My OperaArtificial Intelligence
Friday, May 26, 2006 10:27:16 AM
When I looked into AI a generation ago, I needed some practical result to convince my friends that AI belongs to the near and far future. I also asked myself: How can I get AI programs onto a computer near me? The answer was expert systems. One example is a program that can diagnose illnesses, but it did not convince my surroundings. Knowledge engineers interviewed experts. The program was given all the latest knowledge in medicine, but was not clever in using data on doctors, hospital, and the patient herself. A computer program alone can not treat sick person, but depends on its user having common sense.
Go to AAAI (American Association for Artificial Intelligence) home page and se how large the field of research topics is today. You still find HAL, games, and robots, but I guess also quite a few topics you did not think of. You can also find out more about AI by joining news groups.
Many fear that AI will make human beings “redundant”. But I wonder why doctors still relying mainly on their memory when confronted with a new case? Why can not cars park themselves, and why are library books unable to recommend other relevant information…
This page links to 868 pages around the web with information on AI.
You must be 
Leif Rømckeleirom # Friday, June 2, 2006 10:51:42 AM
Artificial Intelligence to Boost Space-Probe Efficiency
http://news.nationalgeographic.com/news/2006/05/060530-software.html
AI will help unmanned space probes to send better data back to Earth. Even current Mars rovers will benefit from smarter software. After upgrading they will be able to make decisions on their own.
Earth Observing-1 captures high-resolution images of events, such as volcanic eruptions, floods, and the break up of large polar ice sheets. EO-1 has already increased its scientific contribution a hundredfold, while cutting costs by more than a million U.S. dollars a year.
A little older articles:
Robot "Scientist" Said to Equal Humans at Some Tasks
http://news.nationalgeographic.com/news/2004/01/0114_040114_robot.html
Supercomputing Project Aims to Simulate Human Brain
http://news.nationalgeographic.com/news/2005/07/0720_050720_bluebrain.html
zeroGzerog # Friday, June 2, 2006 10:30:38 PM
you got another very interesting post (but I assume you know that).
I don't keep an eye on AI but from what I've heard I can say that it is a field that was greatly harmed by its own prestige. I mean that people did expect computer-based systems which would exhibit intelligence comparable to humans. Movies and science fiction were mainly responsible for those big expectations. So, when all someone gets is partly intelligent systems it seems like a failure.
Yet, creating a machine that thinks and understands its surrounding world is a way too much complicated project (and some argue that it may be impossible). Computers may execute billions of calculations per second, yet their overall computing performance can't be compared to a human mind which processes data in parallel and in ways that we don't know. Perception and reasoning, also, are far more complicated and fuzzy than the capabilities of machines that act on pure mathematical logic.
I think, though, that what has been achieved so far is very importand and is just a part of what is to be seen in the future. For example cars can't park themselves (yet) but they can adjust their fuel consumption very effectively which turns out to be very important.
Leif Rømckeleirom # Saturday, June 10, 2006 3:13:21 AM
I agree with you, the term AI is a bit presumptuous. It has been difficult to agree on how to define Intelligence. Great progress has been made in the research of the human brain, but we still have a long way to go. We even need to learn more about the process of learning. So even when the cars park themselves, AI up till then will be at let down.
I also agree with you: maybe it is in principle impossible for AI to reach human intelligence level. On the other hand when you look at the branches of AI, and the progress made in researching for example pattern recognition, search, learning from experience, planning, and expert systems, then you tend to be optimistic.
leirom
AIMarket # Saturday, September 9, 2006 1:53:57 PM
http://aima.cs.berkeley.edu/ai.html
Leif Rømckeleirom # Sunday, September 10, 2006 10:06:52 AM
When I followed “your” link, I read: “This page links to 849 pages around the web…”, and at the bottom of the page “Modified: Jan 31, 2003”, while “my” link says Changed 09/27/04. Both not very often updated…
I’m not ungrateful, only a little confused by the web pages themselves. Any way, the web sites are both rather formidable. There should be reading material here for years of studying.
Again salamat, and thanks for showing an interest for my blog.
AIMarket # Sunday, September 10, 2006 1:34:53 PM
The link that I just gave I know was update only last week, since it has a link to my web site on it now. Bad directed links would have been deleted from it.
But you are right it does say at the bottom of the page, it hasn't been updated since 09/27/04 ??? Not sure why. I’m not going to bother them about it, web maintenance may not be a high priority.
I think the old Berkeley site isn’t maintained... just guessing but it might have changed academics hands. It doesn't really matter, I think the old site will be redirected to the new one sooner or latter, but they are both handy link pages.