Has artificial intelligence ever been successfully developed and applied ?

Answer 1

That depends what you mean by artificial intelligence. If you mean for it to imply "as smart and creative as a human being" then we are still pretty far from it.
If you mean "able to best a human at a very specific task" then yes, the best computer program can beat the human chess world champion, and there are many other cases where computers will do better than humans.
One has to remember that "artificial intelligence" very often reduces to the intelligence behind the program, which is to say the team of programmers who wrote the code and built the computer.
For now, there is a open challenge to build a system that will pass the Turing test which is, as clumsily as it might be, still the only proposed gage to decide if a computer is "smart" enough (see link).

Answer 2

Quoted from http://www.aaai.org/AITopics/html/aieffect.html

Researchers joked that AI stood for 'almost implemented'. AI research projects that worked well enough to make it on to the market, such as speech recognition, language translation, decision-support software, landing planes, reading hand-written postcards to sort mail, are suddenly no longer AI.
[/quote]

People often confuse AI with AH (artificially human). Games are a great example of the difference. A computer opponent that is intelligent will be able to move from point A to point B and go around obstacles. But if it always does it "intelligently" then it would be a straight line which the gamers would take advantage of quickly. If the game opponent is more human then it would vary it with flanking movements or attack from behind.

Game programmers consider it to be a funny comment on humanity that to make an AI more human, you remove some of the intelligence and add more random. (which does actually tend to win more often)

Answer 3

That would be No and No.

Artificial intelligence, if it wants to live up to its own claims, has a long way to go and they are missing at least a dozen important breakthroughs.

What is usually called AI are not much more than classifier systems with a semi-intelligent assembly of responses in the back-end. You get more "intelligence" out of some birds and apes than that, not to mention your average three year old.

IMHO, there has not been a single demonstration of an actually working AI system.

Answer 4

Depends on how you define AI - google for Turin test - that mimicks human conversation (as opposed to seeking out and blowing up targets in second and third world countries)