- A purely networked company should have the following features: distributed, decentralized, collaborative, and adaptive.
- A network is an information processing factory. The value of each product increases with the increase of knowledge contained in it.
- In a complex adaptive system, you cannot rely on interpolation functions to determine system behavior.
- The modularization of programming is the "reliability" of the program ".
- When you find an error, it means that another bunch of errors you haven't seen are waiting for you.
- There are various reasons for creation, but what we create is always a world. In fact, there are not many creative actions, and they are exactly the re-occurrence of creations.
- Absolute control is absolutely boring. To create a new, unexpected, and truly different thing-that is, something that truly surprises you-you must give up your dominance of the throne and give up the bottom-Layer groups.
- Evolution will be the next stage of contemporary art development. This art can be inspired by the concept of variation and sexual reproduction.
- When the exploration space is large enough, the effective exploration process is no different from the true creation.
- The number of books will continue to increase. It is foreseeable that learning from books will be as difficult as directly studying the universe at some point in the future; searching for a truth hiding in nature is not as troublesome as searching for it in a complex book.
- It is not difficult to find the path to a high point. The evolution in nature and the evolutionary procedures in computer science are good at crawling to the highest point in the global sense in the ups and downs of a mountain, even higher than a mountain.
- Repeated blind behaviors over and over again can only lead to a deeper level of ridiculous behavior, while a group of individuals execute blind behaviors in parallel. when conditions are appropriate, they can export all what we think is interesting.
- The more knowledge you enter for a computer, the slower it will run. For a person, the more knowledge you give him, the more agile his mind. Therefore, we are in a paradox. The more you want to make your computer smarter, the more stupid you are.
- We know that there are only two ways to create structures and their complicated things. One is relying on engineering, and the other is through evolution. In both cases, evolution can make more complicated things.
- As long as the engineering methods cannot be discussed, they can all be imported into evolution.
- Another method of supervised learning is directed to history, which facilitates Borges library, and another method of breeding. In the targeted evolution, selection is guided by the nurturing, rather than natural.
- It is the best for individuals, but not necessarily for species.
- Apart from being able to create things we cannot make, evolution is also worth boasting: it can make things with fewer defects.
- Reliability and no-fault running time will become the top priority of the software.
- Evolution is not beyond our control; giving up some control is just to make better use of it. What we are proud of in engineering-precision, predictability, accuracy, and correctness-will all fade away from evolution.
- The future of control is partnership, collaborative control, and man-machine hybrid control. All this means that the creator must share control with his creations and share the same fate with others.
- Computers and evolution are thinking in parallel; at the moment, they are thinking in sequence.
- Tom Ray: I want to upload my computer to my life.
- Whoever has a flexible external form can get a return -- this is the essence of evolution.
- The evolutionary work is to create all possible possibilities through the space where all possibilities reside.
- Once you have the right questions, you may have some answers.
- Once a weak innovation is received, it can serve as a platform to generate more competitive innovation. In this way, the fire of stars can be restored.
- Minsky: the mind cannot grow well only by accumulating more and more new knowledge. It must also develop a better way to use existing knowledge. This is the school principles: some of the most critical steps in mental development are not only based on acquiring new skills, it is also based on new application methods that use known knowledge.
- John Holland: What a complex adaptive system does is prediction.
- The failure of artificial intelligence lies in the fact that it has developed utility, but has sacrificed the control theory.
- Real-time Control of a complete general-purpose computer is a task that traditional control theory needs but has never done before.
- The real abnormal things that ultimately subvert the model are not even seen as abnormal things at first. They are ignored.
- At first, abnormal things were just the fact that people had observed, and there was no need to explain them at all. These facts are not troublesome facts, but facts. Abnormal things are not the cause of replacement, but the result of replacement.
- This may be a better example of scientific progress.
Books and notes out of control
