OFF: [Fwd: Future Fantastic I robot]

[Richard Lockwood]

Hmmm...  Sounds very dubious to me...!

Cheers,

Rich.

** Supercalifragalisticborussiamunchengladbach **

<Begin quoted material>
_______________________________________________________________________________
From: Dave Phelan on Mon, Jul 22, 1996 11:37 am

I don't believe this... Anyone care to correct me?

Dave

>From: rodrigop at olddipesh.agw.bt.co.uk (Paulo Rodrigues)
>Newsgroups: bt.misc
>Subject: Future Fantastic I robot
>
>I came across this site that is all about
>the future fantastic series with wassaerface from
>the xfiles.
>
>http://www.bbcnc.org.uk/tv/tworld/future.html
>
>I've only seen the one show "I Robot" which
>was interesting even if the delivery was a bit
>silly.
>
>http://www.bbcnc.org.uk/tv/tworld/ff2.html
>
>It reminded me of the time when I did some work
>on robotics while at university. We had built
>some six legged vacuum cleaner robots which used
>a combination of genetic algorithms and neural nets
>to learn and evolve. The sole purpose of these creatures
>was to pick up material from the floor and deposit it
>in a central area where it was later collected. The
>critters were rewarded according to the amount of
>material they were able to pick up.
>
>After a number of tests in the lab we decided it was time
>to see how they would fare in the real world. So we let
>them lose in the college canteen, where they spent their
>first day learning to walk. By the end of that first day, to
>the amusement of the other students, almost all of our
>robotic creatures were successfully walking around
>randomly picking up rubbish.
>
>By the end of the second day some most of the robots had
>figured out that if they walked towards the stuff on the
>floor that they could collect more of it and pretty soon they
>got much more efficient. They worked out that where there
>were people, there was rubbish and they started following people
>around in the canteen dashing after scraps.
>
>We were extremely pleased with progress until the fourth day.
>Some of the robots started exhibiting rather strange behaviour.
>Throughout the week the robots had been perfecting their walking
>behaviour, however some of them still had some quirks. On the
>fourth day we noticed that some of the robots had degenerated
>and developed odd and in some cases amusing walking styles.
>
>It took us three days to work out what was happening and in that
>time their behaviour had become deeply disturbing. It turns out
>that an old lady from one of the basket weaving classes had
>been feeding the robots scraps of food as if they were birds and
>those robots which behaved in the most amusing fashion were rewarded
>with extra portions. By  the fifth day they had started to perform
>acrobatics that we would not have thought possible.
>
>It was only on the sixth day, when we heard the old person scream as one
>of the robots accidentally jumped on her and knocked over her
>hand bag, that we found out what had been going on. But by then
>the robots were learning at a tremendous rate. Once they realised that
>they could generate rubbish by jumping and knocking things over, there
>was no stopping them. They were getting on tables, ambushing students
>at the cash desk and even attacking each other to get the scraps.
>
>Regretably we were forced to end the experiment. After that, funding for
>robotics was slashed at the university and we were not able to perfect our
>creations.
>
>===============================================================
>Project -> http://www.lsec.bt.com/old_ec1/projects/glossi.htm
>Unless otherwise stated my views are my own and not those of BT
>===============================================================
>
<end quoted material>