Artificial Intelligence

I've been discussing this with Mike recently (as he has an MSc in the subject he seems like an appropriate person to involve), and he's corrected some of my misconceptions about AI as well as giving me a lot of food for thought.

I once said to him, "When do you think that AI will be better than people?" to which he replied, "Better at what?". I thought about it for a moment and said "Better at thinking like people." and he returned the ever-so obvious "Never, by definition people are as good at thinking like people as it is possible to be." Which left me stumped at the time and still fairly stumped now.

The main problem with artifical intelligence is knowing when you've achieved it. People want to achieve "Artificial Intelligence" but there's no real concensus on what is meant by "intelligence". You end up with one of two situations, either you have a concrete definition of intelligence, leading to people complaining that "That's not what I mean by intelligence." or you can go by the more inductive "felt" definition of intelligence (which leads to thinks like the Turing test, which is basically a test of how easily fooled we are). Either way round, it's going to be very hard to get people to agree what intelligence is, let alone artificial intelligence, let alone whether we've achieved it.

Also, there's a fairly sizable proportion of people who just don't want to believe that intelligence can be an aritificially created. This means that whatever is produced cannot possibly be 'real intelligence'. It was once claimed that an a chess-playing computer would be intelligent. Now that we have one that can defeat grand-masters they say that chess doesn't require intelligence. This kind of thing also occured with flight. Even after the Wright Brothers' flight at Kitty Hawk there were people claiming that it wasn't "real flight". Now, it could easily be argued that chess playing computers aren't thinking like people when they play chess, but does that really mean they aren't intelligent at all in any way?

So, my prediction for artificial intelligence is that we will continue not having any artificial intelligence until at some point we realise that we've had it for some time, we just hadn't really thought of it like that (i.e. we'll be able to see the path we walked to get there once we've walked there).

This is not me

From Dive Into Mark:

In sixth grade, we got an Apple //e computer. Actually, my father got it so he could type up the book he was writing for his sabbatical. But I found out that by programming it, I could make it do whatever I wanted. Or at least whatever I could break down into simple enough steps that the computer could understand. This had four effects on my personality.

The computer replaced my social life. This went more or less unnoticed.
Because I was constantly programming, I learned to break down everything I did into simple, logical steps, as if I were writing a computer program. I very rarely now get overwhelmed by the amount of work I have to do, because I never look at all of it at once. I take this bit and finish it, then that bit, then the next.
75% of my creative juices went down the drain forever.
I discovered that I was in complete control while I was in front of the computer, and when I did have to interact with people, I wanted that same control. I am not an outgoing person, but I am dominating.
Reactions to this now, 12 years later:

It shows a surprising level of self-awareness not normally found in teenagers.
My sense of humor was etched in stone by then, and it has not changed a bit since.
The rest of my personality has also changed very little. I have mellowed (slightly) in the past 12 years, but not actually changed much.
I would now disagree with the creative-juices-down-the-drain part. I didn't realize at the time how creative programming could be.