Crazy input!

Today I originally went to the company’s new location to check out the environment, and the journey was extremely long. However, halfway through, the person in charge of the company simply said there was no need to go. So I immediately took the subway back and ran to have lunch.

So the next path was all traceable. After lunch, I ran to the cafe, and sat there drinking cup after cup of coffee all afternoon. Then during my lunch break, I spent the entire afternoon listening to the one-hour recording and retraining of the 8 lesson.

The main content also needs to be reviewed again to achieve a better review effect:

When talking about the memory section, the analogy used is actually very familiar and vivid. Computer ≈ human brain, computer memory ≈ short-term memory of human brain, computer hard drive ≈ long-term memory of human brain. The human brain is essentially a Bayesian inference machine based on memory. The core of this memory is the clear, accurate and necessary association structure between clear, accurate and necessary concepts (models) one after another. In order to better calculate and generate more meaningful content associations, the best way is to improve memory. Remember, it must be memorized by rote, rather than memorizing things by establishing additional memory associations. The advantage of this is that the more you memorize, the more powerful your brain will be, and the more powerful you will be, the more content you can memorize. Then the performance of your memory-based Bayesian inference engine will be better. Because in order for your Bayesian inference engine to calculate, it needs to continuously make further inferences based on existing evidence, and this existing evidence must come from your daily memory content. Therefore, any meaningful speculation you can make can only be based on content you are already familiar with. Therefore, if you don’t have the concept of a certain field in your mind to do something that needs to be done, it will be really difficult to do it. At the same time, in order to do well, you have to force yourself to learn. In turn, after further study, your creations will become more outstanding. Ultimately, a virtuous cycle of learning is formed.

Here I have a personal experience: I didn’t quite understand this analogy, or it was relatively abstract to me. It wasn’t until I thought about studying some technology giants in the field of AI that I had a very strong desire to learn. And because I relied on output as the driving force to force myself to learn. Therefore, in order to write professionally enough, I was forced to study the underlying logic of AI operation several times more seriously, and by the way, I sorted out and understood the concepts of hard disk, memory, GPU and other concepts. So much so that when I heard that analogy, it seemed so familiar to me…

Teacher Xiaolai’s methodology of “promoting learning through teaching” has long been fresh in my mind. So every time I planned to do something, I suddenly became more and more familiar and even experienced. There is no longer a phenomenon of being dissuaded by those so-called “high-level terms”. If you want to install openclaw, you can install it without asking anyone. Anyway, I saw someone in the circle of friends deploying it successfully, and I thought it was OK, so I spent some time and energy to get it done. For another example, if you want to study the leader in the AI ​​field, we have the same logic. Those who can Google will never ask for help. So after thinking about it for a while, I became more and more comfortable with the technology stocks at hand. I have been using AI continuously for the past three years, and I have paid for it and used it to do real things before I can realize that the bubble of AI is really redundant. The essence of the one-person company that the teacher talked about in class was also instantly understood. A department is just a folder, what an agent manages is just a bunch of markdown files, and a management structure is just a tree structure of files/folders and other analogies can only be understood after using it personally…

So in order to better learn these methodologies, I had to start typing a lot, because this is the key to improving my understanding speed! There are a lot of charlatans on the market, but they can deceive me temporarily, and after a while, I will immediately see through them and leave them completely naked. As a result, good habits have been developed even more diligently. For example, apps that use a large number of paid AI to replace entertainment are really extremely trustworthy. Many bad habits were quickly replaced by my need for productivity. At the same time, my note-taking system also began to be set up desperately. The association between modules constitutes the structure of knowledge. Therefore, it has become my endless interest to retell and organize expressions by changing patterns and scenes.

Let’s sum it up!

• Learning is the real production -Production is real learning
• At any period, mainstream ideas are basically not the result of independent thinking