[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

starship-design: Re: unmanned missions



answering david, and paul, in that order.
David,
>Well, I think we're far from manually dexterous enough robots to 
>perform
>Galileo spacecraft.  So much trouble from one stuck antennae that 
>would
>take a guy with a wrench about five minutes to fix.
----------------------------------------------------
The galileo spacecraft was not provided with any mechanisms to repair
itself. On a large starship, small, versatile automatons for repair
would be provided, complete with proper tools and parts for the job. 
The computers for control of the spacecraft systems could be made highly
redundant, with each computer having primary responsiblity for one task,
and secondary or supervisory control over several others.  
************
Paul,
"Aggresion is a tough engineering problem"  -S. Pinker
A common assumption is that any AI will think "like a human" and
therefore be subject to human mental troubles.  This is theoretically
possible, but almost completely useless for practical applications. 
Even a sentient computer would still lack emotions, concepts like
"dignity" and other human traits which we ascribe to the HALs of the
future.  Malfun- fun- functions :) in a thinking machine would NOT lead
to antisocial behavior, but rather machine paralysis (which is why
seperate, highly redundant systems are in order).  Mental illness in
humans is mostly due to chemical imbalances, environmental pressures,
structural defects, or combinations of these.  "Electrical" problems
result in seizures.  Humans exhibit aggressive behavior because we are,
for most of our existence, omnivorous primates living in small bands
with fluid social structures, and our evolution favored a certain amount
of aggresion toward both each other and other species. Computers are
different, being programmed by (supposedly!) intelligent beings for a
specific purpose.  Moral here: don't anthorpomorhize.
Nels Lindberg