Don Dailey wrote: > We do not have a wall, we have a mountain. Whatever. I'm looking for the ladder to help us up this wall (or the sledgehammer to knock it down :-), you're looking for ropes to help you up the mountain.
> Darren mentioned that I would say the algorithm is scalable. Apparently > this is a reference to Moores law. ... No. Just that giving MCTS algorithms more computer power makes them stronger. I'm referring to your scaling study. But, that was between MCTS programs, except for 10-kyu gnugo, and it was on 9x9. I think if you had been able to include some human dan-level players into the playing group, and done it on 19x19, the results would have shown scaling breaks down; it will turn from looking like a straight line to looking asymptotic at around shodan level. Just *looking* asymptotic: when you repeat your experiment in 200 years with your 70 trillion times faster computer I'm sure you'll see it is getting stuck at 6-dan instead. But, for our practical purposes, I think we'd be better off discussing the "ladder" to get us over the wall, than putting ourselves into suspended animation. >> And David Fotland commented [1] with effectively the same thing: current >> MCTS programs won't be European 1-dan no matter how much computer power >> they have. (Instead algorithm issues need to be fixed.) > David Fotland may have said that, but he didn't mean that. If I gave David > Fotland a computer that was 73,786,976,294,838,206,464 times faster... It was in the context of a bet where the computer needs to be at European 2-dan level (or stronger) in 18 months time. David is saying "won't be European 1-dan no matter how much computer power they have (given the amount of computer power you could possibly manage to take to an event site, even assuming Moore's Law holds for another 18 months)" Thankfully he elided that, so as not to bore us :-). Darren >> [1]: >> http://answers.polldaddy.com/poll/1614035/?view=results -- Darren Cook, Software Researcher/Developer http://dcook.org/gobet/ (Shodan Go Bet - who will win?) http://dcook.org/mlsn/ (Multilingual open source semantic network) http://dcook.org/work/ (About me and my work) http://dcook.org/blogs.html (My blogs and articles) _______________________________________________ computer-go mailing list computer-go@computer-go.org http://www.computer-go.org/mailman/listinfo/computer-go/
