--- BillK <[EMAIL PROTECTED]> wrote: > On 7/1/07, Tom McCabe wrote: > > > > --- BillK <[EMAIL PROTECTED]> wrote: > > > > > On 7/1/07, Tom McCabe wrote: > > > > > > > > These rules exist only in your head. They > aren't > > > > written down anywhere, and they will not be > > > > transferred via osmosis into the AGI. > > > > > > > > > > They *are* written down. > > > I just quoted from the FIDE laws of chess. > > > And they would be given to the AGI along with > the > > > rest of the rulebook. > > > > I just explained to you why the rules would be > > meaningless. > > > > > If the rules are meaningless, why did you give such > a silly example as > a chess-playing AI shooting the opponent?
I shouldn't have said "meaningless"; maybe a better word is "irrelevant". Why is a chess-playing AGI shooting the opponent "silly"? It's one of the most obvious ways to win at something- cheat. > That is similar to a football playing AI breaking > all the opponents > legs then scoring 2,000 goals. Yes, it is. This does not make either of them less realistic. AGIs do weird stuff. > If the AI follows the rules of each particular > competition then it > won't do such things. The rules of football, I hope, do not include any specific instructions against breaking people's legs. > Of course an AI can try to find loopholes in written > rules, which is > why humans put in 'catch-all' clauses when they > write instructions. Can you name a catch-all clause that works so well that it covers any possible bad thing the AGI might do to an opponent? > (Humans cheat as well, you know). > > Your chess example was silly. That's all I am > pointing out. Why is it silly? It may sound silly, but it seems to me to be a realistic way that a chess-playing AGI would act if given that kind of power. > > > Your other point that an AI is uncontrollable by > humans and can do > just as it pleases, is a major concern. Agreed. > But it is > unlikely that such > a creature would be wasting time playing chess with > humans and > shooting them. To an AGI, "wasting time" is anything that doesn't serve the supergoal. Shooting the chess opponent serves the supergoal of "winning at chess", therefore it is not wasting time according to its own internal time-wasting calculator. > You seem to enjoy argumentation, but please try to > reason sensibly. AGIs do not work in a "sensible" manner, because they have no constraints that will force them to stay within the bounds of behavior that a human would consider "sensible". > > BillK > > ----- > This list is sponsored by AGIRI: > http://www.agiri.org/email > To unsubscribe or change your options, please go to: > http://v2.listbox.com/member/?&; > - Tom ____________________________________________________________________________________ Food fight? Enjoy some healthy debate in the Yahoo! Answers Food & Drink Q&A. http://answers.yahoo.com/dir/?link=list&sid=396545367 ----- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=4007604&id_secret=8899106-6d43f7
