Jonar Nader speaks about the world chess champion who played against Deep Blue, IBM’s supercomputer. Jonar explains the use of heuristics and algorithms. To listen to an excerpt from the radio broadcast, please click on the green play button below.
Here is a transcript of the audio file.
Host: Last Sunday Garry Kasparov was beaten at chess by a computer. Kasparov was the world champion. This is significant because he is also rated as the all time greatest. So, are computers smarter than human beings? What is the significance of phrases such as ‘Computers that think’? Does artificial intelligence pose a threat to humans? At what point do we think the blob of metal will start to dictate its own terms? With me is Jonar Nader the manager of Technology at the Australian Information Technology Society. Were you happy about the win of Deep Blue the IBM computer?
Jonar Nader: Well, there are two kinds of people in the world saying, ‘Go,Go, Go for it Deep Blue, and then there are the others saying ‘Oh please not another milestone of some computer’s supremacy!’ I mean it’s bad enough to fight with cloning and other issues in life and now computers are starting to have brains, but I think that there is a lot of misconception about that because that computer did not necessarily have a brain it had a very clever set of algorithms and very clever set of heuristics and I’ll explain what they are in a moment.
Host: Although Kasparov himself and other people who were watching it were suggesting that the computer was applying some forms of strategy.
Jonar Nader: Well there are two sides to the computer, the hardware and the software. The hardware folk certainly primed it so that its’ 256 processors could work in tandem ever so brilliantly. The software folk had a huge data base. They had, for example, the last 100 years of the best opening moves and the last 100 years of the best closing moves so it had a database on which it could draw. Unlike older computers from back in the 1950’s when they started this notion of artificial intelligence, back then the reason computers couldn’t win at chess was because a computer was trained or programmed to try out every possible combination, and because there is a three minute limit per move, the computer always bombed out because the computer just didn’t have the power to do them all. Now today what they are using is a term called Heuristics. A heuristic is a rule of thumb, so that the computer can use rules of thumb rather than absolutely rigid algorithms. That is where people are starting to see rules of thumb or heuristics as thinking, whereas they are not. They are just a bit of guess work. You can programme a computer to do guesswork for you and that is why it has won.