From: Eric Burton (brilanon@gmail.com)
Date: Sat Jan 16 2010 - 13:57:09 MST
OK neuron count decides how much of the top of the screen that makes
up most of the video their neuron weights automatically represent on
and are pulled up or down by the colour channels in, well, that's the
morphogenic field because it's local and forced. Then they have
sensors and motors they can evolve into this field for morphogenic
field tampering if it helps or outright telepathy along those lines...
well, what I'm going to do is make that window a display into a sparse
array, and do morphogenics the same way... no sorry, the morphogenics
will be on a smaller version of the psychic screen you see there --
local still
But yeah, instead of evolved sensors and motors for that screen I'll
force sensor and motor connections for each neuron into this other
window which is a display of a sparse array like space. Also and
firstly to motors for discrete movements over that sparse array.
Forcing the sensors and motors would be too intense for the shared
screen but these would be private -- spawned at random huge
coordinates in sparse-space -- faded to 0, to preserve the RAM
advantage, and they could move away from their own screen if their
noise was a problem. So even without knowing what they were doing a
lot of critters would execute a random walk in this space and leave a
fading trail like pheromones. Then with some scan command to return
the distance of the nearest populated cell and its colour they can
find each other. When windows overlapped they could have telepathy. I
think this is what happens with psilocybin, on Earth
That's all. Like I say in the info, I'll do like a 4 day run at 25 fps
(2.5x less critters) and compress 10M frames into about two minutes
soon, soon. Then I think I'll get that sparse array in there
The change to scroll speed seems to have been advantageous but I think
they're getting smarter too. New code soon
soon
flamoot ,_,
On Sat, Jan 16, 2010 at 3:18 PM, Mike Dougherty <msd001@gmail.com> wrote:
> On Sat, Jan 16, 2010 at 12:55 PM, Eric Burton <brilanon@gmail.com> wrote:
>> Well I'm a little worried now that a ratio of scrolls up to scrolls
>> over that the height of the field is divisible by could be stunting
>> their ability to comprehend its dimensionality, I mean I remember
>> something about prime numbers and aliens, and I think about radio
>> communication with those when trying to train the critters in
>> critterdrug a lot... it's almost comparable except the messages these
>> have to deal with aren't strictly in sequence so it isn't enough to
>> provide just row length, they need to be shown the columns as well.
>> Which is what I'm worried about
>>
>> I'll throw up new source later, but sooner if it turns out to have
>> been important
>
> without context regarding what we're seeing.... this video is not
> terribly exciting. Not that many youtube videos are any good for a
> slew of reasons, but it would be nice to know what we're looking at.
>
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:05 MDT