I'm sorry, but this entire approach is the very approach who's potential for achieving AGI is scientifically proved by the revision to the science I am proposing. It's based on an assumption. A n unproved hypothesis that it can reach AGI. I am merely proposing the science needed to prove whether or not computers, in location (e) right in the diagram, can do the 'computation' natural brain physics does in a manner that makes a robot brain indistinguishable fro a natural brain. I predict it will be proved trivially true: yes, you can do it, but you'd never bother because the model is so vast the computer would drain the universe of resources.
It is a brutal fact that so far the entire enterprise has been based on an unproven hypothesis, and I am trying to convey the formal structure of the science needed to prove it. You do not prove it by assuming it is true. Anyone is free to simply march on and use computers to automate things based on models of brain function. Perfect sound thing to do. It's potential to reach equivalence with natural tissue, however, remains an unproved conjecture. I'be already referenced the key paper twice: Pylyshyn, Z.W. (1980). Computation and cognition: Issues in the foundations of cognitive science. Behavioral and Brain Sciences 3, 111-132. The best part of this paper is in the commentary, in a rejoinder involving John Searle (who's been banging away at this issue of 'causality' or 'causal powers' for decades) is: Pylyshyn writes: If more and more of the cells in your brain were to be replaced by integrated circuit chips, programmed in such a way as to keep the input-output function each unit identical to that of the unit being replaced, you would in all likelihood just keep right on speaking exactly as you are doing now except that you would eventually stop meaning anything by it. What we outside observers might take to be words would become for you just certain noises that circuits caused you to make. This is an amazing statement. There we literally see the birth of the 'silicon replacement' thought experiment and the 'substrate-independence *hypothesis*'. The equivalence of brains and computers is a very complex and unresolved matter. Pylyshyn describes computers as a 'metphor' used in cognitive science. As an artificial brain, the context in which their literal equivalence with nature is scientifically established, it remains a hypothesis to this day. What computers have done so far can be offered as only half of the system of science needed for the real proof. I simply want to do the science of that proof, and when you do that, what it looks like is what I have been writing. Just like Einstein's gravitational waves were a mere hypothesis for a hundred years until the empirical work was done, the proof of the 'substrate independence hypothesis' is waiting. 60 years so far. Like LIGO, the proof is big-science. I'm just about to re-do the silicon replacement thought experiment, only using modern neuroscience and silicon chip tech. You get two different outcomes, depending on choices and the 'granularity' of the conversion. At the end, I'll put the whole discourse together as a single narrative and then look at how to get it into the literature. I am <1/3 of the way through. Hang in there! colin On Mon, Jul 1, 2019 at 6:37 AM Nanograte Knowledge Technologies < [email protected]> wrote: > The following - fascinating - lecture provides a view of the REAL state > of AGI in the world. The follow-up Q&A raises important thoughts and > questions all researchers should be including in their work. If you have > something constructive to contribute to this state of AGI, then good. > > Else, just catch up and read along for the sake of interest. > > Thanks Ben. Much appreciated. > > https://www.youtube.com/watch?v=Zt2HSTuGBn8 > <https://www.youtube.com/watch?v=Zt2HSTuGBn8> > Ben Goertzel:From Here to Human-Level AGI in 4 Simple Steps > <https://www.youtube.com/watch?v=Zt2HSTuGBn8> > Talk at Czech Institute of Informatics, Robotics, and Cybernetics ( > https://www.ciirc.cvut.cz/) on May 21st, 2018 Abstract: AI technology has > entered the mainstream of ... > www.youtube.com > > > ------------------------------ > *From:* Steve Richfield <[email protected]> > *Sent:* Sunday, 30 June 2019 20:08 > *To:* AGI > *Subject:* Re: [agi] ARGH!!! > > Colin, > > At GREAT risk of representing the opinions of others here ... > > The AGI argument seems to be that a computer need NOT work anything like > brains work to be able to problem solving and "think" - that people here > can discover another way, without ANYTHING to guide them, or even > suggesting that such a thing might be possible. Working on such a thing > seems more an act of faith than an act of science. > > Indeed, I have pointed out in past postings that AGI has become a religion > - to create their own God. > > Trying to inject science into religion has so far met with universal > failure, and I expect no better here. > > I suspect the best you/we can do is make our case to attract those rare > dropins who really ARE interested in science rather than religion, and make > weekly postings for newbies to notice - at least until being cast out by > the religious freakos here. Notice the calls to cast out Arthur from the > group. > > Steve > > > On Sat, Jun 29, 2019, 8:57 PM Colin Hales <[email protected]> wrote: > > > > On Sun., 30 Jun. 2019, 12:20 pm Costi Dumitrescu, < > [email protected]> wrote: > > What is the most likely biological cell for an AGI robot to invent first? > > > Please be patient. I will get to my proposed design. Fundamentally, it's a > 3D cellular automaton and each cell is literally made of the physics it's > made of in the brain except it uses electrons/holes as charge carriers. > > The design has cell migration, dendrite/axon growth and shrinkage, cell > genesis and apoptosis. It has adaptive cell function, action potentials and > ephaptic coupling. It has no model. It literally is a brain. But not > organic. > > Patience ! > > I am more interested in getting everyone to understand AGI is not a > computer science project. It's empirical neuroscience. > > Regards > Colin > > > > > > > > *Artificial General Intelligence List <https://agi.topicbox.com/latest>* > / AGI / see discussions <https://agi.topicbox.com/groups/agi> + > participants <https://agi.topicbox.com/groups/agi/members> + delivery > options <https://agi.topicbox.com/groups/agi/subscription> Permalink > <https://agi.topicbox.com/groups/agi/Te7e0d9ec1ecb60aa-M7613000d15a26bb8f3aeea4e> > ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Te7e0d9ec1ecb60aa-M91161cba1df7391d7eb18187 Delivery options: https://agi.topicbox.com/groups/agi/subscription
