My Tiny Brain 2 by Karakanlud
<Updated> This is a prototype for the SFPI challenge. This bot is built with a more complex brain structure; A pseudorandom relaxed signal, a sensory decipherer for wall detectors, excitable motor ganglia and a 2 min memory!
complex
malloc
complicated
complacated
basicai
sfpi
memory
neuron
brain
electronics
Comments
-
Schmolendevice 11th Jul 2016
(from prior to your response) I can understand the "neuron triplets" for each motor output of the movement ganglia whereby one neuron inhibits a direction if it knows it is moving towards BRCK while the third neuron excites the creature to move in the opposite direction of that BRCK stimulus. -
Karakanlud 11th Jul 2016
Schmolendevice: Thanks for the review! The random singal is not a proper method for sure, I'm using it for the same reason why I use cross checking now. Until a better method is found, I assume that the creature's movement in absence of an external stimuli is random. -
Karakanlud 11th Jul 2016I'm trying to imagine it as building up a Schroedinger equation; You have a basic module (logging all the perceptions in a given time). Then it can be upgraded by adding modules which log the sequence of the walls. Then maybe one that log walls in 15s blocks. So, in a nutsell, adding more and more approximative methods will possibly make a good simulation of a real memory.
-
Schmolendevice 11th Jul 2016So far, it seems that Karakanlud is using MERC's expansion to partly simulate short term synaptic plasticity. I'm uncertain of the purpose of the "chaos core." Also, once again I'd argue that "random" is not realistic unless one could say that it was an evolved "internal decision querier and driver for sensory change" that produces internal chaos in the absence of external chaos. For now I debate the utility of the "signal multiplicator."
-
Karakanlud 11th Jul 2016Yeah, it's a thing to be done. :/ Maybe I'll try to add a similar deciphering mechanism for memories instead of the current cross checking. Then it could be combined with the random signals to give a more realistic decision system. Also, the memory could be upgraded by addig different types of modules.
-
Schmolendevice 11th Jul 2016The fundamental structure of the brain with all of its lobes and cortical columns I believe evolves to define for the creature a set of "instinctual goals and impulses" that it can "query/decide through" whereby the feedback loop of propagated actions in relation to external changes in sensory data has us relate our internal processes to cause and effect, allowing us to learn about the outside world and indeed, "logically" reason upon future decisions.
-
Schmolendevice 11th Jul 2016@MexLP I haven't read this "officially," but I'd argue that almost nothing about the human brain _itself_ is random. I'd regard it as a highly tailored, developed and evolved machine possessing its own special "neural architecture" whereby the seeming "chaos" of its neural structures and "chaotic plasticity" is rather the result of _chaotic sensory experience_ coming from the outside world as well as some internal changes in neurotransmitter production.
-
MexLP 11th Jul 2016
@Sylvenia isn't a brane doing random decisions , that in the end are "logical" for us , beacuse random is an other word for possibility of whatever , e.g. you can put 2 things "random" on a table (ifinite chances of "random" positioning), in the end it has it own "possible order" , am I right? -
Sylvenia 11th Jul 2016
Now for the constructive critisism: your "brain" makes decisions off of randomized signals, then cross checks them with the enviroment. Your actual brain moves your body by making LOGICAL instead of RANDOM decisions. I do not know much about technology though, so how you could do that is beyond me. Maby you could try to program a sequence of codes to, say, finish a maze, and add remembering codes of where it has been. -
Sylvenia 11th Jul 2016This is truly amazing and deserves front page for the next decade. + 7.8 bill. (one for every person on Earth)