Improved Design

  • 15 Replies
  • 3812 Views
*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Improved Design
« on: October 07, 2018, 01:16:47 am »
I changed the cube to a dodecahedron to enable incoming signal triangulation. Thus enabling light source naming, which should allow the system to pick and choose which signals it processes. Which should increase the plasticity of the system and give the genetic algorithm more options to play with. The small black rectangles are solar cells (signal detection). The red =led's (signal transmission).
The big black square = battery charger. The grey dome is a green light detector, which will control what is learned with the genetic algorithm.

New_Design" border="0

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #1 on: October 07, 2018, 04:10:18 am »
I cant think of how your signal triangulation is going to work?     
But I do know that it would be good if you could do it,   because your system would be more parallel.


*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: Improved Design
« Reply #2 on: October 07, 2018, 04:24:23 am »
When a light flashes on a different neuron, several solar cells on "our" neuron will detect the light. But they are all at different angles to the light. So each cell will detect a different signal strength. This creates a unique power distribution for each possible incoming signal. It could learn to pass or block signals based on this "name" "code" "power distribution"  like a combination lock.

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #3 on: October 07, 2018, 05:11:43 am »
But I dont see how that would work for simultanious signal travel -  wouldnt the value be disturbed if 2 fired at once?

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: Improved Design
« Reply #4 on: October 07, 2018, 06:36:57 am »
Good point. I'm not sure if that would be good or bad. Probably both. In some cases it might help the system, (then the genetic algorithm would keep it the way it is). If it was preventing some goals of the system, then it would keep changing various parameters until that merging of signals stopped, or became tolerable or even beneficial. That's my hope anyways, that if you get enough nodes together, and you have a system which makes incremental improvements, then it will eventually find a way (possibly counter intuitive or roundabout), to make it all work.

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #5 on: October 07, 2018, 07:02:44 am »
Hrrmrmm....  How is signal confliction supposed to help in any meaningful way at all - its a malfunction.

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: Improved Design
« Reply #6 on: October 07, 2018, 07:29:50 am »
How else do you reduce many sensory inputs into a single choice or action? Maybe if two signals are similar enough to interfere with each other, then it is beneficial to compress them into an average at some point. Maybe after they have served their purpose as individual signals.  That's all speculation though. I've got no idea how it will work beyond the basics of each cell.
One way to find out   ;D

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #7 on: October 07, 2018, 11:37:43 am »
How else do you reduce many sensory inputs into a single choice or action?

I dont mean when two cells are sending to the same cell at the same time,   I mean the confliction between cells sending differently everywhere,  wouldnt there be random spill off to adjacent faces?

[ed]  ah blrug it,  i dont even know what im saying... [/ed]

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Improved Design
« Reply #8 on: October 07, 2018, 11:53:02 am »
Are you making coordinating nanobots already :) ?

I wonder how nanobots will swarm and move stuff, like one pulls out this molecule and the other knows to push the skin fold aside to help and...how... Maybe a hierarchy of functions. Function is save dying limb/human. 50% is another bigish smaller funxtion open arm, but watch out for others. Then this function has mini functions in which some groups in this group of nanosbots do different tasks in sync...All whilst flying in the air as a blob
Emergent          https://openai.com/blog/

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #9 on: October 07, 2018, 11:57:13 am »
Are you making coordinating nanobots already :) ?

Nanobots would work really well!!!!! 
if they all evolved together they would learn the rule very quickly,  if they could communicate to each other,  each nanobot another trial attempt at doing the job.

But this thing that hopesome is doing, is more like one nanobot only,  not a swarm of them.

*

LOCKSUIT

  • Emerged from nothing
  • Trusty Member
  • *******************
  • Prometheus
  • *
  • 4659
  • First it wiggles, then it is rewarded.
    • Main Project Thread
Re: Improved Design
« Reply #10 on: October 07, 2018, 12:52:01 pm »
a nanobot that consists of nanobots! Nice! This could be said for all levels.
Emergent          https://openai.com/blog/

*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #11 on: October 07, 2018, 02:02:11 pm »
a nanobot that consists of nanobots! Nice! This could be said for all levels.

The thing in a thing of all levels.

is a fractal.
is recursion.
is nested for loops.

a stack implemented with recursion?

Programming.

If you could go for all of space,  for all of time, for all things, for all dimensions, for all animations. etcetera etcetera,  you can have the model of anything...   and if only you could run the iteration,  it would be a perfected model.

Now having a perfected model, is possible for a man to achieve, I feel...  so I wonder what god can do?

*

HS

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1175
Re: Improved Design
« Reply #12 on: October 07, 2018, 07:58:47 pm »
How else do you reduce many sensory inputs into a single choice or action?

I dont mean when two cells are sending to the same cell at the same time,   I mean the confliction between cells sending differently everywhere,  wouldnt there be random spill off to adjacent faces?

[ed]  ah blrug it,  i dont even know what im saying... [/ed]

No, I see what you're saying now. You're completely right. There would have to be an intensity thresh-hold so the random noise didn't get passed on. Else It would try to pass along something resembling the universe background radiation.

Are you making coordinating nanobots already :) ?

I wonder how nanobots will swarm and move stuff, like one pulls out this molecule and the other knows to push the skin fold aside to help and...how... Maybe a hierarchy of functions. Function is save dying limb/human. 50% is another bigish smaller funxtion open arm, but watch out for others. Then this function has mini functions in which some groups in this group of nanosbots do different tasks in sync...All whilst flying in the air as a blob

No, but that sounds like a fun project. Necessary to prevent aging in a system like mine. Since i'm making larger scale models of neurons, I guess I'll start with centi-bots. I think it would be great to put cameras on them, and then have them be remote controlled by a traditional computer, it would make them cheaper/smaller.


*

ranch vermin

  • Not much time left.
  • Terminator
  • *********
  • 947
  • Its nearly time!
Re: Improved Design
« Reply #13 on: October 08, 2018, 08:55:11 am »
Here a did a diagram if i were to do this ->


[ed]  the oscillator isnt connected to the battery properly...  bugs... [ed]

I cant actually finish this one,  because Ive got no experience with filters (its a low pass with a high pass.),  and i
cant make a pure sinusoid oscillator,   ive only got square wave oscillators.   But I wonder,  you could maybe get an approximate sine wave with a low pass filter? 

But it would be leaky.
« Last Edit: October 08, 2018, 10:45:37 am by ranch vermin »

*

Korrelan

  • Trusty Member
  • ***********
  • Eve
  • *
  • 1454
  • Look into my eyes! WOAH!
    • YouTube
Re: Improved Design
« Reply #14 on: October 08, 2018, 10:03:02 am »
It thunk... therefore it is!...    /    Project Page    /    KorrTecx Website

 


OpenAI Speech-to-Speech Reasoning Demo
by ivan.moony (AI News )
March 28, 2024, 01:31:53 pm
Say good-bye to GPUs...
by MikeB (AI News )
March 23, 2024, 09:23:52 am
Google Bard report
by ivan.moony (AI News )
February 14, 2024, 04:42:23 pm
Elon Musk's xAI Grok Chatbot
by MikeB (AI News )
December 11, 2023, 06:26:33 am
Nvidia Hype
by 8pla.net (AI News )
December 06, 2023, 10:04:52 pm
How will the OpenAI CEO being Fired affect ChatGPT?
by 8pla.net (AI News )
December 06, 2023, 09:54:25 pm
Independent AI sovereignties
by WriterOfMinds (AI News )
November 08, 2023, 04:51:21 am
LLaMA2 Meta's chatbot released
by 8pla.net (AI News )
October 18, 2023, 11:41:21 pm

Users Online

321 Guests, 0 Users

Most Online Today: 396. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles