Attracting Stars: Thinking Like a Scientist

 

---

Attracting stars

The most powerful skill is the star-absorbing method, learn what you encounter, and learn what you learn.
 

Metacognition = patience + deliberate block practice + feedback on key points of thinking/mental/body process + regulation

Our general cognition is external cognition (world, things, subjects, problems), in order to do that thing better.

Metacognition is the cognition of oneself (learning process + refinement, thinking process + refinement, method steps + refinement, emotional process + refinement, mental process + refinement, cognitive process + refinement, body process + refinement), is for thinking upgrade.

• After every unexpected event, it is the beginning of metacognition
• After a long period of mediocrity, it is also the beginning of metacognition
• Learning to use new knowledge is also the beginning of metacognition

What is the difference between humans and other animals if they do not have metacognition ability.

Most animals behave instinctively, coded in their genes to make them react in a certain way. To change, it may take more than ten years or even hundreds of years of evolution.

We humans are different. We can constantly review our thoughts within a limited life, and then change our behavior strategies. In a sense, we can even evolve every day.

The brain is layered (the rational brain is higher than the emotional brain and the instinctive brain), and the high-level brain can control the low-level brain. The closer the high-level brain is to the real me, the more I can control the brain, instead of the brain controlling me in turn. , this is the most effective means for human evolution.

Most people actually don't understand their own thinking process, thinking principles, and thinking methods; it is because most people have never seriously examined whether their thinking results are indeed correct and reasonable.

When I see people with high IQs, I feel that they are born smart and can easily solve complex math problems. Rather than feel that the other party actually knows the deeper principles behind these numbers.

Seeing people with high emotional intelligence, I envy the other party's social skills, but the real essence is that the other party can understand the deep expectations of others.

The reason why metacognition is important is that it not only helps you understand yourself, but also allows you to understand others better.

Scenario 1 : Time always seems to fly after the past.

• Normal people: I agree, I have similar feelings.
• The Metacognitive Brain: Why Do You Feel This Way? Why even ignore this fact? What am I missing? What else do I need to correct?

Scene 2 : Someone scolds you.

• Normal people: A person with a strong personality will definitely fight back after being hit, until the stronger person loses everything; a cowardly person will always run away, and nothing will be resolved in the end.
• Metacognitive Brain: I got angry after being hit, does this emotion help me? Only doing stupid things, suffering, regretting, things getting troublesome, relationships getting bad, then let me listen in a structured way, divide his words into facts, emotions, expectations, admit the fact of screwing up, appease negative emotions, and act as the other person needs. expectations.

Scene 3 : Husband and wife quarrel.

• Normal people: Just talk about superficial reasons. If you don’t understand yourself, you rush to find the other party. The final result may be that the other party has raised a few different opinions, which has already irritated me, and I don't want to continue communicating.
• Metacognitive brain: Not only expressing intuitive ideas, but also explaining the ins and outs of this idea, and can express it layer by layer, so that the other party can understand deeply.

Scenario 4 : Problems that cannot be solved for a long time

• Normal people: A is angry in love, and B thinks that A is angry because of X, so he repeatedly explains, persuades, and even teases around the topic of X... Sorry, usually doing so may only make A more annoyed, Angrier, why?

• Metacognitive brain: This incident has been encountered many times, maybe this time I should look for the key elsewhere... Where is the key? In psychology books: For advanced animals like humans, some emotions are opposite, and it is almost impossible to coexist. For example, you can hardly be both happy and miserable, excited and depressed, or bored and interesting... So, the solution is quite simple and straightforward: if you can make the other person feel extremely happy, TA can't be miserable, Angry, bored, helpless... Have you really spent enough time thinking about this almost the most important thing? ——What things and events might make the other party extremely happy?

Scene Five : Reading a Book

• Normal people: search for various concepts and concepts, pursue useful and interesting ones, and only read what is useful to them.
• Metacognitive brain: read and compare yourself and the author, perceive thinking, and read every word.

When our metacognition ability is not strong and we cannot think deeply about certain concepts, two extreme situations are prone to occur: total acceptance or total negation.

People with strong metacognitive ability can pay attention to their own thinking and thinking all the time when reading, and reflect, verify, revise, and upgrade their own thinking and thinking all the time.

When you read someone else's writing, you are comparing the other person's way of thinking with your own. See which aspects are the same as what you think and which are different from what you think.

When reading, what you read is not only the text and the principles explained in the text, but also the author's "way of thinking", the difference between the author's "way of thinking" and your own "way of thinking", and , if the author's "way of thinking" has merit, what adjustments should be made to his "way of thinking"?

After reading a book on probability theory, most people may not be able to pass the test even if they take a test, while a very small number of other people become scientists—because they have improved their way of thinking, and since then they can "think like a scientist" "...

Scene Six : Shooting practice.

When practicing shooting, you can be aware of whether your shooting is left or right, but generally you cannot be aware of the state of your fingers (especially the index finger). Not to mention the knees, waist, shoulders, elbows and a series of parts that affect shooting. You have been shooting for a few years and still have a very low hit rate, because you lack awareness; you sometimes wonder why you are lazy and can’t find the reason, and you can only lament that the hand feels bad. In fact, you just want to be aware and cannot perform. Come out; even if you take the time to practice your shot it often doesn't work because your awareness isn't fine.

Scene 7 : Typing practice.

I am typing. When I type, the little finger of my right hand is not very flexible. It does not reach the "P" key smoothly, and I feel this. It's a little stiff and a little short. But I control it to perform the action. After dozens of practices, it became flexible. Although it hadn't made much progress in the previous ten years, because I gave it awareness, it broke through the ten-year barrier in a few minutes.

Scene 8 : Do questions.

• Normal people: I can't do this question.
• Metacognitive brain: When I saw this topic, I made an association first. What do you think of? Think of a formula, or a fixed problem-solving routine. What's next? I started to apply this formula or routine. But then it turns out that it doesn't work. So I got stuck.

Scene Nine : Learning.

You have to check yourself often. You are doing something according to a certain strategy, such as using a certain method to learn English, and you will ask yourself: Is there any way I can do it better? Can I help myself a little more? Is there any other way? Maybe you used the method of memorizing 20 words a day for a while, and felt that the progress was too slow, so you took the initiative to change to the method of memorizing 500 words in a dedicated chunk of time every day. You are very sensitive to methods, rather than going all the way to the dark.

Scene ten : Ask questions.

• Normal people: I don’t understand this question.
• Metacognitive brain: There are many steps in the problem-solving steps and analysis process. Don’t you understand all the steps? it's out of the question. There must be some core steps that I can't understand, and I need to be specific to the nth step. For example, the xx option of this question, if you want to analyze it clearly, do you need to understand step A, and then calculate it according to the xx method? But when I calculated in the second step, I found that it was different from the answer. So which step did I think wrong?

Different questions reflect that the former is accustomed to a state of fuzzy thinking. When his thinking is very fuzzy—reflected in the fact that his questions are very fuzzy, and he does not precisely locate the knowledge points and analysis processes that he has not mastered— He doesn't feel uncomfortable, he can accept this state very well.

If you are a person with a strong desire for knowledge and meticulous thinking, his brain should have an instinct-to understand this thing that he doesn't understand. If you are in a vague state and don’t even know where the problem is, your brain will feel very uncomfortable and uncomfortable. This uncomfortable feeling will guide you to focus further and analyze clearly where the problem lies. .

Higher thinking precision means that your thinking efficiency and utility are higher, which further means that your learning efficiency is higher this time, and your error correction is really useful.
 

---

Thinking Like a Scientist = Metacognition + Pattern Recognition + Machine Learning

The study of science and engineering is not simply based on the length of time. The key is whether you think about it. If you think carefully, you will be able to master it quickly.

Because the way to make money in science and engineering is the same, for example, the core competitiveness of biology and computer science requires 10 years of research and development in a company with a large market.

R&D also has a process in essence. Although super-significant innovation does require talent, many innovations are actually process accumulation.

That is, if you have experience, understand the market, and have worked in a large company for ten years and become top-notch, you can almost come out and do it yourself.

Science and engineering make a lot of money → research and development of big markets and big companies → think like a scientist → discover the characteristics of the topic and thinking patterns → metacognition + pattern recognition + machine learning.
 
From birth, people are collecting, sorting, refining, and processing all kinds of data.

A lot of data is quite secretive, and it is impossible for people who have not been in a specific scene for a long time to collect a large amount of relevant data to form this kind of "pattern recognition/machine learning".

Pattern recognition - I have been input into a certain fixed model (learning other people's thinking patterns), and I make recognition, judgment, choice, etc. based on this model.

Machine learning - I faced a strange world without a clue, and slowly explored by myself, and gradually formed some so-called laws (the method of summarizing the characteristics in the topic and the corresponding characteristics), and this law is a temporary, The less persistent law, I want to use this law and model to collide with the world with a humble heart. If I'm right, deepen that knowledge; if I'm wrong, improve. Then, through this kind of continuous iteration and continuous investigation and knowledge, my cognitive ability will become stronger and stronger.

Pattern recognition is deductive, machine learning is inductive.

Deduction is to have a formed model judgment, and then use this judgment to standardize and measure the miscellaneous information, and then find some information, draw relevant conclusions and make certain predictions. This method is useful for identifying simple information. For example, it is easy to find triangles in a bunch of geometric figures, but it is often full of errors to find the only face in a bunch of faces.

The characteristics of machine learning, in terms of methodology, it is not top-down, but bottom-up. It absorbs a large amount of data without prejudice and prejudice. It generates some patterns from a large amount of unstructured and undirected data, and then looks for things that match this pattern in a large amount of data. At the same time, this process is not completed once, it is an iterative process.

The formation and iteration of this pattern is a slow, deliberate learning process, and once the pattern is formed, its recognition is instantaneous, unintentional, and without thinking.

Relevant to real life, whether Zeng Guofan knows people, or when a peacock sees a male peacock, it immediately judges whether the other is suitable for mating, or when Sherlock Holmes saw Watson for the first time, blurted out "Did you come from Back from Afghanistan”, these processes all contain two elements:

• The fact, the fact that you see before your eyes;
• The story, the story behind this fact.

A "story" is a fact in front of you that is formed at a specific time and scene after a process that may be very long, and recognition is to push back from the fact to the story, or "decode from the fact in front of you and restore it to a story"— — On the one hand, it is a story that restores the past, and on the other hand, it is a story that predicts the future — this is still a problem of "pattern recognition".

• Pattern recognition is a result, machine learning is a process;
• Pattern recognition is for facts, machine learning is for a story.

The role of pattern recognition and machine learning: solving comprehensive problems, finding ideas for solving problems, drawing inferences from one instance, and finding the most direct method.

Steps to use pattern recognition, machine learning:

① You need to do questions (300 questions in 2 months), and summarize the characteristics of each question.

When doing the questions, read in your head silently. The reason why you use the XX method must have the characteristics of XX.

To use pattern recognition, you need to master two points:

• Mastering method: display method and direct memorization. The implicit method needs to be discovered by oneself, and the key steps are compared and defined line by line.
• Identifying features: Features of a method, either from conditions, problems, or intermediate steps. Conditional features can be directly identified, question features need to be identified by comparing different questions, and intermediate step features need to be compared line by line in the problem-solving steps, and continuously compared to count a class of features.

Remember, at the beginning, start from the simple questions to summarize the characteristics of the method, and locate the difficult problems to disassemble each feature.

② Thought tracking is to find the key steps (tacit knowledge only mastered by experts)

Done right, what are the characteristics of using this method?

Wrong, how I thought and wrote it down. Then compare the correct answer which step (key step) am I stuck in?

Find a specific starting point : When correcting a wrong question, I only thought of learning the correct solution, but forgot to think about how I made the mistake (or didn't make it) in the first place?

Doing so has huge implications.

First, doing so allows you to examine your own thinking patterns and uncover deep problems lurking in them.

Learning the answer to a question is very superficial, but improving one's thinking mode is profound and fundamental, and it belongs to the growth of epiphany level. It allows you to develop by leaps and bounds.

Second, discover the loopholes in your knowledge system.

Why do you think of swapping instead of constructors for this question -> You are not familiar with constructors, nor are you familiar with swapping.

③ Mixed problem

The question has more than one feature, and the N features are the combined application of multiple knowledge points, so all the features of the question need to be disassembled.

Take features as indicators.

For example, a question = feature 1 + feature 2 + feature 3, I can’t answer this question because I only understand features 1 and 2, but not feature 3, so I can’t use the formula/problem-solving ideas I just learned, then I need to train features 3.

Step-by-step design questions, from few to many features, from single to combination.

Simple topic = feature 1

Medium Topic = Feature 1 + Feature 2

Complex topic = feature 1 + feature 2 + feature 3

④ Complicated problems, mainly deriving logic chains

There may be 10 steps in the logical chain of complex problems, but normal people can only deduce 3 steps.

• Sequential push: 1+9
• Push back: 1+8+1
• Pattern recognition: 1+3+1+4+1

Forward push + reverse push, and then look for features as intermediate stations to connect with each other, forward push + feature n + reverse push form.

This turns the 10-step logical chain into about 3 steps.

Then write down the logical chain (5why, 5so) and method characteristics of this topic.

⑤ Structural mode

The knowledge points are structured (explicit knowledge), just follow the catalog outline of the book.

The problem-solving thinking is structured (tacit knowledge), and the knowledge points are still not able to solve the problem. Only the problem-solving thinking can solve the problem, and pass the pattern recognition.

The internal structure of the topic (tacit knowledge), the change of the series of topics, through comparative analysis. The two questions belong to a large category, comparative analysis, which parts are the same and which parts are different. Make both questions, disassemble each condition, and compare each condition. The conversion conditions are the same, the same, different, different difference items, points of difference items. Which condition is the same, which is different, and which is transformed into exactly the same

⑥ Associated Learning

Association Direction:

• Imagine yourself as a question teacher, how to design new questions and new conditions
• Forward association (what is the connection with the knowledge I have learned before)
• Backward association (what can this knowledge be extended to)
• Yin Yang (what is the hidden message)
• context logic
• life experience
• Position: What problem does the author write this book to solve, or what is the original intention based on it?
• Thinking mode: In order to solve this problem, what kind of thinking does the author adopt?
• Advantages and disadvantages: copy the advantages of excellent peers, undercover peers to attract outstanding talents, improve the shortcomings of peers as a selling point to beat peers

⑦ Form a closed domain

The topics are endless, but the structured problem-solving ideas are limited. Mastering this chapter is equivalent to mastering all the problem-solving ideas in this closed domain.

 
The improvement of thinking comes from tiredness and discomfort, and the improvement of body comes from soreness.

Don't obey your instincts, don't avoid this kind of discomfort, you have to face up to the difficulties, focus on mastering this knowledge point, and overcome this feeling of tiredness and discomfort, because only in this way can you improve.

Ways of thinking in science and engineering - from a lot of invalid thinking to a lot of effective thinking:

■ Thinking Tracking - Discover thinking differences and write out the author's logic chain

■ Logic chain-discover the principle of things, write the author's why→why→why→so→so

■ Structural thinking - analyze the classification method and write out the possible ideas for each object

■ Pattern Recognition-quick problem solving, statistics of the corresponding relationship between topics and methods

■ Super-fine deliberate practice-growth, the more refined the training, the more obvious the effect

■ Thinking model of science and engineering - thinking to solve specific problems

If you want to grow muscle, the best way to train is to focus on "one part" of muscle mass every day.

How to measure whether today's "exercise" is reasonable? The easiest way is: will the part you practice be sore the next day?

If there is "soreness", it means that the intensity is reasonable.
If there is no "soreness," the intensity is too low.

When a "muscle group" is fully stimulated, let this part rest for 3-7 days before training it.

There's an iron law in fitness: To gain muscle, you must accept pain.

In fact, this is the case in any field. If you want to grow, you must endure a certain degree of pain.

When you repeatedly think about science and engineering problems according to a fixed line of thinking, you are training your brain, changing your brain neural structure, and allowing it to reconnect and strengthen in a more effective way.

These methods to retrain one's brain nerve connections must be a long-term process, requiring at least several months of hard training!

But this way of thinking is a high-intensity use of the brain. This state lasts for up to 45 minutes, and you need to increase exercise and sleep to recover.

Once you don't recover, you're emotionally drained and don't want to do anything.

After enduring those extremely painful moments, the brain's brain power will increase, and it will not feel tired and uncomfortable to do these thoughts in the future.
 

---

I found that the algorithm courses on the market either only talked about methods or only about topics, and did not highlight the connection between topics and methods.

This caused everyone to only think of learning the correct solution, but forgot to think about how they did it wrong in the first place, or why they didn't do it?

My algorithm class is ultra-fine thinking tracking, helping people to examine their own thinking patterns and discover profound problems hidden in their thinking patterns.

Learning the answer to a question is very superficial, but improving one's thinking mode is profound and fundamental, and it belongs to the growth of epiphany level, thinking like a scientist.

• From the microscopic level of thinking, what is the core pattern of this question, and where is it reflected in the question? Why is the answer to the problem-solving steps like this, and how does it correspond to the question?
• Mesoscopic pattern recognition, the reason why this method is used must be because there is a certain feature, usually an implicit feature, which requires refined thinking to come up with
• Structural analysis on a macro level, first positioning several possible problem-solving thinking frameworks each time, and then looking for recognition patterns
• The overall problem-solving ideas are deduced with logical chains, A->B->C
• Topic summary is the connection between features and methods

 

---

Three-dimensional logical framework: horizontal thinking is comprehensive and step-by-step, and vertical thinking digs deep into the underlying logic

Lateral thinking: Diverge the content and make the thinking structure more comprehensive.

Vertical thinking: Converge content and make thinking results more accurate.

How to think comprehensively? (lateral thinking)

When you get a problem/thing, how do you think about it comprehensively?

Lateral thinking, we can always think in three ways:

• Process Dimension: Emphasize steps - use it when making action guides, set the theme -> write framework -> collect materials -> refine framework -> write verbatim draft -> make courseware
• Dimensions of elements: Emphasis on composition - used when looking for reasons, dichotomous to do and not to do, system combination element dismantling, formula dismantling
• Time Dimension: Emphasis on change - share the development of individuals, teams, things, what they did in the past, what they are doing now, and what they will do in the future

How to think more deeply about a problem/thing?

Thinking vertically, we can always think in three ways:

• Make comparisons - do subtractions: what is more important, what is better, what is more urgent, what is lower
• Progressive - Digging depth: from cause to effect, from concept to application, from abstraction to concrete, from phenomenon to essence
• Logical thinking formula: 5why, 5so, learning models of other disciplines

Other thought models:

• Golden Circle - Advice: Why -> What -> How

• SCQA Model - Finding Help: Context, Barriers, Questions, Answers

 

The knowledge structure of science and engineering is divided into: knowledge point structure, problem-solving thinking structure, and associative learning.

If you only learn knowledge points (explicit knowledge in textbooks), you will not be able to solve problems, and you need to master the problem-solving ideas (tacit knowledge of experts) to solve problems.

Problem-solving ideas = refined thinking = thinking tracking + pattern recognition + comparative analysis + internal structure of the topic + positioning and dismantling.

Associative learning (innovative way of thinking) = perception + association + comparison + analogy + example + explanation + application + reasoning + adaptation + derivation.