On Wed, Oct 22, 2008 at 7:16 PM, Pavitra <[EMAIL PROTECTED]> wrote: > On Wednesday 22 October 2008 07:46:51 pm Ian Kelly wrote: >> On Wed, Oct 22, 2008 at 6:39 PM, Pavitra > <[EMAIL PROTECTED]> wrote: >> > Proto-proposal / kibbutz: Instead of having Bayes make proposals >> > till it finds one it will vote for (>= 50%), have it make >> > proposals for a fixed amount of real time and pick the one it >> > likes most strongly. This should result in more Bayes-confident >> > proposals (given a long enough fixed time), which will hopefully >> > also be more >> > human-acceptable. >> >> I agree that it should spend a large amount of real time coming up >> with a proposal. Like, say, a month. > > I don't think it's reasonable to ask its maintainers to commit a large > percentage of their CPU resources to this when the same resources > could instead be spent finding potatoes and aliens.
Or folding proteins, which is more directly useful than either of those other two pursuits. ;-) -root
