INTERVIEW:
Question 1
Why did you choose the game of chess for this program?
Answer
In 1993 we had this product and we wanted to use chess applications as a way to
understand how to solve large complex problems with this computer. Chess is very complex
so it's a good model for studying architecture.
Question 2
When we spoke in Mediamente about this game, we spoke about it as part of artificial
intelligence (AI), as the computer against the human brain. Is this really the case or is
it something else?
Answer
It is not really about AI. The point was to know it better in itself. It is the scientist
studying the machine: it is the program that plays the game, that plays chess. So it's
about how to program the machine to use it for work. We use the computer as a tool, not to
equate the machine to AI.
Question 3
So the structure of the machine is not really modelled on the human brain, but
rather as a super computer?
Answer
Yes, the way the machine solves problems is really different from the way
the brain solves problems. But in chess, you often arrive at the same result. So we use
the machine in a different way. We use machines as a tool. And we can now use this tool to
solve very complex problems.
Question 4
Let's talk about the game. What was the difference between the first game that Deep Blue
lost and the second one that Deep Blue won?
Answer
The first one was in 1996 and the machine had very little sophisticated chess knowledge.
So after we lost to Pascal, we worked with chess experts like Joe Benjamin to try to
capture the expert's knowledge and try to put that into the program. This year the machine
has not only computed or looked very far ahead it, but also played very accurately with
this chess knowledge.
Question 5
One of the Polgar sisters said, if I had been in the place of Kasparov I might have won,
because I am a woman which means I have intuition on my side. Do you agree?
Answer
The machine does not understand intuition. The woman solves problems differently, so it
doesn't understand intuition or psychology. Therefore, it's not affected by that, so the
machine doesn't know if it is playing against Polgar or anyone else.
Question 6
What do you think of the new boundaries of AI. Where do you think we are going to in AI
with these new machines? How can it change our lives?
Answer
AI is a very important area for research in the computer area. Many experts are studying
AI. But the way the program has to learn to emulate our brain takes tremendously long
computations and the technology is not quite there. Today we have some applications in the
use of AI technologies like speech recognition. If you talk to a machine, often then the
machine can recognise your voice and therefore it can accept your voice and transfer that
into digital information. So there are areas which are making progress. But there are very
large general problems which will take many years to solve.
Question 7
A lot of people are afraid of the power of these machines. Is there a real risk that they
may take over from human beings?
Answer
Historically we have been afraid of machines taking over every time we have new tools, new
technology. In the industrial age people are afraid that the machines will take over. But
they make our lives better. So the more we understand technology, the more we understand
how to use the machines, and hopefully we will use them to our benefit. It is important to
make people aware of technology and how technology can help us solve problems.
Question 8
Now there's a new model of Deep Blue, a sort of home model.
Answer
You mean the small version. Today we're using the small version of Deep Blue which has
only one processor. It's about ten times slower and is much smaller than the one used with
Gary Kasparov. Eventually, this will be another personal computer. We already have many
chess programs on the personal computer. So technology is always improving and every day
that goes by, the more things we can do with computers.
Question 9
Deep Blue as a concept, as a team of people working together towards something, is
finished or is it going to expand into something else?
Answer
For the team we will now use what we learned from the chess program Deep Blue for private
technology and for many other applications areas. For instance, we have already started to
apply this to help us find better drugs by using this computer to simulate molecules so
they can combine to form drugs much faster with the help of a machine like this.
Question 10
It's difficult to give the computer the basic common sense that characterises human
beings. Although on the other hand, I don't know if we have so much common sense...
Answer
The computer doesn't have common sense because we don't understand what common sense is.
That's the problem we have. It we understood that, then maybe we could give the computer
common sense too.
|
