No doubt that quantum computers are interesting, but the TRUE problem with AI
is we do not really understand how the brain works yet. So far history has
provided us with nothing but failure of big claims. I believe that trend will
continue. Biology is extremely complex. Scientists over simplify mainly because
we lack necessary knowledge of how things work. The good news is in each epic
failure of HUGE claims, we get more knowledge and very useful tools. For
instance, they now have visual processing that can identify skin cancer. This
is touted as AI. I disagree. It's advanced visual processing that can compare
it's learning and make a decision. While impressive as a learned response
decision, it is not abstract thought. It is however what we expect from
computers. The ability to enhance our lives through tasks more suited to
computers.
I'm not against the idea we will one day conquer true AI, but I do believe it
will be long after anyone reading this email is alive. Why? The history is long
on this goal and the progress is extremely slow.
So rest easy the Terminators are not coming just yet `,~)
John Vaughters
On Tuesday, November 13, 2018, 10:24:33 AM EST, Gregg Tracton via TriEmbed
<[email protected]> wrote:
That is why quantum computers are so exciting: the structure of "holographic"
computing is far superior to neurons, much faster than even a neutron net (like
by billions), and the interconnects can be controlled by software (rather than
neurons that are mostly hard-coded). Plus, it's hard to get multiple brains to
work together, but easy for computers, even at the board level.
And if we were to compare apples to apples, then the basic computing element in
brains would be the molecule (like dopamine), not the neurons that accumulate
them and trigger switches. Neurons are more at the level of ALUs or even DSPs,
not transistors.
--gregg
On Tue, Nov 13, 2018 at 8:17 AM John Vaughters via TriEmbed
<[email protected]> wrote:
Interesting article.
Regarding this quote from the article:
"One of the first fundamental differences between the brain and computers is
how their “smallest units” function. Brain neurons can have multiple
connections and react to impulses in a range of different ways. Computer
transistors, by comparison, are switches that, while can be connected to other
transistors, can only have one of two states."
This is the problem. Trying to simulate biologic structure with silicon. Highly
inefficient and structurally different. Biology is just better than what we can
create at this time. I'm not convinced electronics will ever be able to
simulate a brain, but it is a very interesting approach and we can certainly
learn from it. Not a new idea, but very cool to see this scale.
John Vaughters
On Friday, November 9, 2018, 6:25:36 PM EST, Pete Soper via TriEmbed
<[email protected]> wrote:
U of Manchester probably got a price break for this new machine they've made:
SpiNNaker Million-Core Supercomputer
-Pete
_______________________________________________
Triangle, NC Embedded Computing mailing list
To post message: [email protected]
List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org
To unsubscribe, click link and send a blank message:
mailto:[email protected]?subject=unsubscribe
_______________________________________________
Triangle, NC Embedded Computing mailing list
To post message: [email protected]
List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org
To unsubscribe, click link and send a blank message:
mailto:[email protected]?subject=unsubscribe
--
Gregg Tracton: tired, retired & mostly unattired
(PJ's)_______________________________________________
Triangle, NC Embedded Computing mailing list
To post message: [email protected]
List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org
To unsubscribe, click link and send a blank message:
mailto:[email protected]?subject=unsubscribe
_______________________________________________
Triangle, NC Embedded Computing mailing list
To post message: [email protected]
List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org
To unsubscribe, click link and send a blank message:
mailto:[email protected]?subject=unsubscribe
