Conformity in duplicated personalities

Steve Witham (
Thu, 15 May 1997 21:45:51 -0400

First off I should say that Robin's point about humans holding onto the
niche for generality for long enough, is interesting and I have nothing
to say but hmm.

Damien Broderick writes-

>Once the knowledge base is compiled,
>compressed and debugged, AIs can (presumably) be die-stamped. That might
>lead to dangerous conformity, [...]

Interesting scenario: a mass-produced personality. Everyone would get
to know it intimately (unlike the way everyone knows public personalities
superficially). They'd know its flaws, share tricks for dealing with it...
and meanwhile each of the instances would have the same blind spots to
the same faults. In fact you could get a mass version of that two-person
dynamic where character traits and conflicts get exaggerated.

But you could also have copies that broke the mold in one way or another,
which would provide an interesting kind of example of the potential of
individuality and transcending limitations.


--           Steve Witham          web page under reconsideration