Alt-BEAM Archive

Message #03061



To: beam beam@corp.sgi.com
From: rdraycott@CPL.co.uk
Date: Fri, 7 May 1999 10:32:45 +0100
Subject: Re: Evolution





All this talk of what does and does not drive evolution makes me laugh.
What is it about people that make them think that Bob shannon will be
struck
down from the heavens if his idea isn't sound?

For heavens sake, people, let him be. He's come up with a really good
idea,
and it should at least be given a go. THIS is evolution. When I came onto
the scene a year ago, there was Chiu's 'cool' contest to enter. Now, six
months later, the ideas have evolved so that we have 'innovative' in his
competition, we have another which is organised around finding little balls
of conductive foil, we have yet another which is completely about
behaviour.

I still can't understand why the people who don't like bob's ideas can't
just sit back and let evolution take its course? If his ideas are so
flawed, then let him waste his time on his competition, and laugh when it
doesn't work! Then, when he has learnt his lesson, he will evolve the
competition into something that does take the form of the robot into
account.
It's almost as if those people don't believe in evolution themselves! If
they did, then they would realise that his ideas aren't going to work, and
then his ideas would evolve until they did work.

Instead, people bicker and fight and say "No, that CAN'T work, it's not
evolution, the galapagos islands aren't really a result of evolution,
they're a result of ... umm.... no predators... or.. something... ". If
the
competition idea isn't sound, then write it down in your diary, and have a
big old laugh when the competition doesn't work.


For my part, I think it's a great idea. At my university we have a subject
where the object is to create a robot that runs a maze, looking for a
special arrangement of walls. Once this is found, the robot must go back
to
the start, and get from the start to the finish in the least possible time.

All the robots are based around a MiCE controller board (80196 uC, 256 k
RAM, about 60 input/outputs) manufactured in the university that is
programmed in either C or assembler from a PC. So the hardware is pretty
much the same. The students are left to decide which
motors/geartrains/sensors they will use. Most of the mice are pretty much
the same - same speed, physical arrangement/size, and sensor type/position.

The robot that won the race last time despite it's lack of sophisticated
machinery, was one that was able to move diagonally. There was a long
straight 'staircase' style corridor that went about 6 squares diagonally
along the maze. You had to go along this to get to the finish. All the
other robots went north 1 step, east one step, north one step, east one
step, etc. This one went zipping along straight northeast, and won by a
mile! A clear case of behaviour being very important.

We also have the annual competition where autonomous robots try to pick up
some disks, and deposit them near a target after negotiating a tennis net.
Most of the robots are heavily-modified model helicopters, and none of them
have yet achieved the task successfully. The ones that win are always the
ones with the best behaviour (Which usually means a $2000 student effort),
not the million-dollar effort from team mitsubishi which has all manner of
gizmos stuck all over it.

I say, Go for it. I got my mice board working last night on my iMac
(Keyspan USB to serial adapter, din-8 to 25 pin serial adapter, homemade 25
pin to 9-pin adapter to fix the Dick Smith cable swapping of
transmit/receive, RealPC running the MS-DOS program ACME), so look out! My
entry is coming!

Ben

>
>Ah, but is not entirely true. A famous example that comes to mind is one
of
>the Galapagos finks that use tools (twigs, branches etc.) to fill the
>ecological niche of a woodpecker without having the physical means to do
>so. This increases it's survival ability greatly, without any physical
>modification.
>
>
>You didnt quite get it did you SO I WILL AKE IT NICE AND S L O W
>you are telling me that a bird has a behavior and is not physically
>equiped to do so , well is it using the tool or not! and this long term
>behavior had triggered no anatomical changes OH I see you mean short tearm
>behavioral changes don't trigger the development of a physical changes
>the anatomical mutations dictated physical development on the basis of the
>fittest for the enviroment, those physical mutations dictate behavior not
>visa versa
> the emphasis being on the enviroment as I understand it the Galapagos
>islands are an evolutionry dead end no preditors so the fink aint changed
>much in XXXX years COS IT DIDNT NEED TO its current layout is the fittest
>for its enviroment.
>LIKE I SAID
>
>>ie entity A has a slightly bigger brain and more >synaptic links so can
>>remember more of what is benificial and whats not, >entity A survives
>longer
>>has more prodgeny etc. etc
>>entity A's actual behavior is the result of that >knowlage base ( the
>>physical changes give the advantage).
>>Individuals change behavior on the basis of their >physical circumstances
>,
>
>
>This is not completely accurate either. There is an example of, again,
>birds on the isolated Coco island.
>These birds are members of the same species, but because they have no
>competition they have been able to radiate and fill all niches on the
small
>island. Some birds prefer eating seeds, other insects, but they look
exacly
>the same and breed without discrimination.
>========================================
>here we go again
>they have such a varied diet to fill the untaped recources of their
>ENVIROMENT ie the X entitys eat X fruit while Y entity eats Y insect
>besides the biologists ar discovering that birds apparently of the same
>species are in fact a seperate species, but at this point we get in to the
>hole ball game of species definition and species divergance point
>Also why the hell do you think tha Crocodiles havent changed much in XXXX
>years
>Because the are optimised for their enviroment!!!!!!
>====================================================
>>it would not reflect the population if you only >tested amputees, who
>>ignoring the socialogcal behavirs adapt their >behavior to suit their
>>physical condition .
>>Selective breeding can pick up on some core behaviors >(agressiveness,
>etc)
>>but those exist as a result of the initial physical >evolution of the
>>entity.
>
>
>>Rob D
>
>>Aka "Powerbuilder"
>
>
>
>------------------------------------------------------------------------
>eGroups Spotlight:
>"Kosovo-Reports" - Direct reports from Kosovo/Serbia/Yugoslavia.
>http://clickhere.egroups.com/click/5
>
>eGroup home: http://www.eGroups.com/group/alt-beam
>http://www.eGroups.com - Simplifying group communications
>





------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/group/alt-beam
http://www.eGroups.com - Simplifying group communications

Home