RE: [Asrg] 2. Problem Characterization - Defining spam within consent paradigm

["Bob Wyman" <[e-mail]>]

Barry Shein wrote:
> Although reasonable-sounding these descriptions [of consent] 
> are starting to remind me of those long-ago Artificial Intelligence 
> seminars where you'd get to ponder why some seemingly simple 
> human behavior was so difficult to code into a program. 
	We've had precisely the same issue with software licensing or
Digital Rights Management for years. The issue is one of managing
expectations. The reality is that machines can only manage things they
can detect yet people will often express their rules in terms that
involve things that machines can't detect. For instance, people will
often say that they only want mail from some specific user yet the
machine can't detect identity. The machine can only detect the use of
things which are assumed to indicate identity. (i.e. When if someone
says "Only Bob can send mail" what the machine needs to hear is "Only
someone using Bob's private key can send mail." For a person, "identity"
is an attribute of being. For the computer, identity is a property which
is assumed belong to users of certain proofs of identity. 
	But, this is all getting a bit philosophical. Barry has a point.
As we get into dealing with the issues of expressing and managing
consent, we must all be careful to ensure that people understand the
inherent limitations of machines in this realm. There is only so much
that the machine can do for you.

		bob wyman


_______________________________________________
Asrg mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/asrg