Moving from Robotics to Artificial Intelligence
by MiracleMax
For anyone that grew up in the 20th century, the future promised the dawn of flying cars, jetpacks, and artificial intelligence companions.
We watched them on television and in the movies. Cartoons such as The Jetsons and Futurama, movies including Forbidden Planet, and television series like Lost in Space paved the way for our preparedness for AI in everyday life. Yet here we are, nearly 10 years into the 21st century, no closer to welcoming these beings into our mist than we were in the last century.
For as long as we have written history, man has dreamt of creating machines to help accomplish difficult tasks. s those dreams have become realities, we have moved towards creating a new form of robot, one that possesses the Artificial Intelligence (AI) needed to solve problems.
"The word 'robot' is often loosely used: it can denote nothing more than a box of electronic tricks able to automate some trivial task, or it may indicate a highly sophisticated humanoid system equipped, perhaps, with dexterous fingers to deal cards or play the harpsichord." (Simons 12)
It wasn't until the year 1955 that the term "Artificial Intelligence" was created by John McCarthy and Marvin Minsky.
They used this term "To describe modern computers with some ability to think like human beings." (Angela 41)
The company iRobot has created several home robots that most people are familiar with. Priced anywhere from $99 to $500 each, people can bring Roomba, Scooba, Verro, Dirt Dog, or Looj into their homes to aid in vacuuming, floor washing, pool cleaning, shop sweeping, and gutter cleaning. All of these are time-consuming tasks that many of us do not enjoy.
Many people look forward to the day when movie and television portrayals of artificial intelligence such as Data from the TV series Star Trek: The Next Generation and Sonny from the movie I, Robot become part of our everyday lives.
The history of robotics, or automaton, as it was called, is a long one.
Many of the first robots were for entertainment purposes. When Archytas of Tarentum created "The Pigeon" in 420 BCE, it paved the way for future inventors. It was simply a wooden bird connected to a string which "flew," propelled by a jet of steam.
In 200 BCE it is believed that the earliest "automata" was created by a group of Chinese artisans. Their contribution was a mechanical orchestra. Leonardo da Vinci is credited with creating the first known documented design for a robot in 1495 CE. While he never built his medieval knight, which was designed to mimic human movement, others later created similar robots based on his designs.
It is during the Renaissance period that clockmaking lent its advancements to the further creation of more detailed automata in Europe.
In 1745, the first robot was made to help improve industrial advancements. Jacques de Vaucanson created a punch card driven loom, which was completely automated. Weavers of the French textile industry felt threatened by this new technology. His suggestions and invention were ignored until 1801, when Joseph Marie Jacquard re-introduced the automated loom, this time successfully.
In 1804, "The American engineer and inventor, Eli Whitney, introduces the concept of mass production, using interchangeable parts and the organized construction of subassemblies into complex manufactured items." (Angela 31)
Using this new concept in his Connecticut factory, Whitney was successfully able to mass-produce rifles for post-Revolutionary War America. Whitney's changes to manufacturing lead to a new concept known as the assembly line.
In 1913, Henry Ford introduces the moving automotive assembly line to help make the Model T more affordable.
George C. Devol, Jr., "The Grandfather of Industrial Robotics," expands on the assembly line idea in 1954 with his unimation device.
"Devol's unimation is the first industrial robot, a system designed specifically to pick and place objects in a factory environment." (Angela 40)
On December 29, 1959 Henry Ford's moving automotive assembly line is taken a step further with the introduction of the Unimate industrial robot at a General Motors die-casting plant. By 1961, the Unimate is unloading hot die-casts, cooling components, and delivering them to the trim press, creating greater efficiency on the assembly lines through a process called telecherics.
Programmable Universal Machine for Assembly (PUMA) is introduced in 1978 by Unimation and it quickly becomes the standard for commercial telecherics.
In 1985, robots are introduced to the medical field with Dr. Yik San Kwoh's robot-software interface which allows the steady hand of a robot to perform delicate brain surgery, aided with the three-dimensional CT scan image to help guide the robot to the brain tumor.
SRI is funded by the National Institutes of Health in 1990 to research the possibility of using robots in minimally invasive surgery and remote surgical tasks. Intuitive Surgical is formed in 1995 and creates "The medical technology necessary to apply modern telerobotic technologies to minimally invasive surgery and microsurgery." (Angela 50)
By removing the chance of human error by replacing a shaky hand with a robot, the patient becomes safer and the procedure more accurate.
The Honda Motor Company began developing human-like robots in 1986, with the hopes of integrating them into everyday human lives. In 2000, their hard work paid off when they introduced Advanced Step in Innovative Mobility (ASIMO), a humanoid robot.
ASIMO not only walks, but can also perform daily tasks in society. On December 13, 2005, the latest ASIMO was revealed. This more sophisticated model can interact in a professional environment, serving drinks or answering phones. Sony introduced its AIBO robotic pets in 2000. Using software, the dog can develop from puppy behavior to a mature dog and even obey commands. Although these models are mimicking the Artificial Intelligence seen in movies, they are still based on simple algorithms, and not problem solving on their own.
Inspired by the advances in modern robotics, John McCarthy and Marvin Minsky founded the first artificial intelligence lab in the year 1955 at MIT. By 1963, "Machine intelligence experts soon start attempting to develop artificial neural networks that function in a manner loosely based on how the human brain functions with its network of neurons." (Angela 43)
It is the work of the 1963 Nobel Prize winner, a neurophysiologist from Australia, Sir John Carew Eccles, that inspires them to pursue this development. They started with simple tasks, to see if the computer could learn.
"In 1965 our goal was to build a machine that could do things that children do - such as pouring a liquid into a cup, or building an arch or a tower with wooden blocks." (Minsky 150)
It took hundreds of mistakes and several years of creating a program called Builder to allow the machine to comprehend the task at hand.
The MIT Museum located in Massachusetts hosts an ongoing exhibit titled "Robots and Beyond: Exploring Artificial Intelligence @ MIT."
Patrons who visit this exhibit are actually part of an ongoing experiment. Cog, who was developed between 1997 and 1998 "Is the fundamental hypothesis that the creation of humanoid intelligence requires humanoid interactions with the world." (Angela 380)
Also on exhibit is Kismet, who was developed between 1993 through 2000. Kismet relays its needs and wants to humans through gestures, tones, and facial expressions. A series of mobile robots are also on display as part of MIT's research in assisting humans who have lost mobility.
Artificial Intelligence is very much still in its infancy. To be able to develop thinking robots that would be able to help people in day-to-day tasks would mean greater freedom of time, money, resources, and greater independence. If the current medical robots were able to diagnose and treat patients by using protocols and algorithms, human error malpractice issues could be avoided completely. However, a new problem would arise of mechanical error or malfunction. Even if this were only a fraction of human error, it would be an improvement and would ease the minds of patients undergoing invasive procedures.
In order to move from the current robots that are available for home use to artificial intelligence robots, we would need the Roomba to be able to multi-task. It would need to decide if the floor needed sweeping, mopping, or washing. It would need to be able to hear a plant get knocked over, be able to travel to the plant to see if there is dirt on the floor, choose what function is needed to clean it up, and pick up the plant and put it back in its original position. A robot would need to be able to see a human shiver and choose if it should turn up the heat, get a blanket, start a fire, or get a thermometer.
As prepared as we may be to welcome AI into our everyday lives, we are minimally decades away from seeing it become a part of our mainstream. It will take additional years of funding, research, and dedication to bring the works of John McCarthy and Marvin Minsky to fruition.
In the meantime, the entertainment industry will continue to inspire and tease us with their creative use of AI while raising important questions regarding changes to our society as a result.
References
Angela, Joseph A., Jr. Robotics - A Reference Guide to the New Technology Westport: Greenwood Press, 2007
Minsky, Marvin The Emotion Machine Commonsense Thinking, Artificial Intelligence, and the Future of the Human Mind New York: Simon & Schuster, 2006.
Simons, Geoff Robots - The Quest for Living Machines New York: Sterling Publishing Co., Inc., 1992