Computer Generation of Problems

[hyperpape]

From another thread

Anyway, most go problems are designed for human solution, that is, for depth first local search, even if the search takes the whole board. If bots were trained on such problems, they could very likely do better than humans at them. But such training would take many thousands, maybe millions of such problems, and they do not exist, nor do we know how to have computers construct them.

Does anyone know of efforts along these lines? It would obviously be a hard problem, but several ideas immediately jumped into my head when I read Bill's comment. Lord knows I don't need another project, and this would be a big one, but...

[HermanHiddema]

IIRC, GoTools was/is a program that can generate life and death problems.

[hyperpape]

Looks like you're right: http://lie.ac.brocku.ca/gotools/index.p ... t=features

[Bill Spight]

IIRC, GoTools was/is a program that can generate life and death problems.

Right. I was not clear. I had large semeai, ladder problems, etc., in mind. If you restrict the region of play enough, current programs are good enough. You might contact Thomas and see if he is interested in creating large problems.

[Gomoto]

The current generation of computers has a lot of problems.