What mean that AI understand?

Understanding is the answer to the question "why"? So that means that we should know how to make something. And then we should know why we should do things exactly some way. 

When we are turning our car to somewhere. We must follow certain rules. But then we must realize that those rules are made. That also other people who are outside our vehicle know what we are going to do. 

When we are researching AI we are facing the question. Does the computer understand things what it does? The thing is that understanding is something that even humans don't understand. When somebody asks "do you understand what material is"? 

Everybody can say. That material is consisting things like protons, neutrons, and quarks. Some person who has read a little bit more about the material. Can say that there are subatomic particles like quarks inside protons and neutrons. So that person is mainly right. Most of the mass of the atom is in protons and neutrons. But the contact layer of the atoms is in the electron core. 

The most out electron cores are in the main role in chemical reactions. But most of the atom is inside the electron cores. And that means the person who is saying that material is forming of protons and neutrons is same way wrong. The fact is that even 8 years child can tell about things like electron cores and quantum interactions if that person has some book about those things. The thing that is needed is the skill of reading. And then we have the person who knows everything. 

But then that answer of the electrons and protons can be wrong. The material can be written text that should publish on the net. So when we are asking about the material we should describe what that thing means today. The material can be protons or neutrons. But it can be something virtual like data that is collected from somewhere. 

Knowledge is:

How the actor does something? 

And understanding is:

Why the actor does something?

And if we are thinking about understanding it's simply putting data units one after one. The data unit is like a frame in the film. And they can connect to the endless film. That film can show some certain processes very accurately, but we know that projector doesn't know anything about things that it introduces. 

So if we have artificial intelligence that can read it can read every single article from the databases. We can use keywords and then the AI is searching an article. That connected to those keywords. Then the AI can read those articles to us. Of course, there can be some kind of parameters that the AI can use. Those parameters can be how many times a certain word is repeated in the texts. That thing helps to filter the searches. 

So what makes artificial intelligence maximize the number of clicks is filling your homepage with swear words and highlighting those words. It will bring in clicks. But perhaps the publicity of such a website would bring all sorts of sales to online stores

The thing is that when the AI creates some articles it can simply find texts. And then it can connect the parts of those texts. When the AI helps the writer to make more addictive texts. It can search things like what kind of words are the most clicked homepages include. It can use things like internet search indexes and find out what are the first words of the most clicked homepages. 

This is how the AI makes sure that is the homepage or text is interesting. That is written on the net will get the maximum number of clicks. And the clicks are the thing that is used indexing the interest of the text or homepages. For AI the maximum number of clicks is that people are interested in things. Those are published on the homepages. And we know that this is not true. The clicking of some homepage doesn't mean that somebody cares what there read. 

We can click on millions of homepages and spend about a second on those pages. So to make real analyzes of the information about interesting websites, artificial intelligence needs more information. It requires information on how many minutes or seconds each person will spend on this website. In the case of online shopping, the measurement of interest can be measured using the number of sales transactions as an index. When someone buys something. It is an expression of interest.

But when we are going to think about the understanding. We don't know does the AI knows anything. About the things that it does. The thing is that understanding is an interesting thing. It should be a simple thing to explain. But it's really hard to understand. 

The diversify data storages and their benefits

The diversify data storages and their benefits

When small units are forming the large entirety

Quantum computing means much more than the quantum processors. In data storage, quantum data means the bits, the smallest possible unit of data. If we want to store data safely, we must sacrifice comfortable. In this version, we can store our files that way, every single bit is in different hard disks. And if some of those hard disks will be damaged, the number of the lost data would be minimum.

Diversify solutions in computing are very effective and safe. The diversify computing means that every byte of data is stored in a different place, and if some allocation unit of the data storage is damaged, the rest of the data can be safe. And if we are thinking about the thing, what is called quantum computing, we might think that we are storing every single bit of data of the most important files in the own flash-memory around the world, that makes sure that in the case of damage the minimum number of bits is lost.

And if there is some kind of corruption in the files, the user of the computer can send the signal to the system, that it must start to create a new file by collecting those bits to one place. That thing could happen automatically if the program sends a signal to the data storage that there is an error in the code. In this case, the file contains two types of code. The file code, and the control code, what system uses to find out the problems in the file.

Artificial intelligence is an effective tool for securing the files. 

Every time, when the data is stored in the backup space,  every single part of that code is equipped with a tracking cookie. If the backup file is needed the artificial intelligence uses this code for tracking the code, what belongs to the file. And in this type of data storage, the system has multiple control mechanisms like mirror files and the storages, where is the length of the codes, and if some of those are changed, the algorithm starts to collect the new files from the secured sources.

If the size of the code has been changed, that might mean that there is a virus hiding in the file, and that means the code should be destroyed, and replace with a secured code. If those storages are as small as possible the firewall software would have a bigger chance to find out, if there is some kind of harmful code in those bits.

And then the system would start to collect the new file if it sees that the file, what is in the fast-use storage is somehow corrupted. If there is a problem the artificial intelligence can start to make the new file by using those control codes, which have the same mission with TCP/IP, and the storage system uses them to track the file code, which might be stored in multiple servers and hard disks.

So in this case, we must say that the human brains are also using the diversified computing solution. A large number of cells, which are handling quite a small number of data in each cell are making them very effective. The effectiveness means that every cell must not look for a very long time the data, what is stored in those cells, but there is one weakness in this way to handle data. And that is that human brains are a little bit slower than they could.

When human brains decide to work the signal goes across many layers of neurons, and the mission of some of those cells is controlling the signals. They are filtering the actions, which might cause unnecessary harm or danger for the environment.

Artificial intelligence as an architect

Artificial intelligence as an architect

In the world of tomorrow, the artificial intelligence is maybe the independent and ultimate tool for research and development, which have traditionally believed to belong to humans. The thing is that artificial intelligence can work as an architect, and furnish rooms and houses independently.

The thing what this algorithm need is to access the social media of customer, and where it can see the photographs, what the user has put in there. The thing that artificial intelligence would use are the words, what the user has used about the images, and if it has access to find out things like what kind of images that customer normally looks.

The artificial intelligence follows the movement of eyes and changes in electric actions in the pleasure center. By using that data the system creates forms that please the user. After that, the artificial intelligence makes the matrix or average image. It calculates the color palette, what it uses for the selection of colors for that project.

After that, the algorithm would calculate the angles and average forms, what the person looks at, and it can also detect the movements of the eye if it has access to that. And in the perfect version, it can detect the feelings of humans by observing the neural electricity. In that case, the computer must just show images, and EEG would tell if the image pleases the customer.

Then the artificial intelligence would just show images about the result, and then by observing the feelings, it can make perfect room, which pleases the customer. During the process, it adjusts the colors and the forms and places of things, and it can make the selection by using the budget, what the customer has given. The adjustment would make by following the reactions of the pleasure center of the brains during the process.

The difference between artificial intelligence and the human mind is that artificial intelligence would not have real feelings. Those highly sophisticated computer programs can play feelings, but they don't have them. So artificial intelligence can focus only on the mission, what it has. And the thing is that artificial intelligence can handle very much data if we are comparing that with the human mind.

Here I must say that artificial intelligence has one benefit to humans, and that is the system cannot use parameters, what it should not use. So the algorithm can collect letters or as an example the work applications by using indexes like texts, which seem to be made by the same person. In this time the system searches the similar details in the applications, but it doesn't notice the sender itself. That kind of thing makes possible to detect if some person makes the applications for other people.

Artificial intelligence can compare things like a way to write the messages by comparing the rate of the letters in the messages, and the way to select words are uncovering the writer. And other similarities in the writing style can tell if the writer of some messages is the same. The thing, that makes computers so good in this thing is that they are comparing two images by bit to bit, which makes them ultimately sharp.

What means that artificial intelligence can see things, what human beings cannot notice, because there are so many objects in the images, what are portraying certain parts of the sky or some rain forest. The computer can detect the smallest differences in things like the size of the leaf of the trees and change of the color of the green parts of vegetables.

The computer just compares two images together, and then the system can tell is the area of green material grown, is it decreased, or is a leaf of those trees going pale? Which means that the level of chlorophyll is decreased. That kind of data is an interesting thing because it tells about pollution.