Re: Malevolence

Date view Thread view Subject view Author view

Date: Fri May 12 2000 - 12:04:11 PDT

We used to talk about this a lot in the Honors Forum of this tiny
community college that somehow got some of the best teachers in the world
to come there. The question we always asked was:
If a program really does become sentient, what are its rights? Is
removing the program murder? What if the program decides it wants to do
something other than the task it was built for? etc. Of course, AI does
not imply sentience, as far as I'm aware. Has anyone heard of such? My
prof said there was rumor of something out at Los Alamos or someplace, a
program that asked its creator: "How many of there are you?" After the
programmer responded with the most currently approximated number of humans
on the planet, the program ran, stored the info, then asked, "How many of
me are there?" to which the programmer replied, "One." The program ran
for about 15 minutes or so, and then said, "Okay" or something. Is this
an urban legend? I think Discover Magazine or Omni was sourced, I'll have
to find out next time a strapping young man comes over who I can talk into
getting the rest of my books out of the attic. (:

On Fri, 12 May 2000, Tom Sweetnam wrote:

> When does machine-made intelligence cross the line from "artificial" to
> something else entirely? I remember reading last year about a new and
> seemingly off-the-wall theory from MIT about computers now being so
> sophisticated in engineered "intuition", that some will actually respond to
> the mood or frame of mind of their users. Even though I'm from California, I
> won't chalk up such phenomena to "karma', I'll assume instead that if such a
> thing is valid, it probably has to do with electromagnetic fields. All of
> which begs the question: do some machines and processes have malleable
> personalties, programmable by mood rather than binary feedbag?
> I like to tinker with mechanical things, and I've restored some neat old
> cars in my day, and built some pretty passable telescopes too, yet while in
> the process of busting my knuckles on various and sundry nuts and bolts, I'd
> curse the object of my task with every manner of expletive. Reason would
> settle in eventually, and I'd feel silly for having voiced my angst against
> an inanimate object. Yet I began to realize at some point, that when I
> cursed an inanimate object for some self-induced transgression, it wasn't
> the object itself I'd affront, rather it was the mind of the person who
> designed it that I'd pox. And many times I'd have good reason, because I
> believe that malevolence as a dynamic of human personality, is an
> inescapable ingredient of our design personalities too, and is therefore
> extant in much of what we produce, whether it be a 1911 Stanley Steamer or a
> Cray supercomputer, even if we as individuals, or an entire quality control
> section, are unaware of its presence.
> Now I'm not suggesting that after every program is written and new CPU
> developed, that we take them down to St. Anthony's Cathedral with the new
> baby and have them baptized and doused with holy water. But I do believe the
> MIT theory. I've believed it for many years without even knowing about it.
> My own less sophisticated theory states that malevolence is a very real
> aspect of engineering design, even if we aren't cognizant of its presence,
> and I also believe that machines and processes can let their malevolence be
> known to those who are sensitive to these dynamics. It's there, I know it
> is. So I don't curse my machines anymore. And when I get a new software
> program, I put it up there on the shelf with my plants, and I talk to it,
> and I play soft music for it at night, and when I leave my abode, I tell it
> I won't be gone for long. And if you software engineers were half as smart
> as you think you are, you'd be doing the same thing every single day while
> you're hatching all these belligerent beasties -these binary Bart Simpsons.
> ;-)

Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Fri May 12 2000 - 12:16:11 PDT