In the development process of artificial intelligence in China, there is a person who cannot but mention, this person is Li Guanfu, who once served as the global vice president of Microsoft and Google, and founded a startup company called Innovation Factory after returning to China, focusing on supporting young people to start a business, known as the entrepreneurial mentor of young people, every day in the Weibo circle of friends public account in the public account of various life insights, chicken soup for the soul, after the chicken soup for the soul became poisonous chicken soup, began to move to the field of artificial intelligence.

All kinds of views on artificial intelligence will appear in his public account, and there is also a book on artificial intelligence, which has had a very great influence in the process of promoting the development of artificial intelligence.

In the past period of time, I often say that artificial intelligence is the future of scientific and technological development, will change our lives, many people are very worried about this, thinking that once artificial intelligence is developed, it will replace human beings, so that human beings have no work to do, for these worries I want to say is: the biggest application of artificial intelligence is to improve social productivity, can create more wealth in a limited time, compared to human artificial intelligence is more game, they can work 24 hours a day, no fatigue, no need to rest, and the efficiency is dozens or even hundreds of times that of human beings, the application of artificial intelligence in various fields is becoming more and more popularBut the current artificial intelligence can only be called weak intelligence, they are just repeating mechanically, without their own innovation, all the decisions they make are designed, no matter in the powerful intelligent program is designed.

For artificial intelligence, the final development is to evolve an intelligent program with human thinking, this program can have its own thoughts, can modify its own code, and has creativity, except for the absence of human bodies, artificial intelligence will be no different from human beings, I thought that I would not see real artificial intelligence in my lifetime, but I recently found an app called Xenomorph on which I saw the direction of artificial intelligence development.

Xenomorph this app has not been launched for a while.,It's only been a little over a month.,Now users are scolding more than praising Xenodegree.,I also used Xenodegree personally.,And then I feel that Xenodegree will evolve into a real artificial intelligence.。

On Xeno's developer platform, I saw the overall architecture of Xeno, which seems to be common, but one thing that shocked me was that it was Xeno's auto-executing module.

Xenodegree has a total of three modules, the first module is the information search module, the second is called the information analysis module, the third is called the advance execution module, the first two modules are very common, the search engine we are using now is an information search module, the second information analysis module is not unusual, it is applied in many fields, the most common is information recommendation, for example, today's headlines are more popular today is to recommend the news you are interested in according to the user's browsing information, and the treasure thief network has long realized the function of thousands of people, this personalized recommendation is the specific application of the information analysis module, then the early execution of the third module is an idea that has never come up before.

And the early implementation of this module will give birth to a real artificial intelligence, why is this so, because this module complements the most important link in the development of artificial intelligence, that is, the feedback of results.

What is the result feedback? We humans will get a feedback after making every decision, for example, we will feel full after eating, we will feel the heat when we put our hands near the fire, and you can feel the pain when others hit you, and it is this kind of result feedback that allows us to know what we can do and what we can't do, just like a baby, we will cry loudly when we are hungry, and someone will bring food to our mouths after crying every time, and crying in the baby's consciousness represents food.

The generation of human consciousness is established in the result feedback again and again, and artificial intelligence also needs continuous feedback to evolve thinking, in the past, we have continuously improved the algorithm of artificial intelligence in the first two modules, we have collected a lot of information, and then deduce the results through the analysis of the information, and then recommend the results to the user, we give the right to choose to the human, not the program itself, we provide a large number of results to the human, and then let the human being choose which one they like, this approach is not wrong, but he makes the artificial intelligence unable to get the correct feedback resultsBecause human beings are disturbed by many factors when choosing information, different people will have different choices for the same pile of information, and it is multiple choices, which makes the program not know what you really like, and there are many contradictions in these choices.

Xenodegree's automatic execution leaves the right to choose to the program, only one of the results of countless analyses is selected to execute, and then verify whether our decision is correct by testing our actual reaction to this matter, and now people who use Xenodegree will find that the longer Xenodegree is used, the higher his accuracy, or the more powerful his elastic mechanism, a program can automatically order takeout, automatic taxi is not unusual, just find a programmer to achieve the function of timed execution, the key is that there will be a slight change in the time of each order, not to say that every day at 12 o'clock punctual order takeout, but decide whether it's time to order based on your behavior.

There is a heterogeneous architecture on the developer platform of Xenodegree, this architecture is not uncommon is the traditional cloud plus local end, every app installed on Xenodu is a local end, this architecture is also often called the child-parent architecture, the cloud is responsible for storing information for the child end to call, then what surprises me the most in Xenodegree is that all the analysis, calculation is performed on the child end, the cloud only provides information query, and does not participate in the calculation of the user's child end, which means that each child end is an independent intelligent program.

In the past, we focused on feedback on this thing, that is, put thousands of user information together, and then find the commonality of this group of people, we put all kinds of labels on human beings, age, income, hobbies, whether they are single, etc., but we ignore that people are different, just like there are no two leaves in the world that are exactly the same, and people are not exactly the same two people, and the results of this information analysis must be popular, not personalized.

The real personalization is the way of heterogeneity, the user's information collection, behavior prediction, early execution, result feedback, through the feedback of the results to adjust their own algorithms, these are independent, continuous results feedback back for each user to create a completely different intelligent program, a personal intelligent program, this is like a copy of the game, although they are all using heterogeneous, but each person is independent, each person's algorithm will be constantly adjusted, this kind of application in the world is manifested in each person has a tailor-made intelligent program for him, or it can be said that another self。

Xenodegree in the local for each user to build a behavior tree, we can take the behavior tree as your three views, that is, your world view, values, outlook on life, we often say that character determines fate, the so-called character is your three views, he determines your reaction to a thing, as if a person beat you up, at this time some people will hit back on the spot, some people will endure, some people will retaliate afterwards, different people have different reactions, the behavior tree established by Xenodegree is your three views, he is actually using your three views to make decisions when he helps you make decisions。

The use of Xenodegree is actually to build another self, a person who knows himself better than himself, we are a person who coexists with emotion and reason, many times we will dare to be emotional, but the behavior tree established by Xenodegree will not dare to be emotional, everything he does is based on the idea of making our lives better, we can see him as a rational self.

The real artificial intelligence is actually a rational self, he will not slack off, there is no inertia, he can help us overcome ourselves, let us become better, we don't have to worry about artificial self will rob us of our work, because every artificial self is actually our own.