From the robots of Asimov to Terminator and the androids of the "Westworld" humanity has had dreams or nightmares about artificial beings capable of reasoning but ... can machines think?
Jewish legends spoke of the Golem, a being made of clay that became conscious by inscribing a letter on his forehead. In the eighteenth century a certain Wolfgang Von Tempelen built an automaton capable of competing with the best chess players, later it was discovered that there was actually a person hidden inside the apparatus, but it was not until the 50s of the 20th century when computer technology allowed to glimpse the possibility of intelligent machines. The scientist and mathematician Alan Turing designed a test, if a human being that dialogues with a computer and with another human being can not distinguish between one and another one can affirm that the computer is intelligent.
As early as 1956 there were already researchers designing computer programs that could play checkers and win, solve algebra problems and do mathematical demonstrations. The computers used the method called "Logical" Symbolic "which meant that they used a sequence of instructions called algorithm, carefully designed to solve problems. Such was the optimism that it was said that in 20 years computers could do anything that a human being does, but the computers of that time did not have the necessary processing capacity.
The research was stopped for a long time, until the 80s computers were created powerful enough, at that time the main method was the so-called "Expert Systems" that combined an extensive knowledge base with search algorithms that found their way through all A decision tree like playing 20 questions to find an answer.
These systems were very successful and useful for professionals as for example to diagnose diseases, meanwhile the computing capacity was increasing and was exploited to create increasingly complex algorithms capable of traveling trees of incredibly large decisions like those of chess, with incredibly large amounts of possible plays. So in 1977 "Deep Blue" was the first computer to defeat a world chess champion; Garry Gasparov and he did with this method of "Brute Force" exploring the results of all possible moves and anticipating 8 to 20 movements, is able to analyze 200 million possibilities per second something that a human is unable to do.
Even so for the machines it was impossible to perform tasks that for a human does not involve effort to recognize voices, faces, read, identify and manipulate objects at least until the artificial neural networks were invented, these artificial mechanisms inspired by the functioning of the brain have neurons virtual connected to layers of other neurons and in turn to other layers until they reach a layer that yields a result, the funny thing is that at first they are terrible doing their job but these networks are trained and with each training they learn and refine these results.
Today, artificial intelligence has many applications, helps you decide the fastest route in your cell phone maps, recognizes instructions given out loud, controls enemies in many video games, recognizes faces in certain applications, helps doctors to make diagnoses, translate from one language to another, drive cars and recently Google created an artificial intelligence that built another artificial intelligence and although it is very useful brings with it many questions beyond the technological. The US is already building military drones controlled by artificial intelligence, if they decide to shoot ... Who is responsible for the attacks?
A MQ-9 Reaper unmanned aerial vehicle, has the ability to carry both precision-guided bombs and air-to-ground missiles. Source; U.S. Air Force photo/Staff Sgt. Brian Ferguson.
It is alleged that several programs have passed the Turing test, but does intelligence imply awareness? And if so... are these consciences capable of suffering ?, Do we have rights to create machines that suffer?, Perhaps the greatest concern is that of the famous "Technological Singularity" the hypothetical moment in which the artificial intelligences are improved so much same that exceed the human being and escape our control. Entrepreneur Elon Musk is concerned that artificial intelligence even plans to put an end to humanity, but I do not think that will happen soon.
The images used in this article are Public Domain and can be reused according to your copyright.
If you liked what you just read do not forget to share and support me
BIBLIOGRAPHY
- https://www.nytimes.com/2017/11/05/technology/machine-learning-artificial-intelligence-ai.html
- https://www.theguardian.com/technology/2017/dec/07/alphazero-google-deepmind-ai-beats-champion-program-teaching-itself-to-play-four-hours
- https://www.quora.com/What-is-the-difference-between-the-symbolic-and-non-symbolic-approach-to-AI
- http://www.kasparov.com/deep-thinking-ai/
- https://es.wikipedia.org/wiki/Rabbi_Judah_Loew
- https://en.wikipedia.org/wiki/Expert_system