“To Some Extent, All Machines
“To some extent, all machines and devices designed to replace or augment humans’ skills exhibit
This is an excellent way of highlighting that artificial intelligence has been prominent in mankind’s life from the beginning; from something as simple as an animal trap used by Neolithic hunters 7500 years ago to the likes of Apples Siri and Amazon’s Alexa being used as personal assistants. Artificial Intelligence as a term was coined in 1956 at the Dartmouth Conference in New Hampshire by the computer scientist John McCarthy. But, the idea of artificially intelligent machines has existed since Ancient Grecian times when Hephaestus (a blacksmith) manufactured mechanical servants and a bronze man called Talos.
Further historical events that illustrate the advancement of A.I come from Basile Bouchon and his invention of punched paper rolls and the looming machine that was designed to understand it. For 1725, this technology was revolutionary as it was used as a way to automate the process of creating textiles by using holes punched in a roll of paper to control a needle into either picking up a length of string or not. The loom and paper roll contraption exhibits traits that can be classed as intelligence. It may not have been able to solve the Einstein Problem, but it was a step in the right direction for artificially intelligent machines.
Another inventor that improved upon Bouchon’s punched paper was Joseph-Marie Jacquard with the Jacquard Loom. Joseph-Marie is considered the inventor of the first programmable machine as his loom was design to not only emulate what humans could already do but to make decisions based on the card that was presented to it. The Jacquard loom was created in 1801; over 75 years after the creation of Bouchon’s punched paper. This shows the progression of AI as being more exponential than linear as shown by the number of advancements in the field per year.
Additionally, during the middle of the 20th century a science fiction author by the name of Isaac Asimov wrote a novel called I, Robot. To counter the idea that robots and AI will see humans as unnecessary and irradiate them, Asimov created his Three Law’s of Robotics:
1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey any orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.
These laws, though they have been created for the novel, they can be used help engineers and scientists to evolve artificial intelligence without fearing the consequences of becoming inferior. Sometime after the creation of the three laws, a fourth law was created called the zeroth law, used to add some perspective on the other laws
A robot may not injure a human being or, through inaction, allow a human being to come to harm.
This creates a bit of depth to the three laws as it allows AI to handle morality paradoxes, where it is faced with an option where either one person dies or three people die, the robot would save the three people as it will have a smaller impact on humanity.
A test that was devised in 1963 to allow someone to determine if the artificial intelligence has evolved enough to be classed the same as human intelligence is the Turing Test. This was created by the cognitive scientist, Alan Turing, to allow for any sceptic’s disbeliefs in AI not being of a similar intelligence to be debunked. This can allow engineers to track the progress of Artificial Intelligence can allow for adjustments to be made based on the answers. Other advancements that were made in the 1960s were from Kenneth Colby and Joseph Weizenbaum. Kenneth Colby created a program that simulated someone who had a paranoid personality and an obsession with the Mafia. Its responses followed a more aggressive tone and had a dislike of the police and of authority. Whilst the program created by Joseph Weizenbaum was one that was more interested in being a therapist. They’re responses are based on allowing someone to come to terms with any troubles they have and talks to them in a much calmer and overall nicer tone by comparison to Joseph Colby’s program. These inventions are not true artificial intelligence as they can only recognise key words and phrases which trigger a vague response which includes words relating to the topic at hand, if this is not achieved then to ensure a conversation is continued it just uses a neutral sentence or refers to an older topic too revive that conversation. A further step in the evolution of AI is the creation of Terry Winograd’s SHRDLU a much more advanced version of Colby and Weizenbaum’s programs. This program was a lot more advanced than previous attempts as it did not simply just respond to sentences containing key words, it could actually move objects on command and use problem solving to achieve the command given. Terry Winograd designed a virtual world for SHRDLU in which there different shapes and blocks of varying colours. SHRDLU could interpret commands to move and stack these blocks whilst also being able to rearrange them to fit the command. If a command is conflicting with what there is in the virtual space then it actively looks for an alternative or plausible method of completing the task. Another way SHRDLU shows that it has heard and understood a command is by acknowledging it, similarly to how a human would say yes or ok when asked to do something, SHRDLU does so to and he can also answer questions about his actions. When asked “Is there a large block behind a pyramid?” the program then responded with “YES, THREE OF THEM: A LARGE RED ONE, A LARGE GREEN CUBE AND THE BLUE ONE.” This shows us not only is he able to recognise voice commands but he has a memory and can recall what has happened and where all of these objects are relatively in the virtual world. An effective way to illustrate the progression of artificial intelligence is to observe the different creations of the Westinghouse Electric and Manufacturing Co developed between 1927 and 1940. The first creation was Herbert Televox, this was a simple robot that could answer phone calls made to the company, the only way for it to communicate was to make some grunts and buzzes until it was given the ability to speak a couple of sentences. The next creation from Westinghouse Electric and Manufacturing Co was Mr Telelux a robot whose brain consists of two photo-electric cells that interpret light of different frequencies and intensities, which are interpreted into different actions for the robot to do, such as operating an electrical switch, it may sound mediocre but this is was created before World War 2, which was when RADAR was invented, which is a communication system to locate projectiles. The penultimate invention created by Westinghouse Electric and Manufacturing Co was Elektro, a seven foot robot that could seemingly communicate with people and could even smoke. There were a few creations before Elektro were like Katrina Van Televox and Willie Vocalite though they weren’t anywhere near as successful. When first introduced, Elektro was a walking, talking machine that was able to smoke using bellows behind a mouth and motors to control it. It also was able to be controlled through voice commands which caused vibrations in the system, enabling Elektro to do the action, be it walking using the motors in his legs to lift and reposition his feet. It was able to communicate with people provided they used certain phrases or questions; which were enabled due to there being pre-recorded responses programmed into Elektro simply played on a record player. This was a breakthrough at the time and Elektro was not alone, he had a companion by the name Sparko, a robot designed to look like a dog. This robot had the same capabilities as Elektro in that it could bark, sit and beg and could be trained to sit if a bowl of dog food was placed in front of it. Unfortunately for AI there was what was deemed as a “winter” where interest and progression in the field slowed due to a lack of financial support from governments because of the criticism it was receiving. This occurred during 1974-80 but after 1980 there was a surge of interest in refreshing this area of research. This happened due to the British government wanting to give the Japanese researchers competition. Another winter came for research into artificial intelligence in 1987 when the markets crumbled, though as previously seen researchers do not want to let this area of development slip away so in 1993 there was another resurgence of activity and it has not experienced a “winter” since. A significant breakthrough for artificial intelligence came from IBM’s Deep Blue program which was the very first AI to beat Russian chess Grandmaster Garry Kasparov at a game of chess, which happened back in 1997. This is great because it shows a giant leap from Colby, Weizenbaum and Winograd’s programs which could answer some question to Deep Blue being able to adapt to the moves played by Kasparov and also to outsmart him. This also outlines the exponential growth of progression in this field as if we take into account it took 160 years for the responsive programs to be created from Jacquards Loom but it has only take them 30 years to construct a program that was able to beat a human being at a game that they had master over years of training. Another, more recent breakthrough was with the introduction of Siri into Apple products such as their iPhone, iPad and more recent iPods. This programs is a software design to be the user own personal assistant, being able to help in most situations such as needing the weather, a taxi number, train schedules to even typing emails and text messages to others. Something that we have seen happen recently is that pop-culture has taken to the idea of artificial intelligence becoming an integral part of human existence with the fields representation in popular movies such as The Terminator, Isaac Asimov’s I, Robot was transformed into one as well as children even being introduced to this concept through movies like Wall-E. The use of these movies is not only used to entertain people but to also educate them on different problems that could occur if robots were to be given feelings or if they were to be used in the wrong way. This also enlightens the viewers on the “Three” laws of robotics.