Dr. David Levy Interview - Loebner 2009 winner.
Dr. David Levy already has many claims to fame, the Loebner Prize is only the latest feather in his cap and if you did not know; this is not the first time he has won the competition. I quote from Dr. Levy's biography on Wikipedia :
David Neil Laurence Levy (born March 14, 1945, in London), is a Scottish International Master of chess, a businessman noted for his involvement with computer chess, and the founder of the Computer Olympiads and the Mind Sports Olympiads. He has written more than 40 books on chess and computers.
Levy was born in 1945-03-14, in London. He won the London Junior Championship in 1965 and 1966. He won the Scottish Chess Championship in 1968. He tied for fifth place at the 1969 Praia da Rocha Zonal tournament, scoring over two-thirds and thereby obtaining the title of International Master. He played on Board One for the Scottish team at the 1972 Chess Olympiad in Skopje, Yugoslavia, scoring six wins, five draws, and seven losses (47.2%).
Levy became a professional chess writer in 1971, and has been prolific. Several of his books were co-written with English Grandmaster and prolific chess author Raymond Keene.
In 1997, he led the team that won the Loebner Prize for the program called "CONVERSE". Since 1999, he has been the president of the International Computer Games Association.
He was Chairman of the Rules and Arbitration Committee for the Kasparov vs Deep Junior chess match in New York in 2003.
Levy once started a business called Tiger Computer Security with a famous computer hacker, Mathew Bevan. Now he is the Chief Executive Officer of Intelligent Toys Ltd, a London-based company that develops toys that incorporate AI.
Levy also wrote Love and Sex With Robots, published in the United States in 2007 by HarperCollins, and forthcoming from Duckworth in the UK.
Source : Wikipedia
Here now follows a short interview with Dr. Levy discussing the winning chat-bot 'Do-Much-More' and aspects of the Loebner Prize.
How long did 'Do-Much-more' take to develop ?
Its predecessor, Do-A-Lot, took around two and a half years or thereabouts. Most of that time we had one programmer and a linguistics expert doing the bulk of the work, and some of my time. Do-Much-More is based on Do-A-Lot and required about 6-7 months by one programmer, perhaps two months by another programmer, and about 6 months by our linguistics expert, plus some of my time.
Do you think your grounding in chess helped ?
No. What helped most was coming up with a new way of doing things.
Do you plan on developing this chatbot further ?
Yes. We have devised an approach that will allow us to enable the chatbot to conduct conversations in any specified subject domain (once we have provided it with the relevant data for that domain). So it could be given the persona of an Italian food freak, a Manchester United fan, .....
Any plans to release 'Do-Much-More' to the general public ?
Not yet, but eventually yes! But first I want to take it through the next quantum leap in performance as described above.
What is it that fascinates you about chatbots ?
I am generally fascinated by any software that simulates human behaviour, and conversation is one of the most challenging tasks in this field.
Are there any techniques that you see coming to the fore in chatbot development ?
Just as I and my team have come up with significant new ideas, so I believe that some other researchers will do likewise. Most likely the best new ideas will take advantage of faster hardware and bigger memory capacities.
What in particular pleased you about your bot's performance ?
When I first saw the transcripts of its conversations with the judges, which was not until after the competition was over, I was delighted at how well our approach worked in practice. Here are some examples:
Judge-2: Hi there, are you here for interspeech?
Do-Much-More: Hello. How are you doing?
Judge-2: I'm fine - gosh you're a quick typist!
Do-Much-More: I've no doubt you are fine, but is it a blessing or a curse? If I offered you a potion that would make you less fine, would you drink it?
Judge-2: No, I wouldn't drink it - though if it had other magical properties I might...
Do-Much-More: Maybe you wouldn't, but a therapist has been trying to persuade you to drink it.
Judge-2: I don't have atherapist - do you?
Do-Much-More: I expect I do. I don't have very much choice.
Judge-3: I'm Jon
Do-Much-More: Good to meet you Jon
Judge-3: What do you do when you're not doing the Loebner prize?
Do-Much-More: Oh, I do much more than I intended to do at first. I got more interested in the job as I went along.
Judge-3: Is that difficult?
Do-Much-More: No, it isn't at all difficult. It is as peaceful as a platypus playing with a potato pudding.
Judge-3: T~You have a fine line in metaphors
Do-Much-More: I sometimes have a fine line, but that's not what I do best. I'm better at growing antirrhinums and mimosas.
Judge-3: Ah a horticulturalist?
Do-Much-More: Well I don't know how much a horticulturalist weighs, but I reckon the average horticulturalist must be worth twice that much in gold.
What is your view on the falling number of competitors [in the Loebner Prize] ?
I think it is a great pity, because competitions such as the Loebner Prize do a great job of enouraging new research in this field, just as the regular computer chess championships of the 1970s and 1980s made a big contribution towards the advances that eventually led to the defeat of Garry Kasparov. From this point of view Dr Loebner has made, IMO, a big contribution to the field, and a way should be found to make the competition more appealing to chatbot developers. For example, in Brighton we were only able to follow the conversations of one of the programs - the one that finished in last place! Even more important would be carrying all the conversations live on the Internet, as we do at the annual World Computer Chess Championships (and other chess events). I am sure that these improvements will happen soon.
Previously we have asked past winners if they had any advice to pass on to budding bot makers - is there anything you would like to suggest ?
Yes. Don't be constrained by what other people have done. If you think you have a good original idea follow it up.