Brain algorithms

[evildaddy911]

Quote:

The thing is there is no way of knowing if humans are more efficient at chess than we are

does this mean your a computer?

[mirhagk]

Actually that is a very good question, is our brain a turing machine? Many people think that it is, although it has yet to be proven either way.

Also it is possible to rule out which algorithms our brain uses, as seen by this study. This could easily be done with chess, by comparing different board positions to the time it takes for both a computer and a human to make the optimal move (the best way to ensure that the optimal move is always selected is to have the board always be one move away from a checkmate). If the time taken by the human is directly proportional to the time taken by the computer, then we use the same algorithm, otherwise (assuming sufficient trials) we can safely assume that is not the algorithm our brain uses. We can actually even take that one step further and compute the relationship between the time taken on various boards, giving us a rough estimate on the time complexity of human's algorithms. If the time complexity is less than a computers (which I very much suspect is true) then our brain is more efficient at chess than a computers.

[evildaddy911]

of course, we would need to factor in the speed of that computer versus others, my 2001 machine would take much longer than a brand new top-of-the-line desktop

also, we would need to test a few different algorithms after to determine which one it is
for example:
algorithm | computer | situation
=======================
brute force | fast | #1
brute force | fast | #2
*
*
brute force | slow | #1
*
*
heuristic #1 | fast | #1
*
*
heuristic #1 | slow | #1

you get the picture

[mirhagk]

You actually don't need to factor in the speed of the machine so long as you get a single consistent machine to compare it against. Efficiency has nothing to do with the speed of the machine, it has to do with the time complexity http://en.wikipedia.org/wiki/Time_complexity of the algorithm. So long as the computer's clocked speed is done with multiple trials, and is the same every time, you can easily test with a single computer.

[evildaddy911]

what i meant is that the brain will probably run faster than a slow computer, but not as fast as a really fast one

[mirhagk]

Yeah, and that has nothing to do with complexity, difficulty or efficiency. Just because you have a better computer doesn't make your brute force programs better programs than someone else's efficient algorithms, despite how much faster yours run.