robot's constitution

8.10: pol/robotics 101/constitution not assured:

. the essential problem with robots

is that there is no built-in god-constitution,

with humans there is always the evil of

the supernatural's influence on our emotions,

but that evil always has the welfare of the future in mind,

while a war among robots could lead to a world devoid of life .

. an omnipotent God will make sure that our collective programming

never results in robots devaluing life .

[12.21: this may happen by

bringing us very close to the edge of extinction;

so, we should put much energy into replacing this strategy

with one that is more deliberate and less destructive .

]

. there's going to be a dilemma of one life form vs another,

mutually assuring the destruction of the other:

eg, the matrix-bound humans vs the free humans .

8.25: pol/robotics 101/real he'll begins in the matrix:

. one danger of matrix is no way to kill self,

if dual developers of matrix are at war,

they may use computer hacks to make he'll for the other side

with no death as escape .

9.28: adds/robot constitution/the hal lesson:

. when using robotic police, we need to be careful about

making sure that the constitution is entirely explicit

and that we consider how our values might change

and to define when it's ok to for a subset of us

to say that the values are changed

and it's ok to amend the constitution .

9.29:

. a perfect constitution should never need mending;

rather, it should say what our values are now,

the value space,

the root values are to reach consensus,

or to arbitrate by majority when that's not possible .

. get it all out there on the table .

. make sure the robot is negotiable for the local needs

not just holding the values of the programmer .

12.4: pol/robot constitution/illusion of globality:

. robot must completely separate worlds

so there is noone feeling obliged to

relig'ly convert the infidels nextdoor .