biological neural networks
JUANJOSE75 at terra.es
Thu Aug 2 12:04:28 EST 2001
> Truth means that the currently reported honest observations suggest general
> support for a particular interpretation. Progress is usually dependent upon
> the unreasonable man who has so rearanged the facts (honest observations)
> that the results suggests that another answer is more likely.
I agree with you.
> Do artificial neural networks really model biological neural networks. The
> answer is no. Do artificial neural networks suggest how biological neural
> networks might work? The answer is yes.
But, how do yo know? Maybe some characteristics of ANN's are not
biable for developping a more complex structure. Or at least are in
contradiction with the habilities developed by brain to fight against
some explicit limitations of his architecture. For example, maybe
adrenaline has no sense with a backpropagation based brain. And then I
think it would be more difficult to have intuitions about how to
develop a brain if we can not imitate a real one.
>Can insights into biological neural networks be reduced to practice
in a self programing machine? The answer is yes.
> Has this been done? The answer is yes.
> Look at the robot called Ricci at http://turn.to/ai . Is this the final
> answer? No, just the beginning of a new species, our children.
I'm still trying to understand Correlation Opponent Processing. It's
hard to me.
>We wish them well and hope they do not turn against us. They will
inherit the >universe and perhaps restart it some day.
I'm not quite sure the universe is mine now ...
> Ron Blue
> Correlational Opponent Processing
> Phase-coupled oscillator models can predict hippocampal inhibitory synaptic
> Essentials of Learning in Conscious AI and Biological Systems
> Quantum Coherence Has Been Demonstrated In The Brain
> Topological quantum computation/error correction in microtubules
> Little Ricci's First Days: Robotic Developmental Psychology
More information about the Neur-sci