Seeing red

[emjay]

Say you have a cube with nine 'panels' on each face, like a Rubik's cube, and each of those panels represents a pressure sensor and has a corresponding neuron in the input layer, then it would be interesting to see what representations and expectations form in the network given simulations of what could happen to it... such as just laying on a table (ie nine panels active)... being pushed off a table (nine panels + one on an adjoining face... representing pushing it along with a finger... then some of the nine opposite from where the pressure is turning off as the cube hangs over the edge before falling... then the random way it bounces, rolls, and comes to rest at the bottom, meaning 3 panels each from two adjoining faces if it bounces on an edge, and 1 each from three faces if it bounces on a corner etc... no pattern in the which faces/edges/corners hit first after the fall but you would still expect a general pattern of it's movement to emerge... the rolling motion. Then to 'evolve' this thing you would just add panels/mini-cubes to the surface and add corresponding neurons to the input layer. I'm just curious to see what sort of representations would form and if it could come to represent its own shape. I'm just thinking that about the question of what could be the simplest hypothetical organism to model where you could see, scaled down massively, the principles at work in the human brain.

Sounds a lot like this to me:

Wow, that looks so cool. I've downloaded it now and it looks just the sort of puzzle game I like so I think I'll have a lot of fun with that. 

I see where you're coming from it comparing it to my idea - in that it's got bumpers taking the place of my 'panels' for the robot that he's inside and gets to wire up - but mine was nowhere near as exciting Mine would just be 'presenting' to the network the various patterns that I decided on to represent the events of the outside world and then examining what the neurons learned to detect. But it wouldn't be software like this... it would just be a neural network. But now you've shown this to me it does suggest a way to make it a bit more interesting... to represent my cube... well make it a square... as a sprite in simple virtual world and give it more than just bumpers but also other simple sensors and thrusters for movement. And then perhaps have it 'needing' batteries like in this game and see how it behaves. But that would be a lot more complicated but still a fun project. Who knows, it could help with yours? They could be your little minions?   What do you say... would you like to work on that together? Like we did with Unity?

[bennyboy]

Imagine Robot Odyssey but with much more complex inputs and mechanisms: like sensitivity to varying degrees of different colors of light, different sound frequencies, etc. Then hook it up to ANNs.

Could be a lot of fun, no?

[emjay]

Imagine Robot Odyssey but with much more complex inputs and mechanisms: like sensitivity to varying degrees of different colors of light, different sound frequencies, etc.  Then hook it up to ANNs.

Could be a lot of fun, no?

I'm up for it if you are   But just to be clear, what would you like to do? Your evolution game involving these things or Robot Odyssey as a game with NNs? Because I'm not sure where the little man would fit in if the wiring in the robots of the game was replaced with NNs, unless you mean he could add neurons and connections, individually, in place of the 'soldering' he currently does in the game? Cos that could be interesting... kind of like Minecraft but with neural networks 

[bennyboy]

Imagine Robot Odyssey but with much more complex inputs and mechanisms: like sensitivity to varying degrees of different colors of light, different sound frequencies, etc.  Then hook it up to ANNs.

Could be a lot of fun, no?

I'm up for it if you are   But just to be clear, what would you like to do? Your evolution game involving these things or Robot Odyssey as a game with NNs? Because I'm not sure where the little man would fit in if the wiring in the robots of the game was replaced with NNs, unless you mean he could add neurons and connections, individually, in place of the 'soldering' he currently does in the game? Cos that could be interesting... kind of like Minecraft but with neural networks 

Nope, it would basically just be rock-paper-scissors applied to game objects to get a complex interaction among properties.  Examples:
-Aliens are random colors / attracted to different colors for "mating."
-There are game regions (or objects) where aliens of one color have an advantage over those of other colors
-There are random sounds / attractions / aversions to sounds
-There are game regions (or objects) which emanate random sounds

This is just two variables, but we can already see that: aliens which are attracted to mates in regions which disadvantage them will die out eventually.  Aliens which have an attraction to different sound regions will "pull" apart.

If we made sexual reproduction involving various degrees of child-caring, we'd find eventually that females either become more selective in looking for compatible males, or that they'd look for resource-rich areas that could support a gazillion offspring.

In short, I think with just a few variables, we could kind of simulate a narrow band of the evolutionary process. What if the most resource-rich asteroid was located in a purple zone, and emanated classical music? How many generations would it take for ALL the nearby organisms to prefer that place, that sound, and each others' colors?

This all reminds me of a question for Rhythm and Jorm: is the encoding in DNA an idea?

[Angrboda]

This all reminds me of a question for Rhythm and Jorm: is the encoding in DNA an idea?

What encoding? I don't know what you mean. DNA strands are made up of chemicals, those chemicals in a given environment give rise to chemical reactions varying from the building of a single cell to the construction of entire living organisms.

[The Grand Nudger]

I think you'd have to provide a little more information as to what an idea is, and what you mean by "encoding in DNA".  Do you mean this as an observation, or the explanation for an observation?

[bennyboy]

This all reminds me of a question for Rhythm and Jorm: is the encoding in DNA an idea?

What encoding?  I don't know what you mean.  DNA strands are made up of chemicals, those chemicals in a given environment give rise to chemical reactions varying from the building of a single cell to the construction of entire living organisms.

Well, let's look at your definition of intentionality.  Wouldn't you say, by that definition, that DNA acts with intention over a very long time?  At least a lifetime, for sure, but maybe even over the evolution of a species?

[bennyboy]

I think you'd have to provide a little more information as to what an idea is, and what you mean by "encoding in DNA".  Do you mean this as an observation, or the explanation for an observation?

Well, I'm playing devil's advocate here, by which I mean that I'm trying to ignore my own sense of what mind is, or ideas, and trying to see ideas purely as physical patterns which arise for certain reasons, or which are expressed in certain ways.  So right now, and only right now, I mean that an idea is a representational state, i.e. a partial mirror of an external physical reality.

If DNA represents an idea-- which I think both you and Jorm would say it does-- then that means that mind as we experience it and ideas aren't necessarily the same thing, which would be an important philosophical conclusion we could arrive at even without appealing to qualia.

I don't really want to define idea too narrowly, because I want to draw an analogy between an idea in the human brain, which is short-term and temporary, with other representations of state, like those implied in DNA.

[Angrboda]

What encoding?  I don't know what you mean.  DNA strands are made up of chemicals, those chemicals in a given environment give rise to chemical reactions varying from the building of a single cell to the construction of entire living organisms.

Well, let's look at your definition of intentionality.  Wouldn't you say, by that definition, that DNA acts with intention over a very long time?  At least a lifetime, for sure, but maybe even over the evolution of a species?

According to my definition of intentionality, the intentional representation is part of a recurring, self-sustaining feedback loop in which the thing has behaviors triggered by the representation. In the sense that DNA triggers behaviors such as molecule building, its molecular chain would be about those molecules it is building. However there is no feedback involved, so the DNA molecule isn't properly a representation of the molecule it is building under my definition. But here, in which DNA gives rise to molecule building, is one area that shows the crossover between representation happening and "just stuff happening." In that sense, looking at DNA as 'encoding' the building of certain molecules is a case in which the interpretation of what is happening is firmly "in the eye of the beholder." It is a conceptualization of what is occurring, and that requires high level intentionality of a kind not available to the DNA molecule.

[The Grand Nudger]

Well, I'm playing devil's advocate here, by which I mean that I'm trying to ignore my own sense of what mind is, or ideas, and trying to see ideas purely as physical patterns which arise for certain reasons, or which are expressed in certain ways.  So right now, and only right now, I mean that an idea is a representational state, i.e. a partial mirror of an external physical reality.

Devils advocate for what or whom ?  I think that any definition of idea that does not distinguish between a rock and an idea is likely to yield poor inferences no matter what way you go with it.  

If DNA represents an idea-- which I think both you and Jorm would say it does-- then that means that mind as we experience it and ideas aren't necessarily the same thing, which would be an important philosophical conclusion we could arrive at even without appealing to qualia.

Our conceptualizations of dna are ideas.  DNA itself is an expression of organic chemistry. The concepts themselves are as much(and as little) a code for building life as bernoulli's principle is a code for building a plane.

I don't really want to define idea too narrowly, because I want to draw an analogy between an idea in the human brain, which is short-term and temporary, with other representations of state, like those implied in DNA.

A definition that's too broad runs the risk of defining nothing at all.  You probably answered your own question -in- the question above, we don;t refer to -any- representation of state as an idea. You said it yourself. Expressed in certain ways, for certain reasons. I'm sure you could draw those analogies without any fuss over whether or not dna is an idea, in any case. The things built by dna, for example, are generally temporary themselves, requiring metabolism and growth of new cells or structures to persist, as we do, as living examples of the ship of theseus.