What is artificial Intelligence

 Artificial intelligence (AI) is one of the branches of Computer Science, with strong roots in other areas such as logic and cognitive science. As we will see below, there are many definitions of what artificial intelligence is. However, all of them agree on the need to validate the work through programs.

Artificial intelligence was born in a meeting held in the summer of 1956 in Dartmouth (United States) in which those who later have been the principal investigators of the area participated. In preparation for the meeting, J. McCarthy, M. Minsky, N. Rochester and CE Shannon wrote a proposal in which the term "artificial intelligence" appears for the first time. It appears that this name was given at the urging of J. McCarthy.

The proposal quoted above from the meeting organized by J. McCarthy and his colleagues includes what can be considered as the first definition of artificial intelligence. The document defines the problem of artificial intelligence as that of building a machine that behaves in such a way that if the same behavior were carried out by a human being, this would be called intelligent.


There are, however, other definitions that are not based on human behavior. It's the next four.

https://5fb9e409f32b1.site123.me/

https://5fb9e409f32b1.site123.me/blog/free-cars-for-single-moms

https://5fb9e409f32b1.site123.me/blog/free-cars-given-away

https://5fb9e409f32b1.site123.me/blog/free-government-internet-and-laptop

http://hardshipgrants.simpsite.nl/

http://hardshipgrants.simpsite.nl/cars-for-poor-people/19476-Free_cars_for_low_income_families

http://hardshipgrants.simpsite.nl/cars-for-poor-people/19481-Free_government_laptops_application_form

https://hardship-grants.yolasite.com/

https://hardship-grants.yolasite.com/cars-for-veterans/

http://amusecandy.com/blogs/post/595761

http://amusecandy.com/blogs/post/595762

http://amusecandy.com/blogs/post/595797

http://facebookhitlist.com/profiles/blogs/grants-for-single-mothers

1. Act like people.
 This is McCarthy's definition, where the model to follow for the evaluation of programs corresponds to human behavior. The so-called Turing Test (1950) also uses this point of view. The Eliza system, a conversational bot (software program) is an example of this.

2. Reason like people. What is important is how the reasoning is carried out and not the result of this reasoning. The proposal here is to develop systems that reason in the same way as people. Cognitive science uses this point of view.

3. Reason rationally.In this case, the definition also focuses on reasoning, but here it is based on the premise that there is a rational way of reasoning. Logic allows the formalization of reasoning and is used for this purpose. 4. Act rationally. Again the objective is the results, but now evaluated objectively. For example, the objective of a program in a game like chess will be to win. To achieve this objective, the way to calculate the result is irrelevant. In addition to the definitions mentioned above, there is still another classification of artificial intelligence according to what the ultimate goals of research in this field are. They are the strong and the weak artificial intelligence.

Weak artificial intelligence
It is considered that computers can only pretend that they reason, and can only act intelligently. Supporters of weak artificial intelligence consider that it will never be possible to build conscious computers, and that a program is a simulation of a cognitive process but not a cognitive process itself.

Strong artificial intelligence
In this case, it is considered that a computer can have a mind and mental states, and that, therefore, one day it will be possible to build one with all the capabilities of the human mind. This computer will be able to reason, imagine, etc.


Topics in artificial intelligence
Although there are different points of view on what artificial intelligence is, there is an important agreement on what are the results attributable to this branch of Informatics, as well as to the classification of the methods and techniques developed. We review below the four great topics of artificial intelligence.

1. Troubleshooting and searching. 
Artificial intelligence aims to solve problems of a very different nature. In order to meet this objective, given a problem it is necessary to formalize it in order to solve it. This topic focuses on how to formalize it and the ways of resolution.

2. Representation of knowledge and knowledge-based systems.Artificial intelligence programs often need to incorporate knowledge from the application domain (for example, in medicine) in order to solve problems. This topic focuses on these aspects.

3. Machine learning. 
The performance of a program can be increased if the program learns from the activity performed and from its own mistakes. Methods have been developed for this purpose. There are also tools that allow you to extract knowledge from databases.

4. Distributed artificial intelligence. During its early years, artificial intelligence was monolithic. Now, with multiprocessor computers and the Internet, there is interest in distributed solutions. These range from parallel versions of existing methods to new problems related to autonomous agents (software programs with autonomy to make decisions and interact with others).

In addition to the four topics mentioned above, there are others that are strongly related to artificial intelligence. They are listed below:

a) Natural language.
b) Artificial vision.
c) Robotics.
d) Speech recognition.


As you will see below, many of the most striking applications use some of the techniques related to these topics.

Some applications
To date, many applications have been developed that use some of the methods or algorithms designed in the area of ​​artificial intelligence. In this section we review some of the most attractive existing applications or those that have had historical relevance. However, these are not the only existing applications, as there are methods developed in this branch of Computing that are used in everyday devices or in software used by companies and corporations. For example, on the one hand, we find the search algorithms mentioned above in systems that build schedules taking into account the restrictions of the entities and individuals that participate in them. On the other hand, The learning methods are used to recommend products in the virtual stores and to select the advertisements that are provided to us when visiting certain web pages. Another example is that of fuzzy systems, one of the knowledge representation methods that have been successfully applied to very diverse control problems. There are both digital cameras and washing machines that incorporate a diffuse system inside.

https://topsitenet.com/article/588880-cash-assistance-for-single-moms/

https://topsitenet.com/article/589315-free-laptops-for-students/

https://topsitenet.com/article/589975-personal-grants/

https://www.evernote.com/shard/s441/sh/26d8b002-f8f4-f012-5fa9-379f4dcb7dd3/266a17535801cc03a007cb4d1bacef18

https://www.evernote.com/shard/s441/sh/3fd12c50-b237-5477-fe5b-9fb1129f8484/a0cab7456d0e6420800dbc327689f11d

https://www.evernote.com/shard/s441/sh/0642e379-1dd0-ecaa-bcd4-6e042f3d7840/a5a942a0e589a666d016e6e03a7872ac

http://vedadate.com/member/blog_post_view.php?postId=86458

http://vedadate.com/member/blog_post_view.php?postId=86491

http://vedadate.com/member/blog_post_view.php?postId=86549

https://utah.instructure.com/eportfolios/44124

https://utah.instructure.com/eportfolios/44124/Home/Help_with_paying_bills

https://utah.instructure.com/eportfolios/44124/Home/Cars_for_poor_people

https://utah.instructure.com/eportfolios/44124/Home/Free_cars_given_away

We will now see some of the most relevant applications.

Applications in games
For decades, artificial intelligence has been dedicated to games in order to defeat the best human players. It has always been considered that the ability to play was proper to intelligence. The ladies and the othello were defeated first. In 1997 it was the turn of chess. Now practically only the go remains to be beaten.

  • The ladies.Chinnok, a program developed since 1989 by Jonathan Schaeffer's team from the University of Alberta, became the women's world champion in 1994. The program includes a database with openings of the best players and another of final situations when there are 8 or less chips on the board. The same research team showed years later, in 2007, that when checkers are played perfectly, no player can win. This means that an optimal strategy on the part of the two players can only lead to a draw. The difficulty of this proof is that there may be approximately 500 3 1020 possible boards, or, in words, fifty thousand trillion boards. The problem is a million times more difficult than 4 in a row.
     
  • The chess. Inventions and programs were developed for many years in order to win at the game of chess. However, it was not until May 1997 that Deep Blue defeated human champion G. Kasparov in New York. The program developed by IBM used specific hardware, databases that allowed the program to play perfectly in the final situations with 7 or fewer tiles on the board, and search algorithms of the minimax type to find the best solution in all other cases. . Information about Deep Blue can be found on the website http://www.research.ibm.com/deepblue.
     
  • The go. While the other games have already been defeated, there is currently no computer program that is high enough in go to beat good human players. Go has been considered a much more difficult game than chess for years. The difficulty lies in the dimensions of the board (19 3 19, with 361 intersections), the number of possible moves on each board and the difficulty of defining functions that correctly evaluate a given board. Currently, some programs have been achieved that have a good level on a reduced 9 x 9 board. The programs that have a good performance do not use the same search algorithm as chess (minimumx) but rather UCT.


Applications in robotics
Applications in robotics have been developed from the beginning of computing with a variety of objectives: the automation of industrial processes, military applications and space exploration. While the first robots were oriented to perform repetitive activities, currently a greater autonomy is sought in relation to their ability to make decisions. The evolution of robotics has also gone through its attempt to build human-shaped robots with the ability to walk. Some of the most important achievements in this area are listed below.


Applications in smart vehicles
Many types of vehicles have been built with different degrees of autonomy. Some robots were already mentioned in the previous section. Vehicles that can carry passengers are indicated here.

  • The metro of the Japanese city of Sendai (仙台 地下 鉄). This meter, developed in 1987, was the first in the world with a fully automated driving. Currently, there are many driverless metro lines in the world. The system was made under the direction of Seiji Yasunobu, a member of the Hitachi Systems Development lab. It is based on fuzzy logic techniques.
     
  • Autonomous cars. Stanley was the winning car of the "2005 DARPA Grand Challenge" race. The autonomous and driverless car completed the 212.4 km journey in the Mojave Desert, in the United States, in 6 hours and 54 minutes. In 2007, the "2007 DARPA Grand Challenge" was held, which consisted of traveling 96 km in urban areas (at the George Air Force base in California). The cars circulated processing in real time the traffic rules of the state of California. On 20 September 2011, the Made in Germany car of the Free University of Berlin drove through the streets of this city on a journey of 80 km. The tour was between the International Congress Center and the Brandenburg Gate in central Berlin. The car is completely autonomous, although some information, such as the speed of the journey, they are given to you and are not picked up by your cameras. The car recognizes the presence of pedestrians and traffic lights. For information on the German car you can consult the following website
     
  • Unmanned Aerial Vehicle (UAV). The Global Hwak unmanned aerial vehicle was the first to cross the Pacific Ocean without stops. It traveled from the United States (California) to Australia in April 2001. However, it still needs a pilot at a ground station and other operators to analyze the data. According to Weiss (2011), the main problem with these autonomous systems is that, although they can collect a lot of data, they still lack the processing capacity to process the data in real time and act intelligently on this data.




Conclusions
Computing has come a long way since its inception 70 years ago. The computing power has been doubling every 18 months, following Moore's law. It is believed that, if Moore's law continues to be fulfilled, by the year 2030 the computing capacity of a processor will correspond to that of a person.

https://icube.instructure.com/eportfolios/2354/Home/Government_housing_grants

https://icube.instructure.com/eportfolios/2354/Home/Ford_police_vehicle_grants

https://icube.instructure.com/eportfolios/2354/Home/When_to_divorce_your_husband

https://icube.instructure.com/eportfolios/2354/Home/Government_senior_living

https://aalrc.instructure.com/eportfolios/1377/Home/Apply_for_housing_grants

https://aalrc.instructure.com/eportfolios/1377/Home/How_to_divorce_your_husband

https://aalrc.instructure.com/eportfolios/1377/Home/Police_car_grants

https://aalrc.instructure.com/eportfolios/1377/Home/Social_security_apartments

https://republicschools.instructure.com/eportfolios/3114/Home/Household_grants

https://republicschools.instructure.com/eportfolios/3114/Home/Social_security_housing_for_seniors

https://republicschools.instructure.com/eportfolios/3114/Home/How_to_tell_my_husband_i_want_a_divorce

https://republicschools.instructure.com/eportfolios/3114/Home/Ford_police_vehicle_grants

https://www.articleted.com/article/340609/77690/Government-hardship-grants

In turn, the amount of information stored digitally today is enormous. Search engines like Google store millions of copies of existing web pages, and companies' mail services accumulate our messages by the millions. Social networks record what our interests and our friendships are. Companies keep any information, however insignificant it may be, in case it may be of any use to them in the future.

Naturally, an increase in computing speed and greater storage capacity will mean that systems have more resources to make decisions and that these decisions are made in a more informed way and, in turn, in a more personalized way. 

Comments