Reply to thread

Message: <blockquote data-quote="clearstream" data-source="post: 8631389" data-attributes="member: 71699">I really do not mean to be vexing, the above explains that you believe it can't work, but does not show that it can't work.How will he survive? How will he react? Play to find out! (No one knows how it will go until everyone has had their say. It's not collective storytelling because we are roleplaying and we are gaming. We can have a host of regulatory and constitutive rules... with the exception that instead of rolling dice, someone decides.)Maybe another analogy is Roshambo. The NY Post or maybe it was the Washington Post showed that humans are bad at Roshambo using an AI opponent driven by some kind of Markovian algorithm. The way one responds to this is to partake only of N-player Roshambo, where N is greater than two (you have to extend Roshambo to make this work). A different form of learning AI can still beat that, but for human purposes in small samples (like the sample of every roll/decision your group will make in a year's play) it's fair to call it not predictable (even though each participant can perfectly well predict - in fact decides - what they will choose).</blockquote>