Martin Spano is the author of Artificial Intelligence in a Nutshell, a book that explores the mystified subject of artificial intelligence (AI) with simple, non-technical language. Spano’s passion for AI began after he watched 2001: A Space Odyssey, but he insists this ever-changing technology is not just a subject for sci-fi novels and movies; artificial intelligence is present in our everyday lives.
The story of Garry KasparovRead also:Read more
On April 13th, 1963, in Baku, Soviet Union, now Azerbaijan, Garik Kimovich Weinstein was born. At an early age, watching his relatives solve a chess puzzle in a newspaper, he proposed a genial solution, and his immense natural talent for the game was recognised. In 1978, organisers of a memorial chess tournament in Minsk invited him as an exception to participate. He unexpectedly won. Seven years later, he became world champion after defeating his archrival Anatoly Karpov. From this time onwards, he was almost unbeatable - until 1997. In this year, Garry Kasparov, the name he used from the age of twelve, was considered the best chess player of all time by many. It was rumoured that if someone was to beat this chess-artist, they would have to be inhuman. In May 1997, one of the most watched chess matches of all time took place in New York City. Garry Kasparov on one side, Deep Blue, a supercomputer created by IBM, on the other. A year before, Garry Kasparov won. This time, it was a very close match once again, but Deep Blue won with 3.5 to 2.5. It was the first time a computer had defeated a reigning world chess champion under regular tournament conditions.
What is artificial intelligence?
In short, it is intelligence demonstrated by machines as opposed to intelligence manifested by humans. It is an area of study in computer science that tries to recreate what the human brain does: virtually perceive the world, understand and respond to speech, learn, plan and solve problems. As it is a computer that performs these functions, it is the software that brings about this intelligence.
The computer winning over Kasparov is an example of artificial intelligence. I suppose, being influenced by sci-fi movies like Terminator, RoboCop or Wall-E, some would expect artificial intelligence to be something else. These movies also involve examples of artificial intelligence, but a slightly different kind. Here are the various types of artificial intelligence:
Artificial narrow intelligenceRead also:Read more
Also called weak or applied artificial intelligence because we apply it to one specific domain. Whereas in 1997, only supercomputers like Deep Blue could defeat reigning world champions in chess, nowadays, an average smartphone with a decent chess app possesses so much computational power that it can defeat any human chess master. However, if you try to use this chess program for speech recognition, it will not understand a word. Artificial narrow intelligence is the type of artificial intelligence we have today. It powers autonomous cars and intelligent loudspeakers. It filters spam in emails, displays relevant information on your newsfeed, aids in searching the web, displays relevant advertisements, suggests movies and music you might like and products you might want to buy. It translates between languages, offers the best routes in the case of traffic jams and so on. It is omnipresent in our everyday life without us noticing. It is much better in the task it was trained for than humans, but it can be used only in this specific domain. CIMON, a robot accompanying astronauts on the international space station, is also an example of artificial narrow intelligence because his abilities are limited and do not represent artificial general intelligence.
Artificial general intelligence
Also called strong or full artificial intelligence. It is the Holy Grail of artificial intelligence research. A machine with such capabilities can perceive and respond to its environment in the same way a human can. It can, therefore, execute all intellectual activities a human can. Examples of artificial general intelligence are Hal, Terminator, Wall-E, David from AI: Artificial Intelligence or Ava from Ex Machina.
Although it remains uncertain whether we will achieve such a level of intelligence in the future, according to most scientists working in the field of artificial intelligence we are just decades away from reaching this milestone. There are several approaches that researchers are taking to achieve this level of artificial intelligence. Combining artificial narrow intelligence from several domains might, according to some researchers, be one way. Theoretically, such an intelligent agent would delegate the problem in a particular area to the artificial narrow intelligence it controls that can process this problem. If more domains are needed, it combines the output of these.Read also:Read more
Another examined approach is brain simulation. Why try to create something new when we can inspire ourselves by the object that does the job of intelligence the best in the entire universe: the human brain. With the latest advances in neuroscience and nanotechnology, we could one day reverse-engineer its structure, get down to how it works and recreate it. The Japanese "The Whole Brain Architecture Initiative" and the European "Human Brain Project" are attempting to do just that. Already today, as we see in chapter machine learning, advances in the understanding of the human brain have helped us progress in the field of artificial narrow intelligence.
Other approaches have also been tested. Success in creating artificial general intelligence would have tremendous impacts on our society and view of the world. For the first time in known human history, there could be an intelligent agent with the same intellectual power as we have.
Would such a being be self-conscious or have self-awareness? If yes, the impacts would be even more profound. We would have a sentient being capable of having feelings. It is possible that there would be a rights movement for such beings similar to the animal rights movement.
If artificial general intelligence is capable of any task, it could also reprogram itself. As it would not need to rest and has ever-increasing computational power, it could do so in never-ending recurrent cycles of self-improvement, with each period being an improvement of the previous one. Such an intelligent agent would quickly change from being on a par with humans to superseding us in all areas and exceed all human intelligence combined. This process would lead to an intelligence explosion and possibly to an event called the technological singularity that would bring about changes to human civilization we would not understand. The possibility of such an event is nearer than you might think because technological progress is not linear, but exponential. To fully understand the power of exponential, let's look at the following story.
The story of chess
Nearly 1500 years ago in India, a skilled mathematician invented the game of chess and presented it to his king. The king was impressed and amused and offered the mathematician a reward. The mathematician asked to put a single grain of rice on the first square of a chessboard, twice that amount on the second and so on, the next square doubling the amount of the previous square. He asked just for the amount of rice on the last square of the chessboard. The king laughed at the seemingly humble proposal but executed the mathematician when he found out the actual amount of rice he would have to give him; It would be more rice than humankind has cumulatively produced in its entire history.
So, if you think that the world has progressed rapidly in recent years, then fasten your seatbelts. We can imagine a comparison of someone with a low IQ and someone with a high IQ, and we consider this difference to be significant. An ordinary village Joe has almost the same IQ as, say, Einstein, compared to the IQ of a superintelligent being, which is equivalent to a thousand or even a million. We have no idea how such a superintelligent being would treat us or even if it would think of us at all because from its perspective we might be as unimportant to it as ants are to us. It is improbable it would deliberately want to harm us. However, if we get in the way of it reaching its goals, it would just terminate us. Consider ants once more. We do not usually intentionally harm ants if we come across them. However, when we build a new shopping center and there is an anthill in our way, we destroy it without a second thought.
On the other hand, such intelligence might also be hugely beneficial. Imagine something with such immense power it could read and understand all books and scientific papers ever written in a day. It could come up with ingenious solutions for solving problems like global warming in a matter of hours. It could literally advance scientific progress with the speed of light.
The Slovak Spectator will publish more extracts from Spano’s book in the coming weeks.
2. Aug 2019 at 10:00 | Martin Spano