"Randal A. Koene on Realistic Routes to Substrate-Independent Minds,Teleplace" just concluded (http://telexlr8.wordpress.com/2010/07/04/randal-a-koene-on-realistic-routes-to-substrate-independent-minds-teleplace-july-17-10am-pst) and can continue here.
Realistic Routes to Substrate-Independent Minds (continuation of Teleplace)
(85 posts) (14 voices)-
Editor
Moderator/EditorPosted 1 year ago # -
Giulio Prisco
Key MasterThanks to all those who attended Randal A. Koene talk on Realistic Routes to Substrate-Independent Minds in Teleplace, and thanks to KurzweilAI for creating this discussion forum.
We recorded the full video of the presentation and Q/A in two versions: one from a fixed point of view, and another from the dynamic point of view of a participant who zooms on all slides to read the text better.
Full video of the presentation and Q/A (fixed point of view)
Full video of the presentation and Q/A (participant’s dynamic point of view)This presentation has been a very comprehensive introduction to Substrate-Independent Minds and a very interesting discussion of current research, recent advances, and future possibilities. The Q/A session has been very lively, and there has been no time to ask and answer all the questions raised by the presentation. Please continue the discussion here.
Posted 1 year ago # -
Giulio Prisco
Key Master@Tudor - Berger achievements are spectacular, but what specific result are you referring to?
Posted 1 year ago # -
I enjoyed the discussion. Thanks Giulio.
Before I start posting, I'd like to ask if Randal provided or if we have, anywhere, a clear definition for substrate independent minds. I'd like to look at this in relation to definition of persons (or personhood) and focus it into issues of existence vs./+ consciousness.
Regarding Ted Berger, we do know what he is currently working on, but don't have up-to-date specifics (which would be great to know).
The thrust of what I got out of Randal's talk is that is developing a "knowledge library" and not a method for creating s-ims. This is an important and necessary step, but it is no different than basic strategic planning for focused research projects.
Posted 1 year ago # -
Giulio Prisco
Key MasterThank you for coming and asking interesting questions Natasha. Randal provided definitions at the beginning of the presentation before you arrived, watch the video:
We recorded the full video of the presentation and Q/A in two versions: one from a fixed point of view, and another from the dynamic point of view of a participant who zooms on all slides to read the text better.
Full video of the presentation and Q/A (fixed point of view)
Full video of the presentation and Q/A (participant’s dynamic point of view)Posted 1 year ago # -
Tudor
MemberHello Natasha,
Beyond a knowledge library and dedicated journal, I see ASIM playing the role of key events coordinator, social network community, investor and angel introducing broker, with a long term goal of possibly sponsoring an incubator for critical path projects. Randal correct if i'm too optimistic in my understanding :)
The definition of personhood is a great starting point to understand the logistics needed to identify the critical path for what we are attempting to transfer. I separate personhood into a Universal Me Program, shared by all humans (with a large feature overlap to other primates very likely, briefly discussed in Randal's talk), and the genetic and memory based modifications/settings to that Universal Me Program that make and define the unique individual.
A basic overview of the high level features of the Universal Me Program, as i see them:
-Driven by Appetites
-Constantly on the lookout for attacks to its ability to pursue prosperity via Appetite fulfillment (response with Aggression or Fear depending on size of threat)
-Photographic and Concept Based Abstract Memory system
-Ability to perform parallel what if scenarios with abstract concepts, mostly using logical operations (if lucky)A potential Control Theory for the Universal Me Program (must be degenerate in its ability to overcome simple perturbations, and is very noisy):
-Nested negative feedback loops that have evolved to handle any incoming information in the most efficient way so as to return the Me Program back to its reference state (I call this the Contented Baby System, where any Appetite signal (including boredom) is actually an attack on the reference state, and triggers the negative feedback loops that in turn control memory access, scenario modelling and the body)
As Giulio once said, mapping the behavioral features of the Universal Me Program on to a Control System is a big part of the work that AGI, WBE and ASIM in general focus on. The parameter settings and memories that transform a Me Program into a speaking and socializing individual "ME", pursuing specific goals (though even here there are many universals) seems amenable to being reverse engineered by something along the lines of Bainbridge-Rothblatt interviews, and a future memory prosthesis that lifeblogs the patient's final years at least (enough to deliver continuity) in the worst case, or captures all memories via scanning in the best case.
Regards,
TudorPosted 1 year ago # -
Ted Berger's team reverse-engineered a section of the rat hippocampus and built a chip based on the way that section functions. This chip then sent and received signals to and from the rest of the hippocampus (the section it mimicked having been removed) and it restored functionality with 90% accuracy.
It does not seem unreasonable to suppose that we can further refine this model and achieve a 100% emulation of this section. And, let us suppose this represents 1% of the brain (probably a lot less, it matters not). It does not seem unreasonable to suppose we could progress to 2%...3%..and so on until the whole brain is emulated. You can reduce these numbers to any level you feel comfortable with (each step achieves a 0.0000000000000000000000....1% advance in terms of accuracy and amount of brain emulated) but the logic of this proof-of concept still holds: The brain can be understood and an artificial brain that emulates it to any desired level of accuracy can be built.
Posted 1 year ago # -
So what you're talking about in that(hippocampus) demonstration is simply the processing of signal input/output by undisturbed feedback processes.
But 1% of the brain simulated at near 100% efficiency doesn't mean that the full complexity of such synaptic processes, which are themselves subject to far more switching systems, would approach near simulation of an entire brain.
Posted 1 year ago # -
Tudor
MemberFrom other discussions in the ASIM community, the definitive Proof of Concept was more technically defined as:
"....-> record from and perturb an experimental [neuronal cluster] system so that you capture a nearly complete view of its dynamical phase space
-> [chemically] fix and scan the system
->acquire physical variables from scan
->convert variables to model parameters
->return a highly accurate representation of the original phase space from the model
...-> tweak until we understand how all the scan data maps onto parameters... "according to that definition, would we say that Ted Berger's work was the definitive Proof of Concept?
regards,
tudorPosted 1 year ago # -
Giulio Prisco
Key Master@Tudor: If this is how a definitive Proof of Concept is defined, then I would say Ted Berger's work qualifies. But I miss something from this definition. What I would consider as a definitive proof of concept is a device, designed from the scan data, which can be implanted to replace an original part of a brain system, in such a way as to result in a re-engineered brain that still works as it is supposed to work. Experimentally, not theoretically. Continued below.
@Extie: "Ted Berger's team reverse-engineered a section of the rat hippocampus and built a chip based on the way that section functions". Has the chip been implanted to replace the hippocampus in a rat's brain, and did the rat behave as a normally functioning rat after the implant? It yes, then I would certainly consider this as an almost definitive proof of concept (I say almost because it is difficult to precisely know what happens in the brain of the rat).
Posted 1 year ago # -
Editor
Moderator/Editor@Tudor:
From other discussions in the ASIM community
Hi, could you direct us to those discussions?
Posted 1 year ago # -
Giulio Prisco
Key Master@Editor: the ASIM discussion are in the mailing list of the ASIM community. See the general and contact info in
http://www.carboncopies.org/There are periodic online ASIM workshops in Teleplace. Two workshop have been held until now:
First workshop (with summary video):
http://telexlr8.wordpress.com/2010/06/18/asim2010-1-first-online-workshop-on-advancing-substrate-independent-minds/
Second workshop:
http://telexlr8.wordpress.com/2010/07/11/asim2010-2-second-online-workshop-on-advancing-substrate-independent-minds/The next ASIM event will be the ASIM 2010 Conference, San Francisco, August 16-17. The sessions of the ASIM workshop will run after the Singularity Summit workshop on Monday and Tuesday, as a satellite event to the main Singularity Summit (August 14-15). The Singularity Summit workshop finishes at 5pm on both days, so there will be time to find some dinner before joining us for our evening sessions. The workshop will also be open for remote online access via Teleplace. Contact the organizers to participate:
http://www.carboncopies.org/workshop2I believe Randal is working at a more detailed ASIM writeup.
Posted 1 year ago # -
Tudor
Member@Editor: the ASIM discussions i referred to can be accessed with an invite from Randal, pls inquire
@Giulio: I agree we should increase the Proof of Concept definition, to include the ability to implant the novel device into the original brain and not lose functions.
Posted 1 year ago # -
>Has the chip been implanted to replace the hippocampus in a rat's brain, and did the rat behave as a normally functioning rat after the implant?<
Apparently so, if one can trust reports in Scientific American, such as the following:
'An artificial hippocampus took over from the biological organ the task of consolidating a rat's memory of pressing a lever to receive a drop of water. Normally, the hippocampus emits signals that are relayed to cortical areas responsible for storing the long-term memory of an experience. For the experiment, a chemical temporarily incapacitated the hippocampus. When the rat pressed the correct bar, electrical input from the sensory and other areas of the cortex were chaneled through a microchip which, the scientists say, dispatched the same signals the hppocampus would have sent'.
Bare in mind that this is pop-science reporting in which researchers tend to make grandiose claims about the importance of their work in areas such as health or security in the hope of improving public support and securing funds (or so Geraci said in that Appocalyptic AI book). It might be an exaggeration to say this chip really is, to all intents and purposes, an artificial hippocampus. Maybe it does not yet do everything the hippocampus does?
Posted 1 year ago # -
Glad to see the discussion moving along! I'm gong to repsond to my own post below. (Bty, sorry but I could not find a way to upload a pic.)
One issue I have is “substrate-independent” which is confusing because at first blush, it seems that a substrate-independent mind would exist without any substrate – free of being tethered to a substrate. Nevertheless, he means that it is in depended of being tethered to any one substrate.
I spoke with Randal and his definition: "It is the notion that our minds are computable things. This means they can be implemented/emulated in any Turing-complete hardware (Church-Turing thesis). So, we can use technological approaches such as whole brain emulation (conservative approach) to achieve a transition of mind functions from one substrate (biology) to another. At that point, you have a completely accessible instantiation of mind." Bty, I didn't know that Randal coined "whole brain emulation". That was good to learn.
Now, how can we discuss mind without an epistemology? Randal says he is not focused on philosophy, but that is not the point. We live in a word where words have multiple meanings in multiple domains. It is crucial to first decide if one is an Idealist or Constructivist or Realist or Functionalist (for example). If mind is immaterial, then why the heck copy it?! That seems to be both paradoxical and needless waste of time and energy if mind already exists everywhere. (Obviously I am not an Idealist or a hard-core Constructivist.)
Randal and I agree that issues concerning “self-continuity” are important and while he looks at it in scientific engineering solution ways to develop substrate-independent minds and I look at it as a design enhancement and human evolution to co-exist in different substrates.
Posted 1 year ago # -
Giulio Prisco
Key MasterThe SciAm article quoted bt Extie is Jacking into the Brain.
Natasha: this new KurzweilAI site, based on wordpress, uses the image in your public Wordpress profile. If you don't have one, create one on wordpress.com or use gravatar.com. As far as epistemology is concerned, I suspect we will have routine operational mind uploading technology, used without second thoughts, much before philosophers reach an agreement on what exactly it is that we copy/paste.
Posted 1 year ago # -
Well, that may be so, but so what? Anyway, I'm interested in evolution and framing concepts of continuity and reality -- the generation of knowledge, which is crucial to the continuation of mind (self-continuity). (Bty philosophers came up with that notion.)
We can overlook this perspective as irrelevant, but is it really? I'm not so sure it is irrelevant. In fact, while I have tried to make it irrelevant I realize it is in fact not irrelevant. Social constructs are based on the flow of information and whether it is objective or subjective may not matter when considering coping computations, but what is within the computation? Mind. What is mind? A conglomeration of memories, ideas, perceptions, etc. If those are merely subjective computations, then copying it would require a designer system. If that conglomeration is based on memetic engineering, then that is a different matter. There would be a different protocol for copying mind. Consider prescription drugs. At one point, one prescription fit all. Now we know that the human physiology is unique -- the chemical reactions of each person may be different, based on his/her unique system. Now consider the mind as being far more complex chemistry and where knowledge comes from in our mind does matter. While second-order cybernetics may put all observers within the system, it does not work with the field of human enhancement and self-directed evolution because we are creating ourselves, not just observing our "selves." By and through this reasoning, I do think that each person's "perspective" in obtaining knowledge is viable.
Thanks for the pointer on the pic - stuck with a gravitar image that I can't change.
Posted 1 year ago # -
{i}Pan~
MemberOne issue I have is “substrate-independent” which is confusing because at first blush, it seems that a substrate-independent mind would exist without any substrate – free of being tethered to a substrate. Nevertheless, he means that it is in depended of being tethered to any one substrate.
Indeed. Try Joel Morrison's "Spinbitz"
SpinbitzAnd Nagarjuna's 'Middle Path Buddhism'
NagarjunaNow we know that the human physiology is unique -- the chemical reactions of each person may be different, based on his/her unique system.
–adjective
1.
existing as the only one or as the sole example; single; solitary in type or characteristics
2.
having no like or equal; unparalleled; incomparable
3.
limited in occurrence to a given class, situation, or area
4.
limited to a single outcome or result; without alternative possibilities
5.
not typical; unusualo.O rly http://chuckpalahniuk.net/files/orly.gif
Now consider the mind as being far more complex chemistry and where knowledge comes from in our mind does matter. While second-order cybernetics may put all observers within the system, it does not work with the field of human enhancement and self-directed evolution because we are creating ourselves, not just observing our "selves." By and through this reasoning, I do think that each person's "perspective" in obtaining knowledge is viable.
Now we're getting somewhere.
Second-order cybernetics
Second-order cybernetics, also known as the cybernetics of cybernetics, investigates the construction of models of cybernetic systems. It investigates cybernetics with awareness that the investigators are part of the system, and of the importance of self-referentiality, self-organizing, the subject-object problem, etc.
Ontology: the Cosmic Singularity/ Omega Singularity/ the Ouroboros- the Cosmic Singularity acts as a fundamental feedback logic gate that computes the Multiverse- universes are the decoherred patterns of feedback reflected bit-states forming an attractor space around the event horizon [formed by feed back paths] of the Cosmic Singularity
Cosmology: the universe we observe is virtually produced by a computational version of Lee Smolin's Fecund Universes or Alan Guth's Eternal Inflation where the immanent fractal configuration space of the Cosmic Singularity's feedback computation produces reflections of the Singularity ITSELF throughout the fractal hierarchy- singularities which compute a copy to the Multiverse- each universe with physical laws emergent from computational process of the Singularity [Entropy=Time/accelerated expansion of space- Gravity/locality/space=slowed bandwidth from increased information density- Many World Quantum Mechanics= recursive algorithmic sorting] containing more singularities which again compute the Multiverse- an endless fractal hierarchy of singularities and complex event horizons which repeat the fractal- each a reflection of the Cosmic Singularity- each singularity giving birth to and given birth by the same singularity in an eternal fractal recursion-
-setAI
Seems so simple to me.
But then again.
I'm completely mad ;)
Posted 1 year ago # -
Giulio Prisco
Key Master@Natasha - of course I don't underestimate the importance of understand how the physical brain generates the thinking mind. My point is that we are often able to do something before understanding in detail how we do it. Children do not have formal understanding of the grammar of their native language, but they can understand and speak it anyway, and well enough to communicate with others. Our ancestors were using drums to communicate over long distances much before achieving a formal understanding of the effect of frequency on the propagation of acoustic waves in the atmosphere, etc. So i believe the practical aspects of mind uploading technology will develop in parallel with theoretical understanding, and probably faster.
I often think that the basic consciousness program, the "me program", is essentially the same for everyone ("Universal Me Program") for all practical purposes. If so, the difference between persons is only in their thoughts. In other words, the difference between us is in what we think, not how we think it. Then we can use a standard Turing-Me module able to ignite a database of personal memories, thoughts, feelings, emotions etc. and bring it to life as a valid continuation of the original person (this possibility also validates the Bainbridge-Rothblatt approach in principle).
Posted 1 year ago # -
Computers are often thought of as dualistic, consisting of the 'hardware' on one hand and 'software' on the other.
But are these two things truly independent?
Imagine I have a computer but am running no programs of any kind whatsoever. I have installed NO software, there is only the hardware. What use is this machine? If it is not executing any programs whatsoever is it really a computer or merely something with the potential to BE a computer?
Imagine a program that is not running on any computer. Uninstalled software. Can you use this program? If you are imagining a program is it running on the wetware of your brain, thereby ruining this whole thought experiment? Can software ACTUALLY exist independently of a physical medium?
It seems obvious to me that computers are not dualistic at all because hardware is useless without software and vice versa. One cannot exist in any useful, meaningful sense unless the other is also present. But, of course, a particular piece of software is not tied to a particular piece of hardware. It can be copied to and run on any suitable machine. I think the same can be said of the 'software' of the mind and that this fact will become obvious once we have worked out how to build suitable hardware to transfer/copy it to.
Posted 1 year ago # -
Giulio Prisco
Key Master@Extie some computers are only able to run one program, and some programs are written for just one computer. Actually, this was often the case at the beginning of the computing history in the 50s and 60s. But then we have found a smartest way to do things, with multipurpose computers able to run many programs, and substrate-independent programs that can run on many computers. Evolution may be looking for ways to do the same thing and, through our efforts, it may be finding a way to run its best programs (ourselves) much more efficiently and effectively, as substrate-independent software.
Posted 1 year ago # -
good lecture- I am curious what he would think the ULTIMATE substrate will be- the substrate that will define our civilization for aeons- and if he thinks we will ever truly transcend all material substrates and harness vacuum energy and/or spacetime geometry itself
Posted 1 year ago # -
Hi ipan, Spinbitz looks interesting. Nagarjuna eliminates self, which would be contrary for self-continuity (or continuity of self) :-) and implausible for mind uploading as a material event of computation. Re our living in a simulation, I have to defer to Hans Moravec who I think is the most informed and fluent writer on this topic.
Hi Giulio - Great reply - I agree mind uploading doesn't need to rely on a theoretical approach to develop. Nevertheless, its prognosticators do because it broadens the scope of possibilities of interdisciplinary soft epiphany and hard strategy. It is not how we think in regards to the mechanics of biology or computation but how we think in regards to individual perspective of reality which is based on our relationship with identity and/or nature. What is that? It is a mind? How does it work? What is your reality? How does it see the world - through what filters. How and what are intertwined.
Hi Extropia DaSilva- Yeh, I agree that computers are not dualistic. :-)
Thanks!
/:set\AI - I wonder to! Didn't Randal say that all substrates are comprised of matter. Does emptiness have any density? In design, we consider black to be negative space, but black is comprised of all colors in the spectrum.
Posted 1 year ago # -
Anonymous
Unregistered"black is comprised of all colors in the spectrum"
I think you mean 'white'.
Posted 1 year ago # -
Perception is an interesting phenomena. Most chemists and artists/designers would say that black is a color. This achromatic color contains barely any visible hue because there is very little absorption of light. But this does not mean there is no color! There are multiple shades and types of black. For example, Burnt Umber and French Ultramarine in equal parts will form the richest black - blue-black.
"Here's a simple way to show how black is made: Combine all three primary colors (red yellow and blue) using a liquid paint or you even food coloring. You
won't get a jet black, but the point will be clear."Posted 1 year ago # -
Tudor
Memberthis discussion is paving the way for the next online Teleplace workshop for ASIM, which i believe will focus on the 2nd order cybernetics and other control programs that attempt to describe the system manifesting human self awareness, consciousness, recall algorithms, etc. One prominent topic will be this model (i hope)
http://en.wikipedia.org/wiki/ACT-R
... REAL black means the object doesnt reflect photons, so no color, since color is the perception of photon wavelength spectrums... the Zen answer Mu seems most appropriate on this topic
Posted 1 year ago # -
I am disappointed that I can't make ASIM but I have been in deep discussion with one of its members on RC and with someone else on SR - absolutely contrasting perspectives. Kind of like black and white :-)
I was reading Umpleby's view on a third order cybernetics and while it has some value, it is too weak to really move beyond n-order. <http://www.gwu.edu/~umpleby/recent_papers/2001_what_comes_after_second_order_cybernetics.htm>
And an issue I am currently dealing with is that n-order as not fully satisfactory for engaging substrate-independent minds / whole brain emulation / uploads because n-order hedges on the observer as part of the system and not the observer as a mutable agent/person/being within interconnected systems. Sorry if this is unclear ... one foot out the door.
Posted 1 year ago # -
I have read SpinBitz twice. I enjoyed reading it, but I have to confess that I did not quite understand it. Being an artist, Natasha might enjoy the book for its attempt to convey a new way of vizualizing mathematics, ultimately tying together the philosophical systems of Spinoza and Leibniz (I suspect one needs to be familiar with Spinoza's Ethics and Leibniz's theory of monads in order to make any real sense of this book).
Posted 1 year ago # -
Giulio Prisco
Key Master@Natasha: "how we think in regards to individual perspective of reality which is based on our relationship with identity and/or nature. What is that? It is a mind? How does it work? What is your reality? How does it see the world"
Of course, this has been discussed for decades and will be discussed for many more decades, possibly much after operational mind uploading has been achieved. I don't think anybody has an answer, I am not sure an answer is really necessary, and I am not sure the identity issue is really an issue. At the end of the 19th century many leading physicists tried to grasp the elusive concept of aether, until Einstein found out that physics does not need it (of course some variants of the concept re-surfaced later). I think the analogy is sound.
Perhaps we should not say "I have these thoughts" but "I am these thoughts", In other words, perhaps consciousness should not be thought of as a property of a thinker, but as a property of a train of thoughts.
@Tudor: this is one of the topics that we shortlisted for future ASIM online workshops after San Francisco. Also, a workshop dedicated to the Bainbridge-Rothblatt approach to soft uploading.
Posted 1 year ago # -
Giulio Prisco
Key Master@/:set\AI - "I am curious what he would think the ULTIMATE substrate will be- the substrate that will define our civilization for aeons- and if he thinks we will ever truly transcend all material substrates and harness vacuum energy and/or spacetime geometry itself"
Then, vacuum energy and/or spacetime geometry itself would become the ultimate substrate. According to quantum field physics, the difference between material substrates and vacuum energy or spacetime geometry is more incidental than fundamental.
Sir Arthur C. Clarke, in 2001: "the age of the Machine-entities swiftly passed. In their ceaseless experimenting, they had learned to store knowledge in the structure of space itself, and to preserve their thoughts for eternity in frozen lattices of light. They could become creatures of radiation, free at last from the tyranny of matter. Into pure energy, therefore, they presently transformed themselves; and on a thousand worlds, the empty shells they had discarded twitched for a while in a mindless dance of death, then crumbled into rusty."
Posted 1 year ago # -
Editor
Moderator/Editorvacuum energy and/or spacetime geometry itself would become the ultimate substrate
Right, that idea is supported by theories of three dimensional spatial metrics, which include the The Wheeler-DeWitt equation (http://en.wikipedia.org/wiki/Wheeler%E2%80%93DeWitt_equation) , geometrodynamics (http://en.wikipedia.org/wiki/Geometrodynamics) and "It from Bit" (http://en.wikipedia.org/wiki/It_from_bit). In Programming the Universe (http://books.google.com/books?id=vcExCDmVn-0C&printsec=frontcover&dq=Programming+the+Universe&source=bl&ots=lYBaGbW392&sig=PGwh50zIGNyhxfBRDm3lv4e_lOY&hl=en&ei=6FZJTJaOLYu6sQPlxdFI&sa=X&oi=book_result&ct=result&resnum=10&ved=0CEsQ6AEwCQ#v=onepage&q&f=false), Seth Lloyd takes it further, describing the universe as a giant quantum computer. In that case, we would be a set of dynamic quantum states in that computer. So capturing the essense of a person would be a "simply" :) a matter of mapping the dynamics of their quantum states, based on Wolfram's principle of computational equivalence (http://en.wikipedia.org/wiki/A_New_Kind_of_Science#Mapping_and_mining_the_computational_universe).
Posted 1 year ago # -
Editor
Moderator/Editorfrozen lattices of light
... or perhaps more likely, dynamic lattices of light, modulated based on sets of complex mathematical functions in a supercomputer per Wolfram (ranging from 10^16 computations per second based on a neural model, up to 10^27 based on the Tuszynski c-termini 1Ghz sub-neural model). Interesting question to explore: what about developing an open-source general model of human functions, with shared modular components -- shared modular human qualia? That would reduce the need for a supercomputer per person to one supercomputer for humanity, using shared routines. For example, a memory of the sound "ah" is fairly universal (with some variations). Why store it billions of times?
Posted 1 year ago # -
Editor
Moderator/EditorI forgot to include the Hameroff/Penrose Orch OR model (http://www.kurzweilai.net/forums/topic/stuart-hameroff-defends-orch-or-theory-at-tsc-2010) for mathematical functions and Hameroff's "conscious pilot" model as another way to achieve computational efficiency: http://www.quantumconsciousness.org/pilot.htm. "Autopilot" modes could then become shared routines in the open-source humanity model.
Posted 1 year ago # -
G - I was eliciting the irony of mind because on the one hand it is the most controversial, enigmatic topic mankind has ever known, and on the other mankind want to get a hold of it.. But I am very clear on the circularity of "how" and "what".
Bty, if you "are these thoughts" then it looks like you have no agency and are comprised of random thoughts.
Posted 1 year ago # -
{i}Pan~
Memberor perhaps more likely, dynamic lattices of light
This makes me wonder anew about our light cone...the idea that SETI has been looking in the wrong places for alien life seems to be a common theme lately. I tend to concur. maybe we should be looking in inter-stellar and inter-galactic space for probes. Or the local light cone itself???
Interesting question to explore: what about developing an open-source general model of human functions, with shared modular components -- shared modular human qualia? That would reduce the need for a supercomputer per person to one supercomputer for humanity, using shared routines. For example, a memory of the sound "ah" is fairly universal (with some variations). Why store it billions of times?
This reminds me of subTillions "Earth 2.0" (short excerpt and link):
Earth 2.0:
Rough Notes on a Peristaltic, Vascular Communication and Transportation System for Planet Earth in the Nanotechnological EraJoel Morrison
1/11/20072. Introduction: An active polymorphic, Utility-Fluid* nanotechnological substrate is loosely outlined which could be used to form a multi-scale object, matter/energy transportation, transformation sensor/effector system which could span the Earth, fill and extend the atmosphere and transform it into an active participant in near and sub-terrestrial events. The Utility-Fluid system bears a behavioral macro-level resemblance to John Storrs Hall’s (JoSH) "utility fog," (see http://www.nanotech-now.com/utility-fog.htm) but the similarities between the two systems rapidly diminish at the unit level. Rather, this system uses a "utility fluid" (UFL) model for its architecture as opposed to a gaseous or "utility fog" model, and it is possible that this would provide for a very different set of capabilities at the meso and macro levels.
3. Programmable Matter Units (PMU): The units in this Utility Fluid (UFL) system are similar in structure to vacuous single celled organisms (e.g. amoebas, hydrae, and neurons). The units themselves have a muscular, polymorphic membrane which can expand and contract and modulate their shape at the object surface and subsurface levels, thus imbuing the PMU and its surface detail, texture and properties with rich polymorphic capabilities. When taken together, as a "fluid", the units can slip around each other in a controlled, continuous and fluid manner using sub-surface modulations. these could include: 1. Adaptive Electro-Magnetic Levitation (AEML, to be discussed below), where the surfaces would actually be seperated some distance by magnetic levitation, similar to a gas; 2. Mechanical, 2a. Radial and/or rotary/circumferential movement of sub-surface sub-units and patternings. 2b. Utility Fog Emulation, the unit-surface can also form into radiating arms and thus the individual units can emulate JoSH's foglets in functionality if the necessity arises. The space-filling, "fluid" mode, however, is a welcome alternative to the cumbersome hand-to-hand motion-transfer from unit to unit required by the JoSH method.
3.1. Adaptive Electro-Magnetic Levitation (AEML) -==----==- The polymorphic, surface modulating capabilities make the units necessarily much more complex than the jOSH model, but the basic inter-unit motion-control programming/protocol should be simpler and readily evolvable. The embedded polymorphic, multi-scalar multi-hub, information/power switching network (see below) gives rise to many interesting possibilities. Each fluid unit is a power-switching hub. Thus networks of these units would be able to form higher level adaptable electronic circuits and electricity flow patterns. A surface of densely packed collumns of ~500 microns in diameter, (50 or so units ) could form helical or coil circuits giving rise to a magnetic field. The magnetic field formed from arrays of these AEML cells should be strong enough to provide magnetic levitation at the interface between the capsule and the surface of the muscular contraction. For larger systems the circuits could be formed at larger scales and more power could be drawn through the circuits to provide a more powerful magnetic field. ,(1/( ganglion of sperm brains1\10\1)
3.2. NOTE: [:::::::: .depending upon the possible strength of an AEML cell in the sub-micron region, the units, expanded to space-filling shape, could produce the AEML cells on their surfaces required for levitation. The precise manipulation and synchronization of patterns of AEML fields at the cell-to-cell interfaces could produce the necessary laminar flow forces (electro-magnetic peristalsis) required to actuate a laminar flow. One could imagine many scenarios arising from this capability ... the units could join together to construct machinery for the transportation and launching of other units, aggregates or macroscale objects ....::::.::::.:]
3.3. Polymorphic shell construction
3.3.1. (to be defined)
3.4. NOTE: Morphic gas ((vs. molecular gas)) could also be called magnetic or maglev gas
3.5. NOTE: The core of the unit could contain multiple neural-network supercomputers ?????Posted 1 year ago # -
mattzadak
Memberok so your goal for a substrate is basicly the multiverse
this makes absolute sense according to the holographic universe theory
we are just peripherals of the multiverse
what we are seeking is to be peripherals of computers or really to have computers be peripherals of ourselves
answer bci
i advocate minimaly invasive bci (wireless)
as you said:
"@Extie some computers are only able to run one program, and some programs are written for just one computer. Actually, this was often the case at the beginning of the computing history in the 50s and 60s. But then we have found a smartest way to do things, with multipurpose computers able to run many programs, and substrate-independent programs that can run on many computers. Evolution may be looking for ways to do the same thing and, through our efforts, it may be finding a way to run its best programs (ourselves) much more efficiently and effectively, as substrate-independent software. "
this is the answer LANGUAGE
and interface
all we need is that first little translation, just to bridge that gap, and our computers and brains will take care of the rest in a very short amount of time
this will turn on the faucet
we need to understand that 'binary'
is the fundamental language of the multiverse one or zero on or off here or there
what is our brains primary language how about the multiverse?
we are all made of the same stuff so our coding is also the same
languages are relatively easy to learn especially with teams of people
Posted 1 year ago # -
mattzadak
Memberso what is the most fundamental source
it is frequency
frequency is 'binary'
and it should work the same way as inventing the conventional computer did
you just start with what you can
the point is we are already computers
just as the multiverse is a computer etc
we dont even need the most fundamental binary
that will come when we catch up to 'infinity'
all we need is to learn to directly communicate
once we have system access we can start running programs and writing software
we already do this in both fields to a certain extent
Posted 1 year ago # -
Giulio Prisco
Key Master@Editor, re Hameroff/Penrose. I don't think there is available evidence to strongly suggest that quantum effects play an important role in how the physical brain generate the thinking mind. So I prefer to think horses and not zebras, and I believe the brain/mind system can be explained as an essentially classical (in the sense of non-quantum) system.
Of course, new theoretical and experimental results could change this. If quantum effects are discovered to play an important role in how the physical brain generate the thinking mind, then alternative substrates for running our minds will need to be engineered in such a way as to exhibit similar quantum properties.
Posted 1 year ago # -
If we want an explanation as to why our artificial brains lack the full capacity of human intelligence and consciousness, surely the most plausible explanation is that our models do not yet match the complexity of real brains. Indeed, I expect most robot researchers still make do with the equivilent MIPS of an insect brain.
If, when we finally match the power and organization of higher-organism brains, nothing like generally-intelligent behaviour and consciousness emerges, then we might well look to something like the quantum world (or maybe even way out fringe science stuff) for the missing ingredients. But for now we do not need fringe science explanations for our missing ingredients, because we know our models are missing hundreds+ ingredients that are present in real brains.
Posted 1 year ago # -
{i}Pan~
MemberSo, the reason I posted the link to Spinbitz, and a quote from setai above, is that I think there is a bit of a problem with the way people think about 'intelligence', 'consciousness', 'awareness', etc.
Basically, we have this tendency to view higher level cognitive thought as 'awareness', or more simply: subjectivity itself.
I believe this is an error. One that is simply caused by our association with higher order cognizance or 'intelligence' (or those parts of consciousness emerging from the cerebrum I suppose).
But, there's always this question of whether or not an AI (or an upload, etc.) has what we would call 'subjective experience'. Most people today try to equate this with those higher functions, and those higher functions only. And yet there is this continual debate over whether or not it has the lower order 'subjective experience' (qualia, awareness, etc.). In other words, if it is 'intelligent', does it have a 'self'. Equating 'self' with the higher order emergent consciousness of the cerebral type ignores the problem of low-level awareness, or 'consciousness simpliciter'.
As far as I can tell, this 'consciousness simpliciter' is already inherent/immanent within all matter/energy. That is, everything that exists, has built into it a base level of 'subjective state'.
This is important in AI and/or uploading discussions, because it shows us that we don't have to replicate that particular aspect. Everything already has 'subjectivity', but this is not what we're trying to enhance/augment/copy/replicate, etc. What we want is to simply nurture/grow/develop continual increase in connectivity between all intelligent systems. We are not 'creating' subjectivity, we are enhancing it, on every conceivable level.
In short, I guess it seems the AI community in general (though I give some exception to Ben Goertzel) suffers to some extent a form of 'creatio ex nihilo' when it comes to AI/consciousness.
Edit: This is why when I opened the link at the beginning of this thread, almost immediately I heard/read the words "the mind is a machine", and I immediately closed it, because I already know it's getting off on the wrong premise. It's a false start, exasperating the problems of a philosophy that doesn't actually work any more.
Yes, the brain/mind is 'machine like' in many ways, but this does not, and can not, answer all our fundamental questions about what 'subjective consciousness' is. It's a blind alley, and I wish AI researchers would get hip to that.
Read some Whitehead.
Or Morrison.
Or Heidegger.
Etc, etc, etc.This ultra-materialism paradigm is failed. Time to move on (but not backwards).
Edit2: Whenever I study 'subjectivity', without fail, I come across concepts of 'interiority'. It seems almost impossible to discuss 'subjectivity' without forming some concept of it.
When I look at physics/cosmology, I'm struck by the notion that 'interiority', in physics, can be reduced to a singularity.
It would appear then, that there is a direct, intimate relationship between 'subjective' and 'singularity', and this would explain the ubiquitousness of subjectivity in the universe, and why 'cogito ergo sum' actually works so well.So, I'll leave off, once again (in hopes that you contemplate 'interiority' in more depth), with a quote from setai (and I highly recommend Spinbitz for a far more in depth discussion of 'interiority' vs. 'exteriority' and how it relates to infinites....I can only point at the moon, while Joel and setai got this shit nailed down really well):
Cosmology: the universe we observe is virtually produced by a computational version of Lee Smolin's Fecund Universes or Alan Guth's Eternal Inflation where the immanent fractal configuration space of the Cosmic Singularity's feedback computation produces reflections of the Singularity ITSELF throughout the fractal hierarchy- singularities which compute a copy to the Multiverse- each universe with physical laws emergent from computational process of the Singularity [Entropy=Time/accelerated expansion of space- Gravity/locality/space=slowed bandwidth from increased information density- Many World Quantum Mechanics= recursive algorithmic sorting] containing more singularities which again compute the Multiverse- an endless fractal hierarchy of singularities and complex event horizons which repeat the fractal- each a reflection of the Cosmic Singularity- each singularity giving birth to and given birth by the same singularity in an eternal fractal recursion-
-setAIEdit3:
Also interesting is Ben Goertzel's "Hyperset Models of Self, Will and Reflective Consciousness"Posted 1 year ago # -
Giulio Prisco
Key Master@iPan: you seem to agree with the idea that the "universal me program", subjective consciousness, the bare perception of existence, may be very similar and essentially the same for different sentient beings, which would then be different from each other only in their memories and other aspects of high level cognition.
I certainly do, and therefore I think a promising path to uploading is acquiring a sufficient volume of personal information, basically memories, and running it on a suitable "universal me program" which can be a generic off-the-shelf module.
Posted 1 year ago # -
{i}Pan~
MemberPan
I dont understand what that has anything to do with AI. If consciousness is an attribute of all existence, irrespective of form or substance, then that is one problem solved, what we need is higher cognitive functions. That is what concerns creating AI...
Yes, but from what I see most AI researchers are still eliminative materialists, who believe they can "create" consciousness where there was none before, hence comparing this to a kind of 'creatio ex nihilo'. They look at the problem as literally creating a state of being (subjectivity) in a mass (say, a robot or computer) in which they believe doesn't have any subjectivity. This can only fail, because that bit of mass already has a subjective state, as all matter and energy does. One cannot create a subjective state.
What I'm trying to encourage, is the perspective of looking at it as a project of developing, nurturing, coaxing, and growing more intelligence within systems that already possess a baseline of subjectivity, rather than seeing it as creating something out of nothing.
I believe that when they begin to take this perspective, they will make faster progress. Otherwise, I'd just be spinning my wheels here, having an entertaining, but impractical conversation (I have a friend with a BA in philosophy, I could talk about this stuff all day with him, and he never gets tired of it).
I don't think conscious machines will actually come about until the programmers understand proto-panexperientialism/panpsychism. Then, they'll know what they're actually trying to build, and will stop chasing blind alleys.
Posted 1 year ago # -
virgilic
MemberI don't think conscious machines will actually come about until the programmers understand proto-panexperientialism/panpsychism. Then, they'll know what they're actually trying to build, and will stop chasing blind alleys.
I really dont see the relevance of protopanexperientialism to how a robot must be designed. What would be different in the hardware or the software required?
Posted 1 year ago # -
{i}Pan~
MemberI really dont see the relevance of protopanexperientialism to how a robot must be designed. What would be different in the hardware or the software required?
Well, I think one difference would be to use something like Wolframs simple rules, to build an evolutionary intelligence (in a more robust substrate), rather than 'programming' all the rules into it top-down. The top-down method only gets us expert systems. A bottom up approach might get us something more.
Edit:
Or, consider this, from Ben Goertzel (I'm only copying the conclusion to the paper):8 Conclusion
Suppose the hypotheses presented here are validated, in the sense proposed above. Will this mean that the phenomena under discussion { free will, refective consciousness, phenomenal self { have been "understood"?
According to our panpsychist view, the answer would seem to be "yes," at least in a broad sense { the hyperset models presented would then constitute a demonstratively accurate model of the patterns in physical systems corresponding to the particular manifestations of universal experience under discussion.
And it also seems that the answer would be "yes" according to a purely materialist perspective, since in that case we would have gured out what classes of physical conditions correspond to the "experiential reports" under discussion.
The so-called "hard problem" of consciousness has been ignored here, via sticking with panpsychist or materialist views in which the "hard problem" is not an easy problem but rather a non-problem. The ideas presented here have originated within a patternist perspective, in which what's important is to identify the
patterns constituting a given phenomenon; and so we have sought to identify the patterns corresponding to free will, refective consciousness and phenomenal self. The "hard problem" then has to do with the relationships between various qualities that these patterns are hypothesized to possess (experiential versus
physical) ... but from the point of view of studying brains, building AI systems or conducting our everyday lives, it is generally the patterns (and their subpatterns) that matter.
Finally, if the ideas presented here are accepted as a reasonable approach, there is certainly much more work to be done. There are many diff erent states of consciousness, many diff erent varieties of self, many diff erent aspects to the experience of willing, and so forth. These diff erent particulars may be modeled using hypersets, via extending and specializing the de nitions proposed above. This suggested research program constitutes a novel variety of consciousness studies, using hypersets as a modeling language, which may be guided from a variety of directions including empirics and introspection.Edit:
Another mistake I see people making, is that they don't consider that the 'mind' (proto-experience) is formless.
Plato was right in imagining an 'ideal' realm. However, he was incorrect in thinking that it was an ideal realm of form. The ideal (mind) is formless. Like a point-like singularity.
Purpose often points out that people confuse 'awareness' with 'awareness of...', or in other words, there is awareness, and the object of awareness, which is easy to get mixed up.
Classical AI research has been focused on the 'appearance of...' intelligence, and not actual intelligence. I believe the classical paradigms need to fall before we make true progress towards intelligent machines. The classical physicalist picture doesn't cut it anymore, and since we need to know what the real laws are to manipulate our reality, then we must discover what those actually are, and not cling to outdated world views. And one of those is that a kind of proto-subjective state is ubiquitous in all things that exist. Once we know that, we can work with it. Ignoring it will only lead to failure.Posted 1 year ago # -
Giulio Prisco
Key Master@iPan: even accepting a Panpsychist world-view, a substrate for human consciousness must still be "strong enough" to support human consciousness, Even if a flower or even a rock could be said to be conscious in some sense, they would not suitable as a substrate for a human mind because they do not exhibit a coherent computational activity of sufficient complexity. So, I think the "universal me-program" may well be implementable as a generic off-the-shelf computational module to which specific individual memories can be added, but of course it will need a sophisticated computational substrate. We can run a chess program on a PC, but not on a rock.
Posted 1 year ago #
Reply »
You must log in to post. If you haven't created an account yet, click here.