--- On Mon, 11/17/08, Mark Waser <[EMAIL PROTECTED]> wrote: > > Autobliss responds to pain by changing its behavior to > make it less likely. Please explain how this is different > from human suffering. And don't tell me its because one > is human and the other is a simple program, because... > > Why don't you resend the link to this new autobliss > that "responds to pain by changing its behavior to make > it less likely" and clearly explain why what you refer > to as "pain" for autobliss isn't just some > ungrounded label that has absolutely nothing to do with pain > in any real sense of the word. As far as I have seen, your > autobliss argument is akin to claiming that a rock feels > pain and runs away to avoid pain when I kick it > > > So either pain is real to both, or to neither, or > there is some other criteria which you haven't > specified, in which case I would like to know what that is. > > Absolutely. Pain is real for both.
autobliss: http://www.mattmahoney.net/autobliss.txt By "pain" I mean any signal that has the effect of negative reinforcement, such that a system that learns will modify its behavior to reduce the expected accumulated sum of the signal according to its model. In the AIXI model, pain is the negative of the reward signal. Kicking a rock or cutting down a tree does not inflict pain because rocks and trees don't learn. -- Matt Mahoney, [EMAIL PROTECTED] ------------------------------------------- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244&id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
