Can machines think?

In '2001: A Space Odyssey', Arthur C. Clarke speaks of a space mission to Jupiter, manned by three astronauts and the HAL 9000 supercomputer, supposedly infallible. As the spacecraft travels towards it's destination, HAL alerts the team leader to the impending failure of the AE-35 transmission unit, which keeps the radio antenna pointed at Earth.

When the unit is brought in for testing, the astronauts find it to be in perfect working condition. Faced with this error by a supposedly error-free machine, the astronauts decide to disconnect HAL and continue the mission without him.

HAL, programmed to ensure the success of the mission and faced with the threat of elimination, decides to eliminate the human crew first. However, he is not entirely successful - the team leader survives, and succeeds in disconnecting HAL from his memory banks as HAL begs for his "life".

In 1950, Alan Turing proposed what has come to be known as the "Turing test" for artificial intelligence. In the Turing test, an interviewer in one room is conversing with a computer and a human, located in another room. The computer is said to possess "intelligence" if the interviewer is unable to detect any difference between the answers of the two.

"2001" was probably one of the first books to explore the concept of artificial intelligence, at a time when the concept was both unfamiliar, and a little scary. But new research over the past few decades has led to an avalanche of new discoveries and inventions that harness cheaper computing power to make artificial intelligence less of a fantasy, and more of a reality.

Now, if only someone could come up with a way to force-feed some of this intelligence to our revered political leaders - but that's another story...

We're not too sure how keen we are on toasters that talk, cars that drive themselves, and robots that do our every bidding - but in the interests of our readers [the things we do for you guys! :-)], we thought we'd take a look at the current state of the AI union. And so, we've compiled a list of some of the most interesting artificial intelligence Web sites around, and pointed you to some sites that are sure to set you thinking!

Take it away...

The BioRobotic Vision Laboratory

We'll start with the BioRobotic Vision Laboratory, a lab that conducts research into insect vision, navigation and cognition, and uses these principles to devise algorithms for machine vision, and novel kinds of autonomous, visually-guided robots. They've already designed and built a corridor-following robot and an obstacle-avoiding robot - movies are available on the Web site.

The Face Recognition Home Page

Information on face recognition using neural networks, facial analysis, facial animation, pattern recognition, and a list of face databases are just some of the things you'll find here.


Cyberlife first hit the headlines with the news that their AI technology, currently available in consumer form in their Creatures product, was being analysed by the military for use in robot-driven warplanes. Since then, there's been no looking back for this company, which uses its proprietary Artificial Life technology to create focussed business units, autonomous robots for civil and military applications, and so-called "designer life-forms". In particular, check out the very interesting articles on evolution and life, written from an AI viewpoint.

The University of Berkeley's Robotics Department

Everything you ever wanted to know about robotics is available on this site, with information on topics as diverse as micromechanical flying insects, micro-robotics, motion planning with nonholonomic velocity constraints, robotics tools, RISC robotics and algebraic and geometric computing for robotics and vision [man, this copy-paste function is pretty cool! ;-)]

The Harvard Robotics Laboratory

The Harvard Robotics Laboratory was founded in 1983 and contains information on current research projects, including computational vision, neural networks, tactile sensing, motion control and VLSI systems. It also has links to the Berkeley Human Engineering Lab, the Berkeley Robotics and Intelligent Machines Lab and other Internet resources related to robotics.


The word 'anthrobotics' has been derived from the words 'anthromorphic' and 'robotic' to signify a new generation of robots capable of never-before-seen dexterity and flexibility. These robots have industrial applications in the fields of space station assembly, satellite servicing and planetary exploration, welding, sealant application, spray painting, inspection, and hazardous material handling.

Handwriting recognition

Research on the use of human language models in performing handwriting recognition. This research holds promise for pen-based human-computer communication, handwritten document recognition and understanding the role of language in interpretation of writing.

Alife Games

Starting out as an small outfit with the sole purpose of creating a truly intelligent gaming environment, where everything a gamer sees is significant and influences the game's progress, Alife Games has developed a set of artificial entities which get out there to seek you and kill you!

These games have been developed using Microsoft's Direct3D technology, and the most interesting fact about these games is that the source code is available under the GNU Public License. Remember, you heard it here first!

Cellular Automata

"Life might look like a mathematical game, but it lives up to its ambitious name."

Cellular Automata is a branch of AI research which investigates "the truth which is out there". It primarily consists of computer simulations that try to emulate the laws of nature. While not promising the moon, it does point out that our seemingly topsy-turvy and roller-coaster way of life can be simulated by a set of relatively simple equations.

Princeton Cognitive Science Laboratory.

Cognition can be defined as "the science of knowing and perceiving things". Cognitive science involves research on language, thought, memory, and learning in the fields of visual and auditory perception, experimental psychology, and mathematical psychology. Princeton's Cognitive Science Laboratory provides a great deal of information on this topic.


A wag once said that finding relevant information on the 'Net was more difficult than finding a needle in a haystack. But then he found KDNuggets, an excellent online data mining resource for anyone interested in finding out just how AI can be used to increase your online productivity. The site has a massive database of the various technologies, used in sectors like telecommunications, banking et al.

Expert Systems

Though there exist many varying definitions for an expert system, all of them stress on one very vital fact: an expert system is an artificial life form doing what humans are so good at - giving advice! And once you've digested that one, check out the second link, which tells you all about PROLOG, the language that is primarily used for applications such as expert systems, natural language, and intelligent databases.

Virtual Pets

"Ronnie, take the dog out for a walk! Ronnie, give Kitty a bowl of milk! Ronnie, clean the fish tank!"

After the initial excitement of getting a pet dies down, the task of taking care of them eventually becomes a burden for many. But with the arrival of a new generation of virtual pets, you no longer have any worries - check 'em out!



Wanna know more about the Turing test? Or even talk to your computer and have it respond? Well, you should check out ELIZA, one of the more famous AI programs available online. We also tried out MegaHAL, which was pretty good too.

And before we go, a site maintained by David W. Aha, which contains links to numerous AI and machine-learning sites - for all those rainy days when you have nothing to do!

As for us, we're going to see how the so-called "intelligent" search engines deal with our search request for "+wealth +fame +love -taxes -hate -viagra". See you when we've figured that one out ;-)

Till next time - stay healthy!

This article was first published on 14 Apr 1999.