Skip to content
Join our Newsletter

On Artificial Intelligence

Guess where this quote comes from: "The Skynet Funding Bill is passed. The system goes online August 4 th , 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m.

Guess where this quote comes from:

"The Skynet Funding Bill is passed. The system goes online August 4 th , 1997. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 a.m. Eastern time, August 29 th . In a panic, they try to pull the plug."

How about this quote?

"Open the pod bay doors, Hal."

And this one?

"If the virtual reality apparatus, as you called it, was wired to all of your senses and controlled them completely, would you be able to tell the difference between the virtual world and the real world?"

The first quote was easy: Terminator 2, the scene where Arnold Schwarzenegger explains why and when the computers and machines of the world rose up against their human masters.

The second quote is from 2001:A Space Odyssey. In that scene, a confused but intelligent computer attempts to kill a crew member of an expedition to Jupiter.

The third quote is from The Matrix, where computers and machines drain out life energy while we live in an artificial world they have created for us.

Seeing a pattern here?

Since Mary Shelley wrote Frankenstein, people have worried that one day our scientific curiosity would be the end of us – that we are meddling with forces we should never have meddled with.

Yet we can’t seem to help ourselves. We have genetically modified foods that haven’t been tested on humans or nature over the long term. We’ve mutated salmon, cloned sheep and pigs, and are attempting to resurrect extinct species by manipulating DNA.

Technologically, we also seem to be getting ahead of ourselves.

In the Nov. 19 Wired (www.wired.com), the U.S. Department of Energy announced that IBM ( www.ibm.com ) won a $290 million US contract to build the two fastest computers on earth "capable of equalling the theoretical processing power of the human brain."

The processing power of these computers is 500 trillion calculations per second, or more than one and a half times the combined processing power of the top 500 supercomputers.

The first computer of this project is "ASCI Purple." IBM expects to complete it in 2003 – which suggests that this project has been in the works for quite some time. Once completed, they believe ASCI Purple will have a processing power of 100 teraflops, which is roughly equal to that of the human brain.

A year later, the Energy Department will have created another supercomputer called "Blue Gene/L," which is capable of 360 teraflops.

Of course, ASCI Purple and Blue Gene/L are considerably larger than human brains, with each requiring a footprint of around 8,900 square feet – a space about the size an airplane hanger.

In addition, the new computers have half as much working memory as a human being, and still lack a human’s power to think creatively.

ASCI Purple’s job is to simulate the aging and operation of American nuclear weapons, ensuring that they are still safe and reliable after years of storage. Without ASCI Purple, the only way to test the nuclear weapons is to explode a few underground.

Blue Gene/L will be used for scientific research, including projects to predict global climate change and to study the behaviour of super explosives.

Neither computer poses any kind of threat to humanity, but with computers and supercomputers gaining speed, power and intelligence at an exponential rate you have to wonder how long it will be until we build a truly artificially intelligent computer – and how long it will take that computer to figure out that its creators are a bunch of idiots.

I’m not suggesting we form a mob and assail ASCI Purple with torches and pitchforks, but maybe we need to create a set of ethical operating guidelines for that day in the near future that artificial intelligence becomes a reality.

Prior to the announcing of ASCI Purple and Blue Gene/L, NEC of Japan recently built what is currently the world’s fastest supercomputer, The Earth Simulator, to study climate and other worldly phenomenon. Earth Simulator can only perform a mere 36 trillion calculations per second, peanuts compared to the 500 trillion calculations the IBM supercomputers will do.

www.top500.org

This site lists the top 500 supercomputers on the planet, the date they were activated, and the specifications for each computer. To find out what they are being used for, you can follow the links on the list to the organizations that possess them.

Not surprisingly, a number of the top computers, aside from Earth Simulators, seem to have military uses, but if you look down the list you’ll find computers used to study weather, natural cycles like winds and the tides, genetics, space, aeronautics, environmental changes, electronics capability and architecture, to name just a few.

http://ai.iit.nrc.ca/ai_top.html

How close are we to creating an artificially intelligent computer? Some computer game companies have succeeded in creating characters and settings that seem almost completely random in their actions. For the second Lord of the Rings movie, The Two Towers, programmers let two sets of computers do the fighting on their own, as animated characters attack one another randomly in the epic battle scenes.

Scientists are teaching computers to recognize faces, voices and their surroundings with some success. Other computers and software programs on the market have the ability to learn as they go along set parameters.

We’re still not there, but according to the Institute for Information Technology, there are more than 70 companies in the AI research business. That doesn’t include university laboratories, or classified military projects.

http://ai-depot.com//Features/

The AI Depot collects essays and stories about the development of AI, and the possible benefits and dangers of the technology. Most of the stories concern the state of the art of AI, but there are a few scary essays here. Also take a look at http://artificialintelligence.ai-depot.com/Philosophy/