Re: Single model or framework for reputation/trust

Ed Gerck <[email protected]> wrote:
> >   I'm sorry to disagree with you and Bertrand Russell, but the story
> > is based on misconceptions, as I pointed out previously.  Was the
> > chicken wrong for 6 months, when it predicted that "today, the farmer
> > will feed me?"  No.
> 
> The story illustrates the leap of faith that occurs when reputation
> is used to predict behavior. It's simply not warranted.

  Cut the crap, Ed.  You *deleted* my qualification that, in this
case, reputation was correct 99.5% of the time.  You then talk about a
"leap of faith", as though it was relevant to the discussion.

  To repeat my point again, which you seem to have misunderstood:

	Reputation is a strong indicator of future behavior.

  It's not a PERFECT indication of future behavior.  To claim so would
be idiotic.  I'm NOT claiming it's perfect, so a leap of faith isn't
required.  So top talking about leaps of faith, it's annoying.

> The chicken should never have predicted anything based on reputation
> alone.

  And again, you're talking as if I'm trying to use reputation
"alone".  I'm not.  I never have, and I never will.

  This is a problem I've seen with many discussions.  There's talk
back and forth, but little communication.  e.g.

    Party One: Tool X is useful for solving problem Y.

    Party Two: You're forbidding anyone from using anything OTHER
	       than tool X to solve problem Y!
	       You're claiming that tool X is perfect!
	       You're claiming tool X is the best way to solve problem
	       Y!

    Party One: WTF?

> You can't rely on reputation to predict behavior, as Russel's chicken
> story exemplifies. To do so would be a leap of faith.

  Only if you're an idiot, and think that reputation is perfect.

  But 99.5% correct?  That's useful.

  And wasn't Betrand Russsel the guy who wrote a ~400 page document
trying to prove "1+1=2"?  My recollection is that he failed, so I
can only conclude that he believes it as a "leap of faith".  So a
"leap of faith" isn't a very good argument against anything.

> To rely on means to trust. To be able to trust future beahvior, we
> need more than just past behavior. We need more than reputation.

  Did I ever claim otherwise?

  Reputation is a tool which can be used by an email recipient to
determine how to deal with a message.  It would be stupid to think
reputation is perfect, and it would be stupid to rely on reputation
alone, to the exclusion of everything else.

  Now that that's clear (I hope), can we agree that reputation MAY BE
used as ONE PART of the information available, for input to a
recipients decision-making process?

  And given that (potential) agreement about reputation, I believe
it's useful to have on-line reputation systems.

  Alan DeKok.