From: Samantha Atkins (samantha@objectent.com)
Date: Wed Sep 12 2001 - 04:11:01 MDT
Miriam English wrote:
>
> At 01:31 PM 11/09/2001 -0700, Holden, Steve wrote:
> >Per the 'manual override' comment in an earlier post:
> >My father flies 757 and 767 airplanes for American Airlines, just like the
> >ones hijacked today. It's not possible to disable manual override and put
> >them on a collision course.
>
> This is a good argument for having AIs able to override dangerous
> situations. I know pilots would not be happy about it, but making planes
> "smart" and able to avoid such dangers on their own would make such nasty
> acts *way* more difficult.
>
Not unless the AI is fully sentient (and therefore makes its own
decisions we may or may not enjoy) and its security and the
security of all communications are utterly assured. That is a
big set of ifs. Scratch one tecnical fix for the immediate
future.
- s
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:10:31 MST