Derek On 4/16/08, Derek Zahn <[EMAIL PROTECTED]> wrote: > > Steve Richfield, writing about J Storrs Hall: > > > You sound like the sort that once the things is sort of > > roughed out, likes to polish it up and make it as good as possible. > > I don't believe your characterization is accurate. You could start with > this well-done book to check that opinion: > > http://www.amazon.com/Beyond-AI-Creating-Conscience-Machine/dp/1591025117 >
Very interesting. > Because you are new to the discussion here you probably don't quite get > the topic of this mailing list (AGI); > I think that I do - see comments after addressing your other comments. the system sort-of described in your papers > I described TWO systems. The one in this thread I specifically designed with a mind to eventually emulate YOU, neuron-by-neuron, synapse-by-synapse, in real time. The one mentioned in my "Comments from a lurker" thread mentions Dr. Eliza, that is designed to solve difficult problems in simple ways that billions of people have missed for a million years, and very likely ANY astronomically-sized AGI machine would miss for centuries. It was unclear how AGI was supposed to quickly do something that was only possible after 10E14 human years of wars and other strife, without having to go through, and even potentially cause the same. Proof by example to me, but apparently still not yet to the remainder of this group, is that there ARE really important things that can only be solved inductively, and that socialized AGI-like humans have SO little inductive abilities that even relatively simple concepts have simply escaped human capabilities for a thousand millennia. I clearly understand this because my own native inductive abilities are also in short supply. I had to "hang on by my fingernails" just to get through differential equations. I eventually developed my own assortment of mental crutches to survive my shortfall in native inductive ability, which were subsequently expanded upon to form Dr. Eliza's concept and innards. does not address any of the issues of that topic (as defined in its core > publications and conferences) so don't be too surprised if people here are > not particularly excited about it. > Hmm, I haven't seen a reference to those core publications. Is there a semi-official list? Much of what is presently known about human neuro-anatomy comes to people from the writings of Dr. William Calvin. I was his assistant at the U of W Department of Neurological Surgery. That was AFTER I had performed one of the first neurological simulations and the first known to have categorized inputs via unsupervised learning. We held each other's feet to the fire, Me for being wet-science correct, and Calvin for models that performed good-math computations. Everyone knows about synapses performing weighted accumulations, but few people know that many/most integrate and differentiate, and that inhibitory synapses are typically VERY non-linear with some VERY interesting transfer function, etc. I published a paper at the first IJCNN in San Diego explaining how everything pointed to wet neurons generally computing with the logarithms of probabilities of assertions being true. That simple fact should have guided future research, but lab researchers not being mathematicians, and neither going to NN conferences, this guiding fact as died away like the echo of some long-forgotten noise. When a tree falls in the forest... My son has beliefs that closely match those expressed by others on this forum, and we sometimes have long arguments about what is and is not reasonable for a human scale neural simulation program - beyond more all-too-human stupidity. My son has also developed the best known (and acknowledged as such at an unrelated WORLDCOMP presentation) general purpose neural net simulation program that runs on a PC, that is at once fast, flexible, and well-instrumented. It has good-looking graphics (that look like contemporary test instruments with fantastic abilities) and is able to stick its tentacles deeply into other applications (like flight simulator) to provide interactive input. I give him all the support that I can, but I still question where this is all going. His program is (presently) written in VB.net, converted from its earlier VB. My own personal interest is in living forever, but regardless of how expanded my brain might become, I suspect that I will STILL have the shortcomings that this sort of architecture brings with it, scary though that might be. THAT was part of my motivation for designing Dr. Eliza, which (it appears to me) could quickly (like in a year of adequate funding) grow beyond any AGI's future problem-solving abilities. It may take the likes of an evolved Dr. Eliza to provide the problem solving ability needed to design the AGI that people are discussing here. Steve Richfield ------------------------------------------- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244&id_secret=101455710-f059c4 Powered by Listbox: http://www.listbox.com
