Essay - AI: The Technological Trickster

Jim Erkiletian's amazing essay has just been published.

Moving elegantly among social sciences, computational linguistics and classical AI disciplines, this article analyses the essence of computer thought compared to human reasoning, and emphasises its effects on society and future developments.

http://ai-depot.com/Essay/Trickster.html

Wow! What a great article! Well written, well-presented--I'm bookmarking this one!

In the 1980's, one of my brothers met one of my father's former
students from 1952. Some weeks later he mentioned it to dad,
but couldn't remember the guy's name, only that he'd been in
one of dad's classes at Missouri School of Mines (where he
taught engineering mathematics) some 30 years earlier. And he
thought the name was Morgenson.
Dad said he didn't have a student named Morganson, but there
was one named Morgernstern in '52.
My brother was impressed at this feat of memory, and said so.
Dad said he didn't have trouble with stuff from 30 years back, it's
the stuff from yesterday he forgets.
It's difficult to see how one would build a computer memory that
would duplicate such a condition.
Jim

In the 1980's, one of my brothers met one of my father's former
students from 1952. Some weeks later he mentioned it to dad,
but couldn't remember the guy's name, only that he'd been in
one of dad's classes at Missouri School of Mines (where he
taught engineering mathematics) some 30 years earlier. And he
thought the name was Morgenson.
Dad said he didn't have a student named Morganson, but there
was one named Morgernstern in '52.
My brother was impressed at this feat of memory, and said so.
Dad said he didn't have trouble with stuff from 30 years back, it's
the stuff from yesterday he forgets.
It's difficult to see how one would build a computer memory that
would duplicate such a condition.
Jim

A really superb article!

It makes me think language(-learning) is deciphered - apply the
thought of metaphors to ontologies, you'll get a system that relates
symbols (Words) to the context they got learnt in.

And now, based on the current context we can choose which meaning is
most probable according to the relation between the current context
and areas related to the word which is to be understood... and go on
learning relations between symbols and context - like the weights in
neural networks...?

Cheers, Kim

I love this article. It combines two difficult topics and relates them flawlessly. I would have to say this quite possibly one of the best articles I have read here.