Poll on Coulter.

[Kuciwalker]

Originally posted by KrazyHorse


Look more carefully at the pattern that you've built

Some of those numbers are dividing each other...

In fact, if you've built it according to how I think you did, an awful lot of them are...

I forgot that I had a limited number of 3's and 5's to play with.

[KrazyHorse]

See above. Almost....

[Kuciwalker]

Didn't you forget (3*5)^100?

EDIT: I originally guessed it would be the 101st square, but I thought that was too easy and tried to work it out on paper some more

[KrazyHorse]

No, that's included in it...

In fact, it's included as the last possible combination

[KrazyHorse]

BTW, this was not a multiple choice test. Papers must have justification for their answers...

[Kuciwalker]

Oh, I see.

[KrazyHorse]

Not bad, by the way Kuci

Pattern-building is the most important analytical tool a mathematician possesses.

No proof is built as a computer would build it. Every theorem exists because somebody noticed a pattern in examples they were creating and tried to prove that the pattern was a general principle. Intuition first, logic second.

[KrazyHorse]

There's a reason we haven't built electronic mathematicians yet.

We just let them help us on stuff like the 4 colour theorem...

[Kuciwalker]

I once spent about 2 hours staring at a list I had put together of the first 100 primes, powers of 2, a bunch of other numbers, etc. looking for a pattern. I nearly drove myself nuts

[Kuciwalker]

Originally posted by KrazyHorse
There's a reason we haven't built electronic mathematicians yet.

We just let them help us on stuff like the 4 colour theorem...

We'll see after I study AI next year

[KrazyHorse]

Good luck. Mathematics (and physics) have an art to them, as well as a science...

We can build a computer than can beat the best chess players in the world.

I think it will be a while before we build a computer that can replace a great mathematician.

[Kuciwalker]

I'd rather have a computer that was good at playing civ

[Jaguar]

Originally posted by Kuciwalker
I once spent about 2 hours staring at a list I had put together of the first 100 primes, powers of 2, a bunch of other numbers, etc. looking for a pattern. I nearly drove myself nuts

I drove myself insane trying to prove that given a point and a line, there's only one parallel of the line that goes through the point.

It's unprovable.

[Kuciwalker]

Hyperbolic geometry

[KrazyHorse]

You and I both know that if they put their minds to it they could build one that would beat all of us.

The problem with civ (as compared with chess) is that the game is hopelessly complex. I think that this complexity, however, degrades the additional benefit of the unique best move. I think a computer's micromanagement skills would destroy the best human player's setup if programmed any way better than half-assed (i.e. what we have so far)