Comments on: Synaptic electronic circuits that learn and forget like neural processes http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes Accelerating Intelligence Sat, 25 May 2013 01:48:43 +0000 hourly 1 http://wordpress.org/?v=3.4.1 By: asiwel http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-81757 asiwel Sat, 29 Dec 2012 18:57:35 +0000 http://www.kurzweilai.net/?p=175469#comment-81757 Hi, @uday. Good idea you have. I just went through my college library on-line journal access system and downloaded this article and it reports on a really interesting piece of research. But I can't email that to you, unfortunately. ACS publications - like ACS Nano - are widely available and it seems very likely that your college library would subscribe to an on-line service that provides access to this publication. Hi, @uday. Good idea you have. I just went through my college library on-line journal access system and downloaded this article and it reports on a really interesting piece of research. But I can’t email that to you, unfortunately. ACS publications – like ACS Nano – are widely available and it seems very likely that your college library would subscribe to an on-line service that provides access to this publication.

]]> By: uday http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-81444 uday Sat, 29 Dec 2012 08:04:54 +0000 http://www.kurzweilai.net/?p=175469#comment-81444 i would like to give a seminar on this topic in my college can any one send me the report on this topic i would like to give a seminar on this topic in my college can any one send me the report on this topic

]]> By: melajara http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80535 melajara Fri, 28 Dec 2012 04:29:11 +0000 http://www.kurzweilai.net/?p=175469#comment-80535 @Bri" It’s going to be interesting to see what type of programming language they would have to come up with". IMHO, It will be more and more programming by (showing) examples, the basic tenet of human teaching. Why? Because the sub-systems will be too complex to be explicitly taught what to do. As for us, in some way they'll have to auto-configure. The progressive transition from explicit to implicit programming in the coming decades will be fascinating to watch, stay tuned ;-) @Bri” It’s going to be interesting to see what type of programming language they would have to come up with”.

IMHO, It will be more and more programming by (showing) examples, the basic tenet of human teaching.

Why? Because the sub-systems will be too complex to be explicitly taught what to do. As for us, in some way they’ll have to auto-configure.

The progressive transition from explicit to implicit programming in the coming decades will be fascinating to watch, stay tuned ;-)

]]> By: eldras http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80338 eldras Thu, 27 Dec 2012 22:48:45 +0000 http://www.kurzweilai.net/?p=175469#comment-80338 you can engineer two ways: 1) get a similar output to the real thing. 2) reverse engineer the real thing. We cant reverse engineer the synapse yet and dont understand it. We have a superficial knowledge of what/how it's working. When we crack it that is most of the brain cracked IMO. It;s devilishly complicated. you can engineer two ways:

1) get a similar output to the real thing.

2) reverse engineer the real thing.

We cant reverse engineer the synapse yet and dont understand it. We have a superficial knowledge of what/how it’s working. When we crack it that is most of the brain cracked IMO.

It;s devilishly complicated.

]]> By: Bri http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80335 Bri Thu, 27 Dec 2012 22:45:54 +0000 http://www.kurzweilai.net/?p=175469#comment-80335 Speed is superfluous, though I'll bet it's faster than our aqueous brains. The big news is thier nonlinearity, and configurability. It certainly would get rid of 1's and 0's. Much more like a dynamic, organic brain. This really is phenomenal and could become quite useful. It's going to beinteresting to see what type of programming language they would have to come up with. I wonder if it might be holographic or interference based. Speed is superfluous, though I’ll bet it’s faster than our aqueous brains. The big news is thier nonlinearity, and configurability. It certainly would get rid of 1′s and 0′s. Much more like a dynamic, organic brain. This really is phenomenal and could become quite useful. It’s going to beinteresting to see what type of programming language they would have to come up with. I wonder if it might be holographic or interference based.

]]> By: Editor http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80274 Editor Thu, 27 Dec 2012 21:13:53 +0000 http://www.kurzweilai.net/?p=175469#comment-80274 Operating speed: I will be getting the original papers and will pass that spec on. Operating speed: I will be getting the original papers and will pass that spec on.

]]> By: asiwel http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80149 asiwel Thu, 27 Dec 2012 18:00:01 +0000 http://www.kurzweilai.net/?p=175469#comment-80149 We have two similar articles today, one on memristors and this one. Simulation is one of the best routes to achieve and demonstrate understanding. However, engineering is often much better for designing useful devices and systems. For example, an unanswered question in this article has to do with the operating speed of these nanoionic devices. If they only work as fast as the objects (neurons, etc.) they simulate ... well, that would not be much of an engineering breakthrough. We have two similar articles today, one on memristors and this one. Simulation is one of the best routes to achieve and demonstrate understanding. However, engineering is often much better for designing useful devices and systems. For example, an unanswered question in this article has to do with the operating speed of these nanoionic devices. If they only work as fast as the objects (neurons, etc.) they simulate … well, that would not be much of an engineering breakthrough.

]]> By: John http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80068 John Thu, 27 Dec 2012 16:04:13 +0000 http://www.kurzweilai.net/?p=175469#comment-80068 Seems like trying to simulate feathers structure in order to build flying machines. Seems like trying to simulate feathers structure in order to build flying machines.

]]> By: melajara http://www.kurzweilai.net/synaptic-electronic-circuits-that-learn-and-forget-like-neural-processes/comment-page-1#comment-80024 melajara Thu, 27 Dec 2012 15:01:51 +0000 http://www.kurzweilai.net/?p=175469#comment-80024 Excellent, a very promising news indeed, welcome to the positronic brain, LOL Maybe some day ca++ ions will be used in such circuits like in biological synapses. Excellent, a very promising news indeed, welcome to the positronic brain, LOL
Maybe some day ca++ ions will be used in such circuits like in biological synapses.

]]>