Future Super Intelligent System

Everyone who had started to get tired didn't consciously raise their spirits. Normally, two people came to this proposal to participate in this proposal, a project leader and a technician.

So whether it is a project manager who travels the market all year round or a technician wearing a plaid shirt, naturally he is full of interest in what Liu Fan will say next.

"The reason why our face recognition technology can achieve the scene application you just saw is based on our theory of dynamic feature logic."

Liu Fan started to explain. Some people started recording, some people started recording, some people started taking notes.

"The core of our current mainstream face recognition technology is features, whether it is overall geometric features or local features, whether traditional algorithms or neural network simulations, the root of everything is to capture features and process them. But regardless of the follow-up How dynamic or how high-dimensional the processing method is, the features we capture are static in nature.

It is impossible to avoid being easily affected by the environment. So we have long considered whether we can capture dynamic features instead of static features. Later, we realized that simply capturing dynamic features is still not enough. If you can use big data to mine dynamic feature logic, then no matter how the environment affects, as long as you can build a sufficiently large dynamic feature logic model, even if you rely on very fuzzy Image data may also enable face recognition.

I give a simple example. "

Liu Fan said, point the PPT to the next page. "First of all, everyone here must know a basic principle, that is, one leaf and one bodhi. There are no two leaves in the world that are exactly the same. Growth process.

You can look at PPT. This is the dynamic feature logic we discovered for the first time. After continuously digging these 30 million dynamic faces, the computer got unexpected answers. When people turn their eyes, their eyes There is a functional relationship between the rotation and the muscle changes that occur at the position of this red dot below the cheek. This function is the same between different people, the only difference is the coefficient.

The specific function involves commercial privacy, so I will not show it here. I only make a simple analogy. The relationship between A's eyeball and cheek function is F (X) = y, and the function relationship of B is F (X) = 1.1y. The functional relationship is F (X) = 1.2y, and so on, so when the video captures the variable X of a person ’s eyeball rotation, the corresponding function result is unique, and a dynamic function like this is currently on the human face 26 were found, and we are still working hard.

I believe everyone can understand that the advantages of dynamic functions are very obvious. For example, a criminal covers his face, but as long as he can see his eyes, as long as his eyes provide a variable X, then he can perform several eye contact The calculation of related functions, maybe he also covers his eyes, but the muscle changes in a certain part of the face also have corresponding corresponding functions, so ... "

"You let the algorithm derive the algorithm!" Liu Fan said half of it, and suddenly someone shot him and interrupted him.

When the man in the black plaid shirt said this, he immediately became restless on the court.

They just patronized and listened to Liu Fan saying that they hadn't had time to realize what Liu Fan's algorithm would mean to the artificial intelligence industry if it really existed.

The so-called deep learning algorithm is the study of ability. For a slightly one-sided example, let the computer continuously learn multiplication, and the computer's multiplication calculation speed will become faster and faster.

To use a more life example, many companies now start to develop robots. When you see robots interacting with people, you will feel that the era of artificial intelligence is really coming. But in fact, in the process of robots talking to people, we simplified the process of deep learning. In fact, what should be said is an optimized answer under certain program settings.

For example, the girl told the robot that I was angry. The current thinking model of the intelligent robot is like this. She said that she was angry. According to the principle that women are unreasonable when they are angry, the optimal solution is obtained at this time. , Just apologize, and send her a red envelope.

But how do real people think about it? Is she really angry? Why is she angry? What should I do in such a situation? Apologize if it's not a big deal, but if it's a matter of principle, you can't get used to it, because it may ruin her and yourself.

This is just an example. There are many more ideas for boys in real interaction.

In fact, in this comparison, everyone can easily understand the difference between humans and artificial intelligence. One is to follow the set program to analyze and obtain the optimal solution, and the other is to perform different real-time reactions based on life experience and own emotions. . This kind of immediate response is theoretically completely disordered and unpredictable.

We can find that when people deal with one thing, they go through several processes: perception, analysis, and decision-making.

现在 What the current artificial intelligence does is information input, information analysis, and output of optimal results.

So in terms of information processing, artificial intelligence looks like human beings, but once emotions and disorderly things are involved, the difference between artificial intelligence and humans comes out.

I am going to die for my loved one, and now artificial intelligence will never.

At this time, look back at the algorithm implemented by Liu Fan. If the algorithm can derive various functions based on the continuously input data, the behavior pattern at this time is to some extent simulated: people start from completely ignorant babies. As we grow, we learn more and more about the world.

In fact, for artificial intelligence, the so-called algorithms, functions, and rules may be human survival rules. Isn't human life a process of constantly understanding life.

If you refine the process of human growth, is it possible that you are constantly discovering countless rules? It's just that the brain has dealt with these problems, and we haven't noticed it ourselves.

So when the algorithm can discover the rules by itself, it is equivalent to opening another door for artificial intelligence. Although this direction may not be completely correct, in theory, it is likely to be closer to the door of true artificial intelligence.

But this thing is simple to talk about, it is too difficult to realize, just like Zhang Kaixiang has a strong mathematical background ~ www.NovelMTL.com ~ also has a corresponding conjecture, but can not start. If Liu Fan doesn't have the abnormal plug-in of the system, he may not be able to do this in his life if he wants the algorithm to autonomously mine functions.

Of course, in fact, Liu Fan's current algorithm is far from the ideal state. On the one hand, there is still a lot of space for the accuracy and mining ability of function derivation, because it will be limited by the amount of data and data types. . For example, before this face recognition, Liu Fan's data association inverse algorithm had not made any breakthrough progress.

Although in the past, it is indeed possible to realize the use of unordered data as he had previously conceived, but the previous functions did not break through the traditional mathematical framework of the underlying algorithm. This leaves the question of whether these functions are derived by the algorithm itself. Discussed.

But this experiment on face recognition has made Liu Fan see a breakthrough, and the self-derivation of the algorithm seems really feasible.

Another problem is that Liu Fan's algorithm can only analyze and cannot make decisions.

He tried many methods, but he couldn't realize his decision-making ability.

慢慢 He slowly realized that in order for the algorithm to have decision-making ability, he must break the existing underlying algorithm principles.

But this does not affect the shock caused by Liu Fan's current algorithm to the technical staff present today. Liu Fan and Zhang Kaixiang will have such conjectures, and others may naturally have such conjectures, but no one can overcome the technical difficulties Nothing, so when everyone reacts to what Liu Fan presents if it really means what it means, everyone will have a hard time controlling their emotions.

Everyone's shock was what Liu Fan was expecting. From the moment when the GSCT was positive, Mulong Science and Technology was an army standing on the battlefield and preparing to attack the city. It was no longer necessary to converge ...