally fixed, but also in terms of specific elements, so the audience will still buy the story with a slight change of clothes.

This kind of systematic summary of a field, with the help of a literary term is “deconstruction”: what seems to be a whole thing is actually just a bunch of combinations of common elements, not so mysterious.

China is still an emerging force in modern technology, but there is one area that has long been deconstructed and understood by us, and that is examinations.

If you are applying for a doctoral degree in the United States, universities usually require you to take a test called the GRE, which is organized by a private but authoritative company called ETS, and is designed not to test professional knowledge but to test intellectual ability. More than two decades ago, the GRE consisted of three sections: language, math, and logic, but now it has been changed to include several sections such as analytical writing, opinion questions, response questions, vocabulary, measurement, and experimentation. This GRE has been deconstructed by the Chinese before.

Logic and math are the strengths of Chinese science and engineering students, nothing much to say, the key is language.

More than twenty years ago, the main purpose of the language test was to test your vocabulary, and the main items were analogies and fill-in-the-blanks questions using cold vocabulary, just like the Chinese use out-of-the-way words in couplets. the idea of the ETS was very simple, these words were not the everyday language of the common people, and if you even knew these words and could still use them proficiently, then it meant that your reading was extraordinary, and that you must have a high level of intelligence. Then you can imagine that the default test object of this kind of questions are Americans, Chinese candidates must suffer: even if we love to read can not read all day cold English literature, which is not the same as forcing foreigners to read the literary language?

However, ETS never expected that Chinese test takers would turn the GRE language into a strong point. The reality is that English vocabulary is limited: there are only so many interesting words that are worth using on the GRE. We first came up with the ‘GRE Vocabulary Book’ and then developed the use of computers to do questions to train quick responses. As long as you spend hundreds of hours memorizing these words and adapting to several question types, these are all easy questions. As a result, Chinese test takers are vastly outperforming American students, even though their actual English reading skills are very limited.

Of course, ETS has also thought of some ways, such as keeping the past exam questions secret, but these are difficult to defeat the Chinese exam researchers. So then ETS had to change the GRE question type ……

I think there’s something to think about here, so let’s take another example of the KJU.

The American sinologist Benjamin Ehrman has a book called A Cultural History of the Section Examination in Late Imperial China (Chinese edition, 2022), which was recently discussed in an episode of the Kulozo podcast. One of the big points of the book is how the business of the imperial examinations became more and more involutional.

During the Sui and Tang dynasties, because there were so many sons and daughters of official families that the court had a hard time placing them in positions, the imperial examinations were mainly able to provide a reason for elimination and selection. By the time the Song Dynasty came around, because almost all of China’s aristocracy had been killed off in the previous An Shi Rebellion and Huang Chao Uprising, commoners had a chance to become officials, and the imperial examinations did select a number of talented people. But further on, when the imperial examination became an official system, and the children of various families grew up to take this examination as their only career to do, its role of selection was already very small.

First of all the test must be ‘standardized’. First of all, it is certainly not a test of what new ideas you have - what a joke, even if it is a test of policy, the court’s general policy can not let the candidates casually discuss. Secondly, the content must also be limited, there is a difference in the level of education in different parts of China, so you can’t come up with bullying questions. As Mr. Luo Zhenyu said, the former college entrance examination essay questions can not appear ice cream - why? Because candidates from remote areas may not have eaten ice cream! It’s not fair to people for you to take this test. The imperial examination is also like this, needless to say the twenty-four histories of the ordinary people do not have home, even the “Five Classics” is not the average person can afford to learn, so the Ming and Qing imperial examination only test “four books”, and to Zhu Xi annotated version shall prevail.

Candidates look at the scope of the question is so narrow, this is not the GRE? The content of the four books is so much, and does not allow you to have any new ideas to play, then you just need to give the question bank and model text are memorized, this is a simple test.

But the test must be eliminated, the court out of no new questions can be issued, which made the eight-legged essay. Since the content has been out of all, then we have to continue in the form of volume: eight-legged essay on the form of the text of the strict requirements of the battle, rhyme and so on, you will write can indeed write a very beautiful, but also considered a kung fu.

And in the late stage, the candidates to the eight-legged essay also deconstructed. Especially to the temple examination - that is, to determine the rank of the final, in front of the emperor’s face to answer questions - this level, the presence of all the eight-legged essay doers, how do you distinguish between high and low?

In fact, the temple examination test is mainly …… calligraphy. The article are almost the same, it must be required to roll absolutely neat, there is a little error on the elimination. But the volume of all the same neat, then it can only be calligraphy beautiful stand out. Now we look at those first-place exam papers are marveling at the people write really good, that is called “Pavilion Style”, with the same printing, really great! Worthy of being a scholar! I do not know that he took the first prize depends on this ability.

Shouldn’t the real selection of talent be based on thought? However, the institutionalized examination system can only from the test content to test form, and finally to test calligraphy, more and more emphasis on useless things. Needless to say that creativity can not even begin to talk about, even IQ is not important, the most important thing here is willpower.

Then you say, willpower is also a kind of ability, which is also considered a kind of selection, right? In the past, some scholars thought that the imperial examination at least provided a way for the civilians to rise, which was still meritorious. But the new generation of scholars think that role is very limited. Ancient reality is that the average family can not afford a full-time scholar, and the imperial examination this thing is not a real skill, but it is incomparable time-consuming and laborious hard work, so the late examination is the family’s condition is good.

In fact, the court from the beginning to understand that the imperial examination can not select the excellent talent of the state, but the results of the game can only be so. In fact, the court does not really need what smart people to govern the country together, what a joke, the Qing dynasty is the Manchurian people can let you govern it - the court needs smart people do not cause trouble. The biggest role of the imperial examination system may be the domestication of smart people: you again smart, read at home for fifteen years Zhu Xi and then practiced twenty years Pavilion body will be all the whims and bravado are smoothed out.

This is why I hate calligraphy so much: calligraphy is a test of obedience to start, to walnut carving on the end of the project to play storm in a teacup scroll, is the destruction of Chinese children.

*The truth of this is that sets do work, but if too many people use them, and everyone studies them and understands them, they’re bound to go to involution. *

Isn’t that what’s happening with the college entrance exams now? There are no real difficult problems, you are not expected to engage in invention, in order to both reflect the fairness and elimination function, the test is full of biased and troublesome questions, focusing on the test of the candidate’s willpower.

The value of the college entrance exam is to reward effort, not to select talent. It appears to be fair, but I think it is actually more favorable to the rich. It’s good to have a set, the rich can always rely on technology, you find a famous teacher to guide you and play the questions to understand them, maybe you can improve your performance. For the poor, the high school exam expects you to work hard, not mutate.

If there are a lot of forbidden areas in a field, people are afraid to try new elements and can only work on fixed sets, the result is bound to be an inside roll. You look at the current costume drama, the other day I read a microblogging hot search called “2023 four big must explode TV series”, found that these four plays are not only all romance idol drama, and the poster is a man and a woman, all with the wedding photo like …… I really want to ask a question. Are we Chinese not worthy of better programs?

The rational choice is that you should roll first if no one else has started rolling. But if everyone else is rolling, you can’t go against the grain and probably have to roll with it.

Evolutionary Ideas 2: A List of Sets

Reader David Tam: How can we, as individuals, not be affected by these sets? Or minimize the impact of these sets on us?

Reader Yuxiao Gu: Sometimes it’s hard to break “social norms” even when you know the new routines.

Wei Wei Gang replied -

The thing about scams is that if you can tell that it’s a scam, and the merchant is promising something that it can’t do, then you can immediately tell that it’s not only not worth the money, but it’s also probably harmful. But for the set-up, because it is just a psychological device, even if you can see that there is a set-up here, you can’t tell if it is not worth that much money. What’s more, as business becomes more and more advanced, various merchants use sets, and that makes it even harder for you to choose.

The way to spend rationally is to shop around, study product reviews and user comments before buying, and especially not to buy something you don’t need. But as we’ve talked about, System 2 is very energy intensive, and if you have the work to spend half an hour researching which laundry detergent is more cost-effective, why not use that half an hour to learn something new and put your mind to earning money instead of saving it?

So sometimes it’s hard to be confused and forget about it. It’s like when your kid asks you to buy a toy for a cute reason, and you know that reason is unscientific, but since your kid is so cute, why not buy one? It can be rational to spend a little money for the shopping experience.

Evolutionary Thoughts 3: Prove Yourself

Reader Green Lantern Listening to Rain: Mr. Wan, you said these methods are based on the perspective of the boss or the company, then for the team or individual within the company, I provide tools but need to change everyone’s habits, people are emotionally resistant, in addition to increasing the degree of professionalism and labor illusion, even with the addition of these methods, it is still difficult for people to accept, how to increase the trust of others, how to make things can be carried out smoothly How to increase people’s trust and how to make things go smoothly?

Wei Wei Gang replied-

It is very difficult. A consumer’s willingness to buy something can be all about feelings, because spending money is a very simple action. But you have to say it’s too painful to learn a new tool from scratch. And the perception is different. Take the recent AI-assisted programming, it’s obviously easier to work with, but there are a lot of programmers who are just resistant to it now, uncomfortable with it, unwilling to use it. I’ve heard of programmers who were laid off just because they refused to use AI and were no more efficient than newcomers.

To get people to want to use a new tool, you’d better be able to prove that it can increase efficiency by say ten times. You have to fully understand the difficulty. Sometimes even if it’s a tenfold difference, you end up having to switch people.

From the Daily Lesson: Laxity is a race to the bottom, but harshness is a dusty business

Reader Attorney Wu Feng: The cognitive issue of custody mentioned in this article, if left to the judgment of AI, will it be the same as the human’s choice?Will the AI’s answer twice also contradict itself?

Wan Wan Gang-

The situation of AI is now clearer, divided into two categories: ‘specialized AI’ and ‘language models’.

Dedicated AI is to train a model specifically from scratch using data from that particular field. For custody issues, it is to put together and train who the children were awarded to after the divorce of a couple in history, how well the children grew up later, and what all the indicators of the couple were at that time. Its output is not affected by the way the question is asked; it’s more rational. Because the human brain can’t take so many factors into account at once, the model will think differently than the human brain, and may even go against our intuition - but it’s actually better, in practice.

The language model AI, or GPT as we use it, on the other hand, has a human-like mind. The fact that its results may be more accurate when you ask it to think step by step about the same question shows that it has human-like intuitive errors. The answer it generates has a lot to do with the way you ask the question.

That’s why I think questions like custody are best left to dedicated AI. but that’s a problem for you guys in the legal profession because you’re asking for a human to have control, and you need explanations that the human brain can understand. There is no good way to do this right now.

---