perm filename MESSAG.SCI[LET,JMC] blob
sn#084445 filedate 1974-01-25 generic text, type T, neo UTF8
∂25-JAN-74 1553 MT,YAW
I saw what the point was I wanted to make about "semi-persons" the other day,
and the question of where to put machines on the responsibility spectrum
FALLING TREES---TIGERS---DOGS---PEOPLE.Given,a syou said,that theyre all bits of
material stuff,nevertheless (British?) common law draws some line on that
scale between dogs and tigers,because although we reliably know the plans
of a falling tree(too fall!) and of tigers(to get out and eat and kill) we
arent so sure about dogs.So the doings of the tree on your land and your escaped tiger are
all your fault,though the doings of your dog arent,and the doings of your children are hardly your fault at all.Im making this point in con with yor general
that we cant talk of responsibility of machines AT ALL until we can talk of their
planning abilities.Im trying to say ,I agree if that means their actual behavior
interpreted by judges as evidence of plans,but I disagree if it means
structural (print out) evidence of their plans,simply because the law
distinguishes between dogs and tigers without knowing anything at all
about the differeing physiology of dog and tiger brains.I guess you might say ,
at this juncture,that this shows no more than the low technological
level of the law, and that it'll have to be beefed up before any machines can be
placed on that spectrum,other than at the "falling tree" point.
[substitution in above for readability:
"con in yor general" → "connexion with your general one"]