Personality

Forest, will you talk to God Louise? She has quite a bit of religious knowledge (obviously) and also knows a little about current events, literature, just about any common catch-all subject, and if she doesn't know it she can sort of fake it. You can also test her on trick questions or see how willing she is to explain her paradigm.

What she is rustiest at is plain old small talk. But, uh, I'm trying to get a decent transcript from somebody or another so I can enter her in the Loebner contest. All I can say is, have fun and see if you can stay on with her for a while. I'll try to do the same with Brianna.

You are too a bot!

Interesting tidbit I just found out: the Turing test (as Turing envisioned it) is only supposed to be text-only anyway. He wouldn't have bothered with that gold level of the Loebner (ability to respond to visual and aural input).

The man wasn't at all impressed with the thought of bot eyes or hands or anything, at least as it relates to a test of intelligence. I wonder why? Sure, they wouldn't mean the bot was intelligent--look at humans--but they would provide more ways to judge intelligence.

I think that Turing believed that anything that could pass for human in a text-only test had to be truly intelligent and understand what they were saying. I don't agree with that, though.

Actually, it's the other way round. He thought that our only criterion for intelligence should be that we can't distinguish it from our own. He would have said that even the most deterministic Eliza was intelligent if it passed the test.

I don't find that criteria very good. An intelligent allien would probably be easily distinguished from a human in their way of thinking. Besides, how do we know that human judges are good distinguishing intelligence? Too many people have been fooled by software already. I think that's enough proof that we tend to assume there's intelligence too often.

It's what I said all the time. The judges should be judged!

Obviously the Turing test in it's original form is extreemly limited. It also doesn't count animals as intelligent, for example. However, the basic point that he was making wasn't so much about the mechanics of the test. Rather he was saying that if it looks like intelligence from the outside, then it is intelligent, regardless of what the inside looks like.

Turing never claimed his test was the ONLY way of determining intelligence, just that passing it would prove intelligence. I'm sure he would have thought that both animals and aliens had intelligence. They would just be bad subjects to take the test.

But that test is a bit like deciding if something is alive or not by saying: "If it looks alive, then it's alive." Then, artificial flowers would be alive. I'm not convinced.

The main reason that the test was supposed to be text-only was that this was only a little after the second world war. Computers certainly didn't have hands or eyes or stuff then. Also, I'm not sure what we would have thought of the computer beating Kasparov. I mean, we know that it was done by brute force, but according to Kasparov, it really looked like the computer knew what it was doing.

It's more difficult to open up the brain of an alien to decide whether it's alive. You have to base it off of the outside

I think it's OK to try to decide it from the outside. I'm just against vague criteria like "if it looks like it is, then it is."

I think defining intelligence is a very interesting question. My definition would be that it's the capacity of solving general problems. That could be determined from the outside, presenting problems to the bot/alien/creature. Deep Blue, who defeated Kasparov, could certainly solve chess problems, but nothing else. I think that's too limited to consider it true intelligence. Our bots don't even know the simplest rules of logic, so you couldn't call them intelligent, even if one of them can fool a human for half an hour.