reply to post by vjr1113
Yeah, I copy-pasted. Let me repost so that there is no confusion. I'll put the article in quotations:
Please allow me to link to a BBC story entitled "Can computers have true artificial intelligence?" My comments at the end:
"The Turing Test
The benchmark for the success of AI that Turing suggested in his original paper of 1950 was about communication.
If you were talking online with a person and a computer, could you distinguish which was the computer?
Since we can only assess the intelligence of our fellow humans by our interaction with them, if a computer can pass itself off as human, should we
then call it intelligent?
There are some very good candidates out there that are getting close to passing The Turing Test, including this one, cleverbot.
Interestingly this hurdle is more and more being regarded by those in the field of AI as a red herring.
Even if a computer passes the test, it does not mean it understands anything of the interaction.
In fact I was recently put through a thought experiment called The Chinese Room devised by philosopher John Searle, which challenges the idea that a
machine could ever think.
I was put in a room with an instruction manual which told me an appropriate response to any string of Chinese characters posted into the room.
Although I do not speak Mandarin, it was shown I could have a very convincing discussion with a Mandarin speaker without ever understanding a word of
The Chinese room problem Continue reading the main story
Chinese room problem A message is sent to a non-Mandarin speaker in "The Chinese Room"
Chinese room problem Armed with an instruction manual of possible questions and answers, the non-Mandarin speaker is able to match the characters and
select a response.
Chinese room problem The Mandarin speaker believes he has been talking to another Mandarin speaker. But the person in "The Chinese Room" has no idea
what he has said.
Continue reading the main story
previous slide next slide 1/3
Searle compared the man in "The Chinese Room" to a computer reading a bit of code. I didn't understand the Mandarin so how could a computer be said to
understand what it is programmed to do.
It's a powerful argument against the relevance of Turing's test. But then again, what is my mind doing when I'm articulating words now?
Aren't I just following a set of instructions? Could there still be a threshold beyond which we would have to regard the computer as understanding
partial quote end
I tend to believe that this problem can be resolved more easily through organic evolution than through human design. The Chinese Room problem is an
impossible problem if a machine is programmed. One can never tell the difference between an algorithm and genuine understanding, which is often gained
through experience on the tabula rasa of the mind.
However, if there is enough experience, and an organism is able to pass genes down, as well as culture, then an accumulated understanding emerges that
is an open-ended ability to adapt as opposed to an algorithm. You can think of it as a chaotic interplay of a number of algorithms, not as a single
algorithm. This defines intelligence in my book.
Now, if I am wrong, and we can create artificial life, then by definition we no longer have a computer. We have silicon based life. At that point, the
organisms created become their own organisms and not artificial by definition. We cannot then program what we have created. It becomes autonomous. Are
we ready for that?
end of old post
OK, so that is the original post with the article still in quotes. My comments come after that, so we don't confuse my ideas with the BBC.
As for the issue of Nature, please see my original post where I try to invoke randomness as the best reinforcer of intelligence and intuition.
edit on 16-4-2012 by EarthEvolves because: More needed