It's certainly not and was never intended to be a good example of intelligence, but the fact that it fooled many people says more about the people of the time than anything else.I haven't read the whole paper - got to the exchange with 'Eliza':
User: Men are all alike.
Eliza: In what way?
User: They’re always bugging us about something or other.
Eliza: Can you think of a specific example?
User: Well, my boyfriend made me come here.
Eliza: Your boyfriend made you come here.
User: He says I’m depressed much of the time.
Eliza: I am sorry to hear you are depressed.
User: It’s true. I am unhappy.
Eliza: Do you think coming here will help you not to be unhappy?
User: I need some help, that much seems certain.
Eliza: What would it mean to you if you got some help?
User: Perhaps I could learn to get along with my mother.
Eliza: Tell me more about your family.
If this is a good example of intelligence and it actually fooled some of the participants, we are in trouble. I've had conversations like this and the quality of the replies indicated distraction - talking on the phone while checking emails, for example. Women are particularly good at this and have absolute faith in their ability to 'multitask'.
Weizenbaum tells the story in his book "Computer Power and Human Reason: From Judgement To Calculation" (I first read it in college, just a couple years after its publication). Some people aren't good judges of what such a program is doing for various reasons - apparently they want to believe it is understanding them, and its responses give enough of a veneer of "understanding" for them to latch onto and fully believe there was "something there listening to them. His own secretary was this way, and he had had her help in the programming and/or collection of data used in the program, and this made him even more surprised at her reaction.
As the article states, this was 1965, when most people had no clue what a computer might or might not be capable of. People thought when they saw a tape drive they were seeing a computer.
I've heard similar things about the invention of the phonograph - when people heard it for the first time they couldn't believe it was what is was, a machine reproducing the sounds of the human voice. They thought Edison, or whoever was recorded on it, was somehow "throwing his voice" like a ventriloquist. But just about everyone in the last century would easily recognize the sound as a highly distorted or "worn out phonograph" as we have heard such things before and have a context for them, despite how "obvious" it may seem to us.
IBM's "Watson" that just won Jeopardy, on the other hand, uses perhaps ten orders of magnitude more computer resources that Eliza. Eliza can easily run on an Apple ][. I saw the Nova episode on Watson, it's quite interesting - just seeing the thing correctly answering a question, as I had seen on earlier video clips, seemed almost magical, but the Nova episode explains enough of its operation that it now seems to me more like "just a trick." It certainly has no more real understanding than Eliza:
Smartest Machine on Earth | NOVA | PBS Video
What's fascinating about Watson is the way that it can analyse a plain language question and interface that with a stored database. Imagine what that could lead to - a search engine that actually finds what you want rather than giving you tens of pages of vaguely relevant results, or a Helpdesk that actually understands your angry ravings and gives you an appropriate solution to your problem.
I recall a small article I read some 20-25 years ago in an American magazine.
The article described the “teaching” procedure of an experimental computer program.
People working for the the project, were feeding-in data to the “memory” of the computer for many years. This data consisted of text paragraphs from newspapers, magazines, periodicals, books, encyclopaedias e.c.t., picked up in random from every principle.
One day, they input a single phrase: “Napoleon is dead.”
The computer responded: “And Nelson felt sorry” (or something like that, my memory fades)
Regards
George
In the nova program, there was discusison of another attempt at creating a "thinking" machine that was not so successful. It reminded me of an earlier program that had people feeding information to a machine and it would output correlations that it "found" in the data. At the time, some of the correlations were wrong, and even humorous, but others were actually insightful. I wonder if that is the same one you are talking about?
All tech-help call centers use simple computerized flowcharts to diagnose the problem. All such flowcharts inevitably lead to "Reformat the hard drive and reinstall the operating system."
There must be a higher level flow chart that makes sure that the more technically capable callers get assigned to the least capable advisers with the thickest accents!
What would be really interesting is if the judges themselves were evaluated ito their ability to discern what they're talking to.
I'll swear some people I've known couldn't spot the difference if the contestant was a broken keyboard, and would probably be thinking of marrying it after 5 minutes because it's such a "good listener".
I do the "good listener" thing quite well myself, mostly because I can't be bothered to talk when I don't have anything to say, and have sometimes been quite astonished at peoples misperceptions and misinterpretations of silence.
For example, I have received compliments like "You're so interesting to talk to!" and "What I really appreciate about you is your honesty", after half-hour conversations during which I have uttered absolutely nothing except acknowledgment noises.
I'm not surprised Eliza did so well.
I'll swear some people I've known couldn't spot the difference if the contestant was a broken keyboard, and would probably be thinking of marrying it after 5 minutes because it's such a "good listener".
I do the "good listener" thing quite well myself, mostly because I can't be bothered to talk when I don't have anything to say, and have sometimes been quite astonished at peoples misperceptions and misinterpretations of silence.
For example, I have received compliments like "You're so interesting to talk to!" and "What I really appreciate about you is your honesty", after half-hour conversations during which I have uttered absolutely nothing except acknowledgment noises.
I'm not surprised Eliza did so well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.