Subj : Re: Neuralink
To   : Moondog
From : Dennisk
Date : Wed Aug 19 2020 22:21:00

-=> Moondog wrote to Dennisk <=-

 Mo>   Re: Re: Neuralink
 Mo>   By: Dennisk to Underminer on Tue Aug 18 2020 09:05 pm


 > No, we are not essentially computers.  You won't find in our brand and gates
 > or gates, flip-flops, microcode.  Our memory works completely different, as
 > does our processing.  We don't work in binary, or definite, strict logic.  O
 > brain uses a vastly different model, so any comparison which goes into more
 > depth than "electrical signals" is useless.  Computers don't understand
 > context, aren't conscious, and cannot think.  They don't fall for visual or
 > auditory illusions.
 >
 > I think when we map how the brain works, we will see that it works using an
 > alien computation model to what our machines do.
 >

 Mo> We are machines.  Organic electro chemical machines that are way more
 Mo> complex than any concepts we currently use to crunch numbers.  Some
 Mo> functions of our bodies have analogs in non-organic mechanics and
 Mo> electronics, howver they are smaller components of more complex
 Mo> systems.  Memory systems alone blows scientists minds because they are
 Mo> aware of generic nodes, however the data is stored and multiplexed in
 Mo> ways that currently don't make sense.

A flagellum is microscopic machine in the cell.  Even the RNA transcription
process is done by machines in our cells.  Our mind interelates different
parts.  For example, when we see an object, our mental image isn't just the
image, but there is contexual information overlayed onto it.  We don't just see
a glass of water, we also see it in a particular context, even if we are not
consciously aware of it.

A good example of how different we work, is how we can easily not be fooled by
someone wearing a T-Shirt with a stop sign, but self-driving cars are.  It
isn't just processing of the image, it is connecting it with everything else we
know, and our bias towards analysing things as objects.

We may one day make a mind like ours, but I think we are far further from it
than we think.  The most interesting part of AI is machine learning through
evolutionary techniques.

.... MultiMail, the new multi-platform, multi-format offline reader!
--- MultiMail/Linux v0.52

.