Artificial Intelligence

By Ed Raymond
Staff Writer

Coming Soon–A Billion Words On The Head Of A Pin

In the 1968 movie “2001: A Space Odyessy,” a computer named HAL controlled most of the operational functions of an international space station named Discovery—besides keeping track of all the humans on board and answering diverse questions that would have made Google proud. HAL was a very smart piece of artificial intelligence. HAL actually should have won the Academy Award for his role. When the humans on board think that HAL has flipped his transistors, they meet in a space pod so HAL cannot hear their discussion. But HAL has the ability to read lips.

When the real space shuttle Discovery docked 43 years later at the International Space Station in February, 2011, one of its passengers was Robonaut 2, a 300-pound robot with a head, two arms and two hands, and a torso. To celebrate another HAL, NASA should have played “The Blue Danube Waltz” as Director Stanley Kubrick did in the movie when a space craft docked. That would have been fun—and significant. Robonaut 2 was designed to be trained by astronauts to assist them in both inside and outside maintenance of the space station. On spacewalks, as an example, the robot will not only carry some of the tools, it will use them if the task requires great strength. In the short span of less than two generations we have moved from science fiction to science reality.

From Man To Computer

Over 400 years ago Shakespeare quilled these famous words in “Hamlet” in describing humans: “What a piece of work is man! How noble in reason! How infinite in faculty! ...in action how like an angel! In apprehension how like a god!” We have reached a point in the development of artificial intelligence in 2011 where we can e-mail: “What a piece of work is a computer! How noble in reason! How infinite in faculty! ....in action how like an angel! In apprehension how like a god!”

In over 100 hospitals in the United States we now have cart-sized robots called TUGs made by Aethon that travel the halls delivering drugs to the right patients among hundreds of other rather routine tasks, including delivering medical supplies and equipment to operating rooms and nursing stations, delivering meal trays as ordered by patients, and assisting nurses in delivering blood samples to test stations.

TUGs are programmed with a concise map of the entire hospital from basement to top floor. TUGs are equipped with sonar and lasers that keep them running into people and other obstacles. If doors to facilities are closed they can signal hospital staff to let them in. TUGs are programmed to enter only unoccupied elevators to keep patients and visitors from freaking out. If a person enters the elevator on another floor, the TUG reminds them to push the right floor button so “it” can get off. Monitors keep track of all TUGs in the hospital. If a patient or staff member starts to fool around with a TUG you might hear this response: “What did I ever do to you?”

The Saga Of A SWAT Robot

We have all seen law enforcement robots used to inspect suspicious packages left in airports and other public places. Some models can climb stairs, open doors, sniff for bomb materials while sending messages back to handlers. Recently, police in West Melbourne, Florida sent a SWAT team to check out a residence where a man was suicidal and threatening police with a “Holocaust.”

Instead of trying to enter by the front door, the police sent in a $65,000 camera-equipped robot to check out the interior of the house. The robot opened the front door and proceeded to check out all rooms. When the robot pushed open the door to the bedroom, a naked man with an AK-47 came out and started to shoot at the robot. The robot came away with a blown-out camera and broken cables and wires, but no humans were hurt. After its heroic act, the robot was transported to a robot hospital to recover from its wounds.

The Japanese and South Koreans are ahead of us in the development of working robots. South Korea presently employs about 30,000 foreigners to teach English, but plan to gradually replace the “expensive” foreign teachers with English-speaking robots. Several schools in 2010 used robo-teachers in classrooms. Now, 18 schools have English robo-teachers on “staff.” The robo-teachers, brightly-colored, fat “androidy” types, will be part of a huge effort by South Koreans to make their students competitive in English in the international arena.

Every 18 Months Technology Doubles Its Capacity

In 1974 a megabyte of random-access computer memory called DRAM cost $34,820. Today you can buy that capacity for about a penny. An economic research paper has essentially proven that technology doubles its capacity about every 18 months. In the business this “proof” is known as Moore’s Law.

Intel, an American company which is the largest manufacturer of computer chips in the world, is spending about $15 billion in building chip factories and conducting research this year. They are going to produce a chip from 22-nanometer technology that will allow Intel to pack 10 million transistors in a space the size of this period. This is very tiny stuff. Much smaller than the width of a human hair. A human hair is approximately 80,000 nanometers in width. A nanometer is one billionth of a meter wide. The testing of such a tiny chip involves the possibility of up to 1,500 defects in the width of a human hair. If you think you are well-scrubbed, read what chip workers have to do to become “clean.”

A typical Intel microprocessor today has a minimum of one billion transistors capable of switching off and on at about 300 billion times a second. This stuff is beyond boggling. We are also capable of measuring tiny things. As an example, the recent Japanese earthquake shortened the earth’s day by slightly more than one-millionth of a second and shifted the earth’s axis by 6.5 inches. One 24-hour rotation takes 86,400 seconds. If we keep having earthquakes some day it might be a big deal.

This May Be What Life Will Be Like In 2100

The book “How Science Will Shape Human Destiny And Our Daily Lives By The Year 2100” by Michio Kaku contains the predictions of 300 scientists who are working in the fields of computers, medicine, nanotechnology, space exploration, and energy. This is Twilight Zone stuff.

:: We will have x-ray vision and be able to move things with our minds.
:: We’ll operate computers and other devices by blinking while wearing wired contact lenses.
:: Micromachines taking up space smaller than a period will perform surgeries inside the body.
:: Toilets will check our excretions daily to determine if we have any disease.
:: MRI machines will be about the size of cellphones and will be kept in a bathroom drawer     at home.
:: Sensors in clothes will call 911 if we are hurt.
:: Zoos will have animals roaming around that are now extinct. We may be able to clone       Neanderthals and give them a choice of attending Harvard or living in a zoo.
:: Robots will be designed to hunt, track, and kill humans, perhaps even Osama.
:: Several major American cities will be underwater because of global warming. Others will     be surrounded by enormous seawalls.
The Challenges Of Checkers And Chess Are Over
The human brain in this technological age is constantly being challenged—by the human brain! Google Chairman Eric Schmidt (Google currently has one billion users a week) says this about computers: “One way to think about computers and people is that the computers are really good at some things that we are no good at, at all. They remember billions of things incredibly accurately, and none of us, even the most brilliant of all of us, do that. And they can search across that in, literally, milliseconds. And come up with all sorts of interesting results that even we, with all of our wonderful reasoning and intuition, can’t do….(Humans) have the ability to recognize patterns in ways that computers are unlikely to for a reasonably long time.”

It may be tragic, or it may be a sign from some superior being, but in the beginning of this technological age there is no man on earth capable of beating a computer in the game of checkers or chess. The simple game of checkers was solved about a decade ago. Dozens of computers running from 1989 to about 2000 mapped completely the 500 billion possible plays in the game of checkers, effectively turning the checkers championship to a computer like HAL.

When the IBM computer Deep Blue beat world champion chess master Garry Kasparov in 1997 it signaled to us: “Keep up or get the hell out of my way!” IBM’s Watson, the world champion Jeopardy champion, reinforced this idea by trouncing two champion Jeopardy players at one sitting. But Watson weighs 1,190 pounds, took four years to program, and uses 2,800 processors. Our brains weigh about three pounds, took six million years to program, and has one billion neurons as processors. Another consolation we have left at this point is that Deep Blue and Watson still have a great deal of trouble getting the proper meaning of this sentence: “One morning I shot an elephant in my pajamas.”

Computers May Put The Bottom Feeders Out Of Business

The Ford Motor Company has spent the last three years working with Nuance software so Ford owners can talk to their vehicles about a wide range of complex problems. 2008 Ford models recognized 100 words. In just three years Ford has increased the total to 10,000 words and phrases. It gives a lonely driver somebody to converse with. Heck, Shakespeare only used 20,000 different words in writing all his plays and poetry!

But this is where the computer can separate the chaff from the buckwheat, as my Dad used to say. Back in 1978 the Justice Department sued the CBS network for breaking anti-trust laws. The network paid many lawyers and paralegals over $2.2 million to examine six million documents for evidence they felt they could use in its defense. In January, 2011 a new business called Blackstone Discovery of Palo Alto, California helped analyze 1.5 million documents in a legal case and submitted a bill for less than $100,000.

Computer programs have become so sophisticated that they can discover relevant terms and words in documents at lightning speeds—and at the same time extract relevant concepts that expert lawyers can then examine at their leisure. A chemical company recently examined legal cases going back 20 years to see how accurate their lawyer teams had discovered relevant documents. New “discovery” software used to determine the accuracy of human lawyers found that humans were only 60 percent accurate! Computer and software experts estimate that as discovery software increases in sophistication it has the potential in the shift from human to computer to reduce manpower to one lawyer for work that 20 years ago would have required 500 lawyers. If this is true, ambulance drivers will have to sneak around roads and streets to avoid the chasers.

Inventor Ray Kurzwell, recognized as the leading authority on the future of artificial intelligence, predicts that, “By 2045 we’ll multiply the current level of intelligence of what is already a human-machine civilization a billionfold, which is really not so fantastic when you consider how far computers have come already…It’s a singular change in human history.”
In the “2001” movie when HAL’s power circuit is finally unplugged by Dave, It-He says: “Dave my mind is going. I can feel it.” I wonder if we will be able to unplug the HALs of the future like that.

Questions or comments: .(JavaScript must be enabled to view this email address)

[Editor’s Note: By the time this story went to press, Osama had been hunted and killed. It is unknown whether robots were employed in Osama’s death.]