AI Zone Admin Forum Add your forum

NEWS: Chatbots.org survey on 3000 US and UK consumers shows it is time for chatbot integration in customer service!read more..

Interview with Margaret Boden
 
 

http://www.youtube.com/watch?v=5dEXIOiAsaw

Margaret Boden has been a prominent researcher in the fields of cognitive psychology and artificial intelligence for many decades. Her book “Artificial Intelligence and Natural Man” has long been one of my inspirations.

In this interview she answers a number of questions which are of great interest to the members of this forum. She talks at length about topics such as where neural network research is getting it wrong and how it could be improved, and the strengths and weaknesses of IBM’s Watson including the real reason it was able to win Jeopardy.

My favorite part of the interview is where she talks about criteria for accepting or rejecting artificial general intelligence and what the Turing Test was really all about.

 

 
  [ # 1 ]

I thought it was a nice interview from the standpoint of seeing a prominent AI person give their views (I used to own her book), good photographic quality, and non-excessive length. As for content, I thought most of what she discussed was either obvious (at least to me) or was already covered in this forum (especially the need for real-world data processing like physically pushing buttons), except for the work on autism (of which I was unaware, but is a high-level enough phenomenon that I believe it’s unlikely to shed light on underlying mechanisms of thought). It sounded like she was alluding to or wanted to allude to Marr’s three-level hierarchy of design. I’ve been meaning to start a thread on that, since that’s a very important topic for all designers of AI systems, and more clearly shows why artificial neural networks are not promising per se. I don’t know if she was unaware of that, or refrained from getting into such details for reasons of understandability.

It was a bit amusing to me that a British woman would be so concerned with machine understanding of Shakespeare: a topic that one would generally associate with the British. I’m not complaining, just surprised to see stereotypes prove so accurate. She’s quite correct about the importance of creativity, but that is something Einstein mentioned long ago, so that was obvious to me (though not obvious to most managers I’ve had!).

This suggested to me it might be interesting to discuss at some length one special ability, namely “creativity,” and its exultant culmination, genius.
  There is another reason for this choice. When talking to “hard” scientists, philosophers, or just high-IQ people who take an interest in psychology, about intelligence, they almost invariably tell me that IQ is not the measure of man, and that they would be far more interested in problems such as those presented by creativity and “genius”—the obvious implication being that science is incapable of dealing with such arcane matters. Now of course it is easy to reply that no psychologist ever suggested that IQ was the measure of man, and that Stephen J. Gould was just as wrong in entitling his book The Mismeasure of Man. IQ testing deals with one aspect of human behaviour, namely cognitive functioning; there are many other aspects of personality that are equally or more important. There is no one measure of man, and to suggest that there is is completely unrealistic. Newton studied the laws according to which bodies attract each other, but he never suggested that his formula was a measure of nature; there are many other laws.
...
  What is obvious is that geniuses have a high degree of intelligence, but not outrageously high—there are many accounts of people in the population with IQs as high who have not achieved anything like the status of genius. Indeed, they may have achieved very little; there are large numbers of Mensa members who are elected on the basis of an IQ test, but whose creative achievements are nil. High intelligence seems to be a necessary qualification for high creativity, but it does not seem a sufficient one.
  This is borne out by another study carried out in the early twenties by the same group of Californian psychologists Cox belonged to. This is a very famous follow-up study of high IQ children (135 and above) who were followed into adulthood to see what became of them as compared with children of ordinary intelligence. Terman, who originated those “Genetic Studies of Genius,” as he called them, selected 857 male and 671 female children on the basis of their high IQs; the mean was 151 for both sexes. Seventy-seven who were tested with the newly translated and standardized Binet test had IQs of 170 or higher—well at or above the level of Cox’s geniuses. What happened to these potential geniuses—did they revolutionize society? Did they monopolize the Nobel Prize? Did they prove outstanding in the arts? The answer in brief is that they did very well in terms of achievement, but none reached the Nobel Prize level, let alone that of genius. They did very well at University, many were cited in American Men of Science (including seven of the women!), the Directory of American Scholars, or Who’s Who in America. They published novels, volumes of poetry, technical, professional or scholarly books, plays, essays, and scientific papers. They took out many patents, made substantial contributions to the physical, biological and social sciences, and became leaders in university faculties.
(“Intelligence: A New Look”, Hans J. Eysenck, 1998, pages 115, 127)

 

 

 
  login or register to react