Re: [agi] Is there any Contest or test to ensure that a System is AGI?
Try this one ... http://www.bentham.org/open/toaij/openaccess2.htm If the test subject can be a scientist, it is an AGI. cheers colin Steve Richfield wrote: Deepak, An intermediate step is the reverse Turing test (RTT), wherein people or teams of people attempt to emulate an AGI. I suspect that from such a competition would come a better idea as to what to expect from an AGI. I have attempted in the past to drum up interest in a RTT, but so far, no one seems interested. Do you want to play a game?! Steve On Sun, Jul 18, 2010 at 5:15 AM, deepakjnath deepakjn...@gmail.com mailto:deepakjn...@gmail.com wrote: I wanted to know if there is any bench mark test that can really convince majority of today's AGIers that a System is true AGI? Is there some real prize like the XPrize for AGI or AI in general? thanks, Deepak *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] SyNAPSE might not be a joke ---- was ---- Building a machine that can learn from experience
Ed, Comments interspersed below: Ed Porter wrote: Colin, Here are my comments re the following parts of your below post: ===Colin said== I merely point out that there are fundamental limits as to how computer science (CS) can inform/validate basic/physical science - (in an AGI context, brain science). Take the Baars/Franklin IDA project It predicts nothing neuroscience can poke a stick at... ===ED's reply=== Different AGI models can have different degrees of correspondence to, and different explanatory relevance to, what is believed to take place in the brain. For example the Thomas Serre's PhD thesis Learning a Dictionary of Shape-Components in Visual Cortex: Comparison with Neurons, Humans and Machines, at from http://cbcl.mit.edu/projects/cbcl/publications/ps/MIT-CSAIL-TR-2006-028.pdf , is a computer simulation which is rather similar to my concept of how a Novamente-like AGI could perform certain tasks in visual perception, and yet it is designed to model the human visual system to a considerable degree. It shows that a certain model of how Serre and Poggio think a certain aspect of the human brain works, does in fact work surprisingly well when simulated in a computer. A surprisingly large number of brain science papers are based on computer simulations, many of which are substantially simplified models, but they do given neuroscientists a way to poke a stick at various theories they might have for how the brain operates at various levels of organization. Some of these papers are directly relevant to AGI. And some AGI papers are directly relevant to providing answers to certain brain science questions. You are quite right! Realistic models can be quite informative and feed back - suggesting new empirical approaches. There can be great cross-fertilisation. However the point is irrelevant to the discussion at hand. The phrase does in fact work surprisingly well when simulated in a computer illustrates the confusion. 'work'? according to whom? surprisingly well? by what criteria? The tacit assumption is that the model's thus implemented on a computer will/can 'behave' indistinguishably from the real thing, when what you are observing is a model of the real thing, not the real thing. *HERE *If you targeting AGI with a benchmark/target of human intellect or problem solving skills, then the claim made on any/all models is that models can attain that goal. A computer implements a model. To make a claim that a model completely captures the reality upon which it was based, you need to have a solid theory of the relationships between models and reality that is not wishful thinking or assumption, but solid science. Here's where you run into the problematic issue that basic physical sciences have with models. There's a boundary to cross - when you claim to have access to human level intellect - then you are demanding a equivalence with a real human, not a model of a human. ===Colin said== I agree with your : /At the other end of things, physicists are increasingly viewing physical reality as a computation, and thus the science of computation (and communication which is a part of it), such as information theory, have begun to play an increasingly important role in the most basic of all sciences./ ===ED's reply=== We are largely on the same page here ===Colin said== I disagree with: But the brain is not part of an eternal verity. It is the result of the engineering of evolution. Unless I've missed something ... The natural evolutionary 'engineering' that has been going on has /not/ been the creation of a MODEL (aboutness) of things - the 'engineering' has evolved the construction of the /actual/ things. The two are not the same. The brain is indeed 'part of an eternal verity' - it is made of natural components operating in a fashion we attempt to model as 'laws of nature',,, ===ED's reply=== If you define engineering as a process that involves designing something in the abstract --- i.e., in your a MODEL (aboutness of things) --- before physically building it, you could claim evolution is not engineering. But if you define engineering as the designing of things (by a process that has intelligence what ever method) to solve a set of problems or constraints, evolution does perform engineering, and the brain was formed by such engineering. How can you claim the human brain is an eternal verity, since it is only believed that it has existing in anything close to its current form in the last 30 to 100 thousand years, and there is no guarantee how much longer it will continue to exists. Compared to much of what the natural sciences study, its existence appears quite fleeting. I think this is just a terminology issue. The 'laws of nature' are the eternal verity, to me. The dynamical output they represent - of course that does whatever it does. The
Re: [agi] SyNAPSE might not be a joke ---- was ---- Building a machine that can learn from experience
Ed,
I wasn't trying to justify or promote a 'divide'. The two worlds must be
better off in collaboration, surely? I merely point out that there are
fundamental limits as to how computer science (CS) can inform/validate
basic/physical science - (in an AGI context, brain science). Take the
Baars/Franklin IDA project. Baars invents 'Global Workspace' = a
metaphor of apparent brain operation. Franklin writes one. Afterwards,
you're standing next to to it, wondering as to its performance. What
part of its behaviour has any direct bearing on how a brain works? It
predicts nothing neuroscience can poke a stick at. All you can say is
that the computer is manipulating abstractions according to a model of
brain material. At best you get to be quite right and prove nothing. If
the beastie also underperforms then you have seeds for doubt that also
prove nothing.
CS as 'science' has always had this problem. AGI merely inherits its
implications in a particular context/speciality. There's nothing bad or
good - merely justified limits as to how CS and AGI may interact via
brain science.
I agree with your :
/At the other end of things, physicists are increasingly viewing
physical reality as a computation, and thus the science of computation
(and communication which is a part of it), such as information theory,
have begun to play an increasingly important role in the most basic of
all sciences./
I would advocate physical reality (all of it) as /literally /computation
in the sense of information processing. Hold a pencil up in front of
your face and take a look at it... realise that the universe is
'computing a pencil'. Take a look at the computer in front of you: the
universe is 'computing a computer'. The universe is literally computing
YOU, too. The computation is not 'about' a pencil, a computer, a human.
The computation IS those things. In exactly this same sense I want the
universe to 'compute' an AGI (inorganic general intelligence). To me,
then, this is /not/ manipulating abstractions ('aboutnesses') - which is
the sense meant by CS generally and what actually happens in reality in CS.
So despite some agreement as to words - it is in the details we are
likely to differ. The information processing in the natural world is not
that which is going on in a model of it. As Edelman said(1) /A theory
to account for a hurricane is not a hurricane/. In exactly this way a
computational-algorithmic process about intelligence cannot a-priori
be claimed to be the intelligence of that which was modelled. Or - put
yet another way: That {THING behaves 'abstract- RULE-ly'} does not
entail that {anything manipulated according to abstract-RULE will become
THING}. The only perfect algorithmic (100% complete information content)
description of a thing is the actual thing, which includes all
'information' at all hierarchical descriptive levels, simultaneously.
I disagree with:
But the brain is not part of an eternal verity. It is the result of
the engineering of evolution.
Unless I've missed something ... The natural evolutionary 'engineering'
that has been going on has /not/ been the creation of a MODEL
(aboutness) of things - the 'engineering' has evolved the construction
of the /actual/ things. The two are not the same. The brain is indeed
'part of an eternal verity' - it is made of natural components operating
in a fashion we attempt to model as 'laws of nature'. Those models,
abstracted and shoehorned into a computer - are not the same as the
original. To believe that they are is one of those Occam's Razor
violations I pointed out before my xmas shopping spree (see previous-1
post).
---
Anyway, for these reasons, folks who use computer models to study human
brains/consciousness will encounter some difficulty justifying, to the
basic physical sciences, claims made as to the equivalence of the model
and reality. That difficulty is fundamental and cannot be 'believed
away'. At the same time it's not a show-stopper; merely something to be
aware of as we go about our duties. This will remain an issue - the only
real, certain, known example of a general intelligence is the human. The
intelligence originates in the brain. AGI and brain science must be
literally joined at the hip or the AGI enterprise is arguably
scientifically impoverished wishful thinking. Which is pretty much what
Ben said...although as usual I have used too many damned words!
I expect we'll just have to agree to disagree... but there you have it :-)
colin hales
(1) Edelman, G. (2003). Naturalizing consciousness: A theoretical
framework. Proc Natl Acad Sci U S A, 100(9), 5520--24.
Ed Porter wrote:
Colin,
From a quick read, the gist of what your are saying seems to be that
AGI is just engineering, i.e., the study of what man can make and
the properties thereof, whereas science relates to the eternal
verities of reality
Re: [agi] Building a machine that can learn from experience
J. Andrew Rogers wrote: On Dec 18, 2008, at 10:09 PM, Colin Hales wrote: I think I covered this in a post a while back but FYI... I am a little 'left-field' in the AGI circuit in that my approach involves literal replication of the electromagnetic field structure of brain material. This is in contrast to a computational model of the electromagnetic field structure. Here is a silly question: If you can specify it well enough to implement the desired result in hardware, why can't you implement it in software? It is equivalent, after all. And if you can't specify the dynamic well enough to implement it virtually, why would there be any reason at all to believe that it will do anything interesting? The hallmark of a viable AGI theory/design is that you can explain why it *must* work in sufficient detail to be implementable in any medium. J. Andrew Rogers /If you can specify it well enough to implement the desired result in hardware, why can't you implement it in software? It is equivalent, after all./ The answer to this is that you /can /implement it in software. But you won't do that because the result is not an AGI, but an actor with a script. I actually started AGI believing that software would do it. When I got into the details of the issue of qualia (their role and origins) I found that software alone would not do the trick. If an AGI is to be human equivalent, it must be able to do what humans do. One of those behaviours is science. Getting the 'logical dynamics' of software to cohere with a 'law of nature' is, I believe, impossible for software alone, because the software model of the dynamics cannot converge on an externally located and intrinsically unknown (there is no model!) knowledge. How a software model of a modeller of the intrinsically unknown (a scientist) can work is something I have had to grapple with. In the end I had to admit that software seemed less plausible than actually implementing the full physics of brain material. Hence my EM approach. The simplest way to get to the position I inhabit is to consider that the electromagnetic field has access to more information (about the world outside the agent) than that available through peripheral nerve signaling. It's the additional information that is thrown away with a /model/ of the electromagnetic field. It's replaced with the arbitrary and logically irrelevant electromagnetic fields of the computer substrate (basically noise). The spatially expressed EM field inherits (dynamics is altered by) information from the external world directly via space.. The EM fields play a very important role in the dynamics, adaptation and regulation of neural activity and none of it is captured with existing neural models - as it acts /orthogonally/ , coupling neurons spatially via EM events, not dendrite/axon routes. It's outside the neurons in the spaces in between. It's the reason cortex is layered and columnar. Cortex is 50% astrocytes by volume. They are all charged up to -80mV. and are intimately involved in brain dynamics. Because the boundary of the cells and space is as much an information source as all the peripheral nerve 'boundaries' (the surface of your body), and the boundary is literally electromagnetism (there's nothing else there!), you can't model it for the same reason you can't model the peripheral nerve signals (you have to have the EM fields for the same reason that you need to have a retina or camera)... by extrapolation everything else follows. The EM coupling effects are the subject of my PhD and will be out in detail ASAP. Bits of it will be published - It's been a real trial to get the work into print. I tried to get a publication into the AGI conference but ran out of time. The original hodgkin-huxley model (upon which all modern neural modelling is based) threw out (or at least modelled out) the EM field. If you look in the original 1952 papers you'll see there are batteries, non-linear, time-varying resistors and ignored and off to one side all by itself, waiting patiently a little capacitor. That little capacitor hides the entire EM field spatial behaviour. If you drew the model properly all the components in the model actually span the dielectric of the capacitor between its little plates. The capacitor is actually linked to lots of other capacitors in a large 3-dimensional mesh. You can't delete (via a model) the capacitors because their dynamics is controlled (very very lightly but significantly) by the external world. So...My approach puts the fully spatially detailed EM field back into the model. The little HH capacitor turns into an entire new complex model operating orthogonally to the rest of the circuit. That capacitor radically changes the real model of brain material. There is spatial coupling to other neurons that happens using the field that has been averaged out and confined inside the dielectric
Re: [agi] Building a machine that can learn from experience
). The laws in T' are the rules of the 'natural' CA. For the physics buffs amongst us: DAS explains why there are 2 kinds of 'theory of everything' MATHS (uber-group-maths-theories - of - appearances) and 'STUFF' (branes, strings, loops, preons etc theories of structure). The reason for the 2 types is that thr former is the T-aspect version of the latter, which is the T' aspect equivalent. The poor buggers doing T' aspect don't know they're actually fighting over the same thing! Only from a DAS perspective does EM-field (the U1 group)-as-consciousness acquire any explanatory authority. /So to explain my chips I have to change the whole of science! /That's 1 impossible thing before breakfast. Only 5 to go. cheers colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Building a machine that can learn from experience
J. Andrew Rogers wrote: On Dec 19, 2008, at 12:13 PM, Colin Hales wrote: The answer to this is that you can implement it in software. But you won't do that because the result is not an AGI, but an actor with a script. I actually started AGI believing that software would do it. When I got into the details of the issue of qualia (their role and origins) I found that software alone would not do the trick. Nonsense, an algorithmic system is describable entirely based on input and output without any regard for its internal structure. If two blackbox systems produce identical output based on identical input, then they are mathematically equivalent in every meaningful sense even if they have radically different internal construction. *I'm not clear how you came to the conclusion that I was discussing an 'algorithmic system'. * You say actor with a script as if that means something important, ignoring that every process in our universe is necessarily equivalent to an actor with a script. Your magical EM chip is, in fact, an actor with a script. *If you mean that the laws of nature persist and govern the regularities around us and the nature of ourselves. Fine! We are all 'actors' in that sense. * *What I meant specifically is that we know already the outcome. It's the script. WE wrote it and gave it to our AGI. Consider my benchmark behaviour the artificial scientist. What 'algorithmic script' dictates the behaviour that will result in scientific encounter with the intrinsically a-priori unknown? If you can write that script you must know the outcome already or be willing to predefine the nature of all novelty! * The simplest way to get to the position I inhabit is to consider that the electromagnetic field has access to more information (about the world outside the agent) than that available through peripheral nerve signaling. It's the additional information that is thrown away with a model of the electromagnetic field. This does not even make sense. Either the software model captures the measurable properties of the EM field or it does not, but either way it does not support your proposal. In the former case, the external input and dynamic *must* be measurable and therefore can be reflected in the software model, and in the latter case it is nothing more than handwaving about something you are asserting exists in the complete absence of material evidence. I'm having a hard time accepting that there is something you can specify and measure that magically has no useful software description. That is not even supposed to be possible as a kind of basic mathematical theorem thing. I mean, you are asserting that some very specific inputs to the system are not being modeled, and if you know this then you can very easily add them to the software-modeled system. You have not explained why this is not possible, merely asserted it. J. Andrew Rogers *My position is that the EM field as it exists, literally, in situ in the brain, is enacting exactly a process of 'input'. To us it appears as, for example, 'visual perception' (that you are using to read this). For the same reason you can't model your inputs - because you don't known them - you can't model the EM field. Yes this looks like an assertion BUT: a)I have preliminary physiological evidence that the EM fields are directly coupled to the distant natural world. It comes from studies of visual acuity in humans. I can detail it if you like. b)DISTANT WORLD _is to_: RETINAL IMPACT is MANY:ONE. This is an inverse problem and unsolvable in principle. Yet humans solve it. Therefore vision, to a human, cannot be an inverse problem - it can't be! So we must have extra data. There are only 2 places to get it: (i) SENSORY I/O and (ii) SPACE. We already use (i) so SPACE it is. A Sherlock Holmes outcome. c)The human brain goes to extravagant lengths (and lots of energy expenditure) to micro-manipulate an incredibly complex field pattern in space. That field pattern is gone in any model of it. d) 75 years of computer-based-AGI failure - has sent me a message that no amount of hubris on my part can overcome. As a scientist I must be informed by empirical outcomes, not dogma or wishful thinking. Apart from that:Over XMAS I am finishing the proposal for an experiment to verify the EM field spatial link. I will be getting money from NICTA. I hope! If anyone out there has US$300K to spare and wants some science excitement let me know! I'll have the results in a couple of years. A human scientist (my benchmark, testable behaviour) is a 'black box' with internal states somehow able to resonate (cohere) accurately with the distant and novel natural world, /not the inputs/. The inputs are redundantly related to the distant natural world on multiple levels. **On balance - I go with the brain as my exemplar. There seems to be sufficient available evidence to doubt
Re: [agi] Building a machine that can learn from experience
Ben Goertzel wrote: Goodness. I have to tell you, Colin, your style of discourse just SOUNDS so insane and off-base, it requires constant self-control on my part to look past that and focus on any interesting ideas that may exist amidst all the peculiarity!! And if **I** react that way, others must react that way more strongly, because I'm rather tolerant of wackiness of most sorts... So, I must suggest that if you want folks to take your ideas seriously, you should try to find different ways of expressing them... ben See the recent post and current neuroscience experimentation. eg Milstein, J. N. and Koch, C. (2008). Dynamic moment analysis of the extracellular electric field of a biologically realistic spiking neuron. Neural Computation, 20(8), 2070-84. There's a long way to go and of course dual aspect science has to be established to make any sense of it at all. I can't help that it's new, confronting and awkward. I didn't make it up. I just looked. This area has been problematic for a reason. That reason is /us/ (scientists, I mean). Be very careful in using the word 'extra-sensory'. and 'wacky'. It implies some kind of imagined space-cadet quackery and fringe science. If a quite clear (but merely relatively unexplored) physical phenomenon enacted at the boundary of the atoms in brain material and space can be held accountable for a phsyically testable outcome - then the word 'extra' will be invalid. There will be 6 basic 'senses'. No magic, just physics. I am as dry as an emprically informed scientist can get. You can't deliver any evidence at all that the processes I am investigating are invalid. Until they are properly investigated I'd prefer to leave these words out of the discourse - and I'll refrain from the use of any similar words in reference to the heartfelt beliefs of others in this list. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Building a machine that can learn from experience
J. Andrew Rogers wrote: On Dec 19, 2008, at 5:35 PM, Ben Goertzel wrote: The problem is that **there is no way for science to ever establish the existence of a nonalgorithmic process**, because science deals only with finite sets of finite-precision measurements. I suppose it would be more accurate to state that every process we can detect is algorithmic within the scope of our ability to measure it. Like with belief in god(s) and similar, the point can then be raised as to why we need to invent non-algorithmic processes when ordinary algorithmic processes are sufficient to explain everything we see. Non-algorithmic processes very conveniently have properties identical to the supernatural, and so I treat them similarly. This is just another incarnation of the old unpredictable versus random discussions. Sure, non-algorithmic processes could be running the mind machinery, but then so could elves, unicorns, the Flying Spaghetti Monster, and many other things that it is not necessary to invoke at this time. Absent the ability to ever detect such things and lacking the necessity of such explanations, I file non-algorithmic processes with vast number of other explanatory memes of woo-ness of which humans are fond. Like the old man once said, entia non sunt multiplicanda praeter necessitatem. Cheers, J. Andrew Rogers One of the tricks with Occam's razor is knowing when you've fallen foul of it. Consider the logic: (a) Human scientists are intelligent (a useful AGI benchmark) and made of the natural world. (b) Humans (scientists) model the natural world. (c) The scientific models of the natural world are amenable to algorithmic representation. does NOT entail that (d) systems that involve algorithmic representation will necessarily be as intelligent as a human scientist (capable of human science). Humans are not algorithmic representations or models of a scientist. We are actual scientists. There's some kind of serious confusion going on here between: (i) intelligence capable of construction of algorithmic regularities (humans) and (ii) intelligence made of the algorithmic regularities thus constructed (a computationalist AGI) (ii) requires subscription to a belief in something extra: that the universe is constructed of abstractions or is a computer running the laws of nature as a program or that magical emergentism is a law of nature or any one of 100 other oddities. Furthermore (i) does NOT entail that the universe is non-algorithmic! I would say that the universe is an absolutely brilliant and wonderful huge and exquisite algorithm. *But it's NOT a model of anything.* It IS the thing itself. I choose to do (i) by using the same actual processes that humans are. That /is/ capable of human intelligence (scientists). /That I know for sure./ That is the only thing we know for sure. I also know that I have to invent (believe) something unproven and extra to make (ii) a route to AGI. Occam's razor prevents me from taking that position. So the argument cuts both ways! 1+1=FROG. On the planet Blortlpoose the Prolog language does nothing but construct cakes. :-) This algorithmic nonsense was brought to you by the natural brain electrodynamics of Colin Hales' brain. and ALL of it (humans, Blortlpoose and cakes) is made of whatever the universe is actually made of, which is NOT abstract representations produced by itself of itself. This is just going to go round and round as usual so I'll jump off the merry-go-round for now. */Xmas is afoot! There's meaningless consumerism and pointless rituals to be enacted! Massive numbers of turkeys and chickens must die! :-)/* cheers colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Building a machine that can learn from experience
Steve Richfield wrote: Richard, On 12/18/08, *Richard Loosemore* r...@lightlink.com mailto:r...@lightlink.com wrote: Rafael C.P. wrote: Cognitive computing: Building a machine that can learn from experience http://www.physorg.com/news148754667.html Neuroscience vaporware. It isn't neuroscience yet, because they haven't done any science yet. It isn't vaporware yet because they have made no claims of functionality. In short, it has a LONG way to go before it can be considered to be neuroscience vaporware. Indeed, this article failed to make any case for any rational hope for success. Steve Richfield DARPA buys G.Tononi for 4.9 $Million! For what amounts to little more than vague hopes that any of us here could have dreamed up. Here I am, up to my armpits in an actual working proposition with a real science basis... scrounging for pennies. hmmm...maybe if I sidle up and adopt an aging Nobel prizewinner...maybe that'll do it. nah. too cynical for the festive season. There's always 2009! You never know merry xmas, all Colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Building a machine that can learn from experience
YKY (Yan King Yin) wrote: DARPA buys G.Tononi for 4.9 $Million! For what amounts to little more than vague hopes that any of us here could have dreamed up. Here I am, up to my armpits in an actual working proposition with a real science basis... scrounging for pennies. hmmm...maybe if I sidle up and adopt an aging Nobel prizewinner...maybe that'll do it. nah. too cynical for the festive season. There's always 2009! You never know You talked about building your 'chips'. Just curious what are you working on? Is it hardware-related? YKY Hi, I think I covered this in a post a while back but FYI... I am a little 'left-field' in the AGI circuit in that my approach involves literal replication of the electromagnetic field structure of brain material. This is in contrast to a computational model of the electromagnetic field structure. The process involves a completely new chip design which looks nothing like what we're used to. I have a crucial experiment to run over the next 2 years. The results should be (I hope) the basic parameters for early miniaturised prototype. The part of my idea that freaks everyone out is that there is no programming involved. You can adjust the firmware settings for certain intrinsic properties of the dynamics of the EM fields. But none of these things correspond in any direct way to 'knowledge' or intelligence. The chips (will) do what brain material does, but without all the bio-overheads. The thing that caught my eye in the thread subject Building a machine that can learn from experience... is that if you asked Tononi or anyone else exactly where the 'experience' is, they won't be able to tell you. The EM field approach deals with this very question /first/. The net EM field structure expressed in space literally /is/ the experiences. All learning is grounded in it. (Not input/output signals) I wonder how anyone can claim that a machine that learns from experience when you haven't really got a cogent, physical and biologically plausible, neuroscience informed view of what 'experience' actually is. But there you go... guys like Tononi get listened to. And good luck to them! So I guess my approach is likely to remain a bit of an oddity here until I get my results into the literature. The machines I plan to build will be very small and act like biology... I call them artificial fauna. My fave is the 'cane toad killer' that gets its kicks by killing nothing but cane toads (these are a major eco-disaster in northern australia). They can't reproduce and their learning capability (used to create them) is switched off. It's a bit like the twenty-something neural dieback in humans... after that you're set in your ways. Initially I want to build something 'ant-like' to enter into robo-cup as a proof of concept anyway that's the plan. So the basics are: all hardware. No programming. The chips don't exist yet, only their design concept (in a provisional patent application just now). I think you get the idea. Thanks for your interest. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Should I get a PhD?
Hi, I went through this exact process of vacillation in 2003. I have a purely entrepreneurial outcome in mind, but found I needed to have folk listen to me. In order that some comfort be taken (by those with $$$) in my ideas, I found, to my chagrin...that having a 'license to think = PhD' (as opposed to actually being able to think), seemed to be necessary. Unfortunately the testicular fortitude of the folks needs careful attention - which is more than merely your assurances that you can think and have a clue. So since 2004 I have been on a weird journey where my ideas have not changed, but my ability to convey the concepts to academics and $$$buckets has improved. Throughout this whole process my ability to think never altered a bit! Having decided to do a PhD...in being in NICTA and melbourne uni I also encountered a whole pile of other course opportunities: commercialisation, Intellectual property, media-handling, presentation skills, interpersonal communications, multicultural and globalised research project management, grant application all of which have rounded out my skills to cope with the post-doctoral process of getting my chips built. I have also met/communicated with a raft of luminaries that I would otherwise not have encountered. So - as a vehicle to connect to established $flows and get $creds - it seems that for me, anyway, a PhD is a necessary step. Mind youif I had already had the $$ back in 2003 I would be building my chips by now... not doing this weird PhD Dance and being forced to kowtow to the spurious, precious and fashion-ridden gods of the academic publishing circuit... in that regard I regard the PhD as necessary collateral damage. like Ben: SHOW ME THE MONEY! :-) cheers the not quite yet Dr Col Ben Goertzel wrote: On Wed, Dec 17, 2008 at 6:12 PM, YKY (Yan King Yin) generic.intellige...@gmail.com mailto:generic.intellige...@gmail.com wrote: If...you want a non-research career, a Ph.D. is definitely not for you. I want to be either an entrepreneur or a researcher... it's hard to decide. What does AGI need most? Further research, or a sound business framework? It seems that both are needed... To venture an explicitly self-centered comment: Since I already have a workable design for human-level AGI ... what AGI most needs is someone to give the OpenCog or Novamente projects a large sum of money 8-D ben *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=123753653-47f84b Powered by Listbox: http://www.listbox.com
Re: [agi] Machine Knowledge and Inverse Machine Knowledge...
And 'deep blue' knows nothing about chess. These machines are manipulating abstract symbols at the speed of light. The appearance of 'knowledge' of the natural world in the sense that humans know things, must be absent and merely projected by us as observers, because we are really really good at that kind of projection. The reason? ...Put any sufficient novelty in front of the machine and you get: a) nonsense/fragility/breakdown. or b) the response which results from a human model of what 'novelty' looks like followed by the output of a human model of what to do with the novelty. Surely the 'knowledge' in these creatures must ultimately be grounded in human style cognition and perception ... which is not perfect ... but it's a novelty-handler light-years ahead of the models of novelty handling we give these critters. Think about it...In order that we bestow on a machine a perfect (human level will do until a better one turns up!) novelty handler, we have to have an abstract model of everything already. If we already know everything, then why build the machine? The usefulness of these machines is their behaviour in the face of ignorance. We get to be clever by being serendipitously 'not wrong' by being allowed to be wrong in a non-fatal way. We get to choose to 'know' something very novel...eg invent a concept which may or may not have anything to do with reality... we then can test it to see if it makes sense as a model of reality (out there in the natural world). We then get to be 'not wrong', as opposed to being 'right'. Our intelligence operates completely backwards to the 'knowledge' models of the critters under discussion. Or, put slightly more technically: the 'dynamics' of a human mind (that represents the gold standard of 'knowledge' and 'knowledge change') and the dynamics of a /model/ of the human mind can part company in significant ways ... In what sense has that departure been modelled or otherwise accounted for in the model? There's an oxymoron lurking in these kind of expectations of our machines...or is it just me projecting? cheers, Colin Matt Mahoney wrote: No, I don't believe that Dr. Eliza knows nothing about normal health, or that Cyc knows nothing about illness. -- Matt Mahoney, [EMAIL PROTECTED] *From:* Steve Richfield [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Tuesday, December 9, 2008 3:21:18 PM *Subject:* Re: [agi] Machine Knowledge and Inverse Machine Knowledge... Matt, It appears that either you completely missed the point in my earlier post, that Knowledge + Inverse Knowledge ~= Understanding (hopefully) There are few things in the world that are known SO well that from direct knowledge thereof that you can directly infer all potential modes of failure. Especially with things that have been engineered (or divinely created), or evolved (vs accidental creations like mountains), the failures tend to come in the FLAWS in the understanding of their creators. Alternatively, it is possible to encode just the flaws, which tend to spread via cause and effect chains and easily step out of the apparent structure. A really good example is where a designer with a particular misunderstanding of something produces a design that is prone to certain sorts of failures in many subsystems. Of course, these failures are the next step in the cause and effect chain that started with his flawed education and have nothing at all to do with the interrelationships of the systems that are failing. Continuing... On 12/9/08, *Matt Mahoney* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Steve, the difference between Cyc and Dr. Eliza is that Cyc has much more knowledge. Cyc has millions of rules. The OpenCyc download is hundreds of MB compressed. Several months ago you posted the database file for Dr. Eliza. I recall it was a few hundred rules and I think under 1 MB. You have inadvertently made my point, that in areas of inverse knowledge that OpenCyc with its hundreds of MBs of data still falls short of Dr. Eliza with 1% of that knowledge. Similarly, Dr. Eliza's structure would prohibit it from being able to answer even simple questions regardless of the size of its KB. This is because OpenCyc is generally concerned with how things work, rather than how they fail, while Dr. Eliza comes at this from the other end. Both of these databases are far too small for AGI because neither has solved the learning problem. ... Which was exactly my point when I referenced the quadrillion dollars you mentioned. If you want to be able to do interesting things for only ~$1M or so, no problem IF you stick to an appropriate corner of the knowledge (as Dr. Eliza does). However, if come out of the corners, then be prepared to throw your $1Q at it. Note here that I am NOT disputing your ~$1Q, but rather I am
Re: [agi] Now hear this: Human qualia are generated in the human cranial CNS and no place else
Trent Waddington wrote: On Tue, Nov 18, 2008 at 4:07 PM, Colin Hales [EMAIL PROTECTED] wrote: I'd like to dispel all such delusion in this place so that neurally inspired AGI gets discussed accurately, even if your intent is to explain P-consciousness away... know exactly what you are explaining away and exactly where it is. Could you be any more arrogant? Could you try for me, cause I think you're almost there, and with a little training, you could get some kind of award. Trent It's a gift. :-) However I think I might have max'ed out. Some people would call is saying it the way it is. As I get older/grumpier I find I have less time for treading preciously around the in garden of the mental darlings to get at the weeds. I also like to be told bluntly, like you did. Time is short. You'll be free of my swathe for a while...work is piling up again. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] A paper that actually does solve the problem of consciousness
Richard Loosemore wrote: Colin Hales wrote: Dear Richard, I have an issue with the 'falsifiable predictions' being used as evidence of your theory. The problem is that right or wrong...I have a working physical model for consciousness. Predictions 1-3 are something that my hardware can do easily. In fact that kind of experimentation is in my downstream implementation plan. These predictions have nothing whatsoever to do with your theory or mine or anyones. I'm not sure about prediction 4. It's not something I have thought about, so I'll leave it aside for now. In my case, in the second stage of testing of my chips, one of the things I want to do is literally 'Mind Meld', forming a bridge of 4 sets of compared, independently generated qualia. Ultimately the chips may be implantable, which means a human could experience what they generate in the first person...but I digress Your statement This theory of consciousness can be used to make some falsifiable predictions could be replaced by ANY theory of consciousness can be used to make falsifiable predictions 1..4 as follows.. Which basically says they are not predictions that falsify anything at all. In which case the predictions cannot be claimed to support your theory. The problem is that the evidence of predictions 1-4 acts merely as a correlate. It does not test any particular critical dependency (causality origins). The predictions are merely correlates of any theory of consciousness. They do not test the causal necessities. In any empirical science paper the evidence could not be held in support of the claim and they would be would be discounted as evidence of your mechanism. I could cite 10 different computationalist AGI knowledge metaphors in the sections preceding the 'predictions' and the result would be the same. SoIf I was a reviewer I'd be unable to accept the claim that your 'predictions' actually said anything about the theory preceding them. This would seem to be the problematic issue of the paper. You might want to take a deeper look at this issue and try to isolate something unique to your particular solution - which has a real critical dependency in it. Then you'll have an evidence base of your own that people can use independently. In this way your proposal could be seen to be scientific in the dry empirical sense. By way of example... a computer program is not scientific evidence of anything. The computer materials, as configured by the program, actually causally necessitate the behaviour. The program is a correlate. A correlate has the formal evidentiary status of 'hearsay'. This is the sense in which I invoke the term 'correlate' above. BTW I have fallen foul of this problem myself...I had to look elsewhere for real critical dependency, like I suggested above. You never know, you might find one in there someplace! I found one after a lot of investigation. You might, too. Regards, Colin Hales Okay, let me phrase it like this: I specifically say (or rather I should have done... this is another thing I need to make more explicit!) that the predictions are about making alterations at EXACTLY the boundary of the analysis mechanisms. So, when we test the predictions, we must first understand the mechanics of human (or AGI) cognition well enough to be able to locate the exact scope of the analysis mechanisms. Then, we make the tests by changing things around just outside the reach of those mechanisms. Then we ask subjects (human or AGI) what happened to their subjective experiences. If the subjects are ourselves - which I strongly suggest must be the case - then we can ask ourselves what happened to our subjective experiences. My prediction is that if the swaps are made at that boundary, then things will be as I state. But if changes are made within the scope of the analysis mechanisms, then we will not see those changes in the qualia. So the theory could be falsified if changes in the qualia are NOT consistent with the theory, when changes are made at different points in the system. The theory is all about the analysis mechanisms being the culprit, so in that sense it is extremely falsifiable. Now, correct me if I am wrong, but is there anywhere else in the literature where you have you seen anyone make a prediction that the qualia will be changed by the alteration of a specific mechanism, but not by other, fairly similar alterations? Richard Loosemore At the risk of lecturing the already-informed ---Qualia generation has been highly localised into specific regions in *cranial *brain material already. Qualia are not in the periphery. Qualia are not in the spinal CNS, Qualia are not in the cranial periphery eg eyes or lips. Qualia are generated in specific CNS cortex and basal regions. So anyone who thinks they have a mechanism consistent with physiological knowledge could conceive of alterations reconnecting periphery
[agi] Now hear this: Human qualia are generated in the human cranial CNS and no place else
Mike Tintner wrote: Colin:Qualia generation has been highly localised into specific regions in *cranial *brain material already. Qualia are not in the periphery. Qualia are not in the spinal CNS, Qualia are not in the cranial periphery eg eyes or lips Colin, This is to a great extent nonsense. Which sensation/emotion - (qualia is a word strictly for philosophers not scientists, I suggest) - is not located in the body? When you are angry, you never frown or bite or tense your lips? The brain helps to generate the emotion - (and note helps). But emotions are bodily events - and *felt* bodily. This whole discussion ignores the primary paradox about consciousness, (which is first and foremost sentience) : *the brain doesn't feel a thing* - sentience/feeling is located in the body outside the brain. When a surgeon cuts your brain, you feel nothing. You feel and are conscious of your emotions in and with your whole body. I am talking about the known, real actual origins of *all* phenomenal fields. This is anatomical/physiological fact for 150 years. You don't see with your eyes. You don't feel with your skin. Vision is in the occipital cortex. The eyes provide data. Skin provides the data, CNS somatosensory field delivers the experience of touch and projects it to the skin region. ALL perceptions, BAR NONE, including all emotions, imagination, everything - ALL of it is actually generated in cranial CNS. Perceptual fields are projected from the CNS to appear AS-IF they originate in the periphery. The sensory measurements themselves convey no sensations at all. I could give you libraries of data. Ask all doctors. They specifically call NOCICEPTION the peripheral sensor and PAIN the CNS (basal...inferior colliculus or was it cingulate...can't remember exactly) percept. Pain in your back? NOPE. Pain is in the CNS and projected (Badly) to the location of your back, like a periscope-view. Pain in your gut? NOPE. You have nociceptors in the myenetric/submucosal plexuses that convey data to the CNS which generates PAIN and projects it at the gut. Feel sad? Your laterally offset amygdala create an omnidirectional percept centered on your medial cranium region. etc etc etc etc YESBrains don't have their own sensors or self-represent with a perceptual field. So what? That's got nothing whatever to do with the matter at hand. CUT cortex and you can kill off what it is like percepts out there in the body (although in confusing ways). Touch appropriate exposed cortex with a non-invasive probe and you can create percepts apparently, but not actually, elsewhere in the body. The entire neural correlates of consciousness (NCC) paradigm is dedicated to exploring CNS neurons for correlates of qualia. NOT peripheral neurons. Nobody anywhere else in the world thinks that sensation is generated in the periphery. The *CNS* paints your world with qualia-paint in a projected picture constructed in the CNS using sensationless data from the periphery. Please internalise this brute fact. I didn't invent it or simply choose to believe it because it was convenient. I read the literature. It told me. It's there to be learned. Lots of people have been doing conclusive, real physiology for a very long time. Be empirically informed: Believe them. Or, if you are still convinced it's nonsense then tell them, not me. They'd love to hear your evidence and you'll get a nobel prize for an amazing about-turn in medical knowledge. :-) This has been known, apparently perhaps by everybody but computer scientists, for 150 years.Can I consider this a general broadcast once and for all? I don't ever want to have to pump this out again. Life is too short. regards, Colin Hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] A paper that actually does solve the problem of consciousness
Richard Loosemore wrote: Colin Hales wrote: Richard Loosemore wrote: Colin Hales wrote: Dear Richard, I have an issue with the 'falsifiable predictions' being used as evidence of your theory. The problem is that right or wrong...I have a working physical model for consciousness. Predictions 1-3 are something that my hardware can do easily. In fact that kind of experimentation is in my downstream implementation plan. These predictions have nothing whatsoever to do with your theory or mine or anyones. I'm not sure about prediction 4. It's not something I have thought about, so I'll leave it aside for now. In my case, in the second stage of testing of my chips, one of the things I want to do is literally 'Mind Meld', forming a bridge of 4 sets of compared, independently generated qualia. Ultimately the chips may be implantable, which means a human could experience what they generate in the first person...but I digress Your statement This theory of consciousness can be used to make some falsifiable predictions could be replaced by ANY theory of consciousness can be used to make falsifiable predictions 1..4 as follows.. Which basically says they are not predictions that falsify anything at all. In which case the predictions cannot be claimed to support your theory. The problem is that the evidence of predictions 1-4 acts merely as a correlate. It does not test any particular critical dependency (causality origins). The predictions are merely correlates of any theory of consciousness. They do not test the causal necessities. In any empirical science paper the evidence could not be held in support of the claim and they would be would be discounted as evidence of your mechanism. I could cite 10 different computationalist AGI knowledge metaphors in the sections preceding the 'predictions' and the result would be the same. SoIf I was a reviewer I'd be unable to accept the claim that your 'predictions' actually said anything about the theory preceding them. This would seem to be the problematic issue of the paper. You might want to take a deeper look at this issue and try to isolate something unique to your particular solution - which has a real critical dependency in it. Then you'll have an evidence base of your own that people can use independently. In this way your proposal could be seen to be scientific in the dry empirical sense. By way of example... a computer program is not scientific evidence of anything. The computer materials, as configured by the program, actually causally necessitate the behaviour. The program is a correlate. A correlate has the formal evidentiary status of 'hearsay'. This is the sense in which I invoke the term 'correlate' above. BTW I have fallen foul of this problem myself...I had to look elsewhere for real critical dependency, like I suggested above. You never know, you might find one in there someplace! I found one after a lot of investigation. You might, too. Regards, Colin Hales Okay, let me phrase it like this: I specifically say (or rather I should have done... this is another thing I need to make more explicit!) that the predictions are about making alterations at EXACTLY the boundary of the analysis mechanisms. So, when we test the predictions, we must first understand the mechanics of human (or AGI) cognition well enough to be able to locate the exact scope of the analysis mechanisms. Then, we make the tests by changing things around just outside the reach of those mechanisms. Then we ask subjects (human or AGI) what happened to their subjective experiences. If the subjects are ourselves - which I strongly suggest must be the case - then we can ask ourselves what happened to our subjective experiences. My prediction is that if the swaps are made at that boundary, then things will be as I state. But if changes are made within the scope of the analysis mechanisms, then we will not see those changes in the qualia. So the theory could be falsified if changes in the qualia are NOT consistent with the theory, when changes are made at different points in the system. The theory is all about the analysis mechanisms being the culprit, so in that sense it is extremely falsifiable. Now, correct me if I am wrong, but is there anywhere else in the literature where you have you seen anyone make a prediction that the qualia will be changed by the alteration of a specific mechanism, but not by other, fairly similar alterations? Richard Loosemore At the risk of lecturing the already-informed ---Qualia generation has been highly localised into specific regions in *cranial *brain material already. Qualia are not in the periphery. Qualia are not in the spinal CNS, Qualia are not in the cranial periphery eg eyes or lips. Qualia are generated in specific CNS cortex and basal regions. You are assuming that my references to the *foreground* periphery correspond to the physical
[agi] test
--- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] Ethics of computer-based cognitive experimentation
Dear Matt, Try running yourself with empirical results instead of metabelief (belief about belief). You'll get someplace .i.e. you'll resolve the inconsistencies. When inconsistencies are *testably *absent, no matter how weird the answer, it will deliver maximally informed choices. Not facts. Facts will only ever appear differently after choices are made. This too is a fact...which I have chosen to make choices about. :-) If you fail to resolve your inconsistency then you are guaranteeing that your choices are minimally informed. Tricky business, science: an intrinsically dynamic process in which choice is the driver (epistemic state transition) and the facts (the epistemic state) are forever transitory , never certain. You can only make so-called facts certain by failing to choose. Then they lodge in your brain (and nowhere else) like dogma-crud between your teeth, and the rot sets in. The plus side - you get to be 100% right. Personally I'd rather get real AGI built and be testably wrong a million times along the way. cheers, colin hales Matt Mahoney wrote: --- On Wed, 11/12/08, Harry Chesley [EMAIL PROTECTED] wrote: Matt Mahoney wrote: If you don't define consciousness in terms of an objective test, then you can say anything you want about it. We don't entirely disagree about that. An objective test is absolutely crucial. I believe where we disagree is that I expect there to be such a test one day, while you claim there can never be. It depends on the definition. The problem with the current definition (what most people think it means) is that it leads to logical inconsistencies. I believe I have a consciousness, a little person inside my head that experiences things and makes decisions. I also believe that my belief is false, that my brain would do exactly the same thing without this little person. I know these two views are inconsistent. I just accept that they are and leave it at that. -- Matt Mahoney, [EMAIL PROTECTED] --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] Ethics of computer-based cognitive experimentation
Matt Mahoney wrote: snip ... accepted because the theory makes predictions that can be tested. But there are absolutely no testable predictions that can be made from a theory of consciousness. -- Matt Mahoney, [EMAIL PROTECTED] This is simply wrong. It is difficult but you can test for it objectively by demanding that an entity based on your 'theory of consciousness' deliver an authentic scientific act on the a-priori unknown using visual experience for scientific evidence. To the best _indirect_ evidence we have, that act is critically dependent on the existence of a visual phenomenal field within the tested entity. Visual P-consciousness and scientific evidence are literal identities in that circumstance. Degrade visual experience...scientific outcome is disrupted. You can use this to actually discover the physics of qualia as follows: 1) Concoct your theory of consciousness. 2) Build a scientist with it with (amongst other necessities) visual phenomenal consciousness which you believe to be there because of your theory of consciousness. Only autonomous, embodied entities are valid, because it involved actually interacting with an environment the way humans do. 3) Test it for delivery of an authentic act of science on the a-priori unknown by testing for ignorance at the start followed by the acquisition of the requisite knowledge followed by the application of the knowledge on a completely novel problem. 4) FAIL: = your physics is wrong or your design is bad. PASS = design and physics are good. REPEAT THE ABOVE for all putative physics END when you get success...voila...the physics you dreamt up is the right one or as good as the right one. If the entity delivers the 'law of nature' then it has to have all the essential aspects of a visual experience needed for a successful scientific act. You can argue about the 'experience' within the entity afterwards...on a properly informed basis of real knowledge. Until then you're just waffling about theories. Such a test might involve reward through reverse-engineering chess. Initially chess ignorance is demonstrated...followed by repeated exposure to chess behaviour on a real board.followed by a demand to use chess behaviour in a completely environment and in a different manner...say to operate a machine that has nothing to do with chess but is metaphorically labelled to signal that chess rules apply to some aspect of its behaviour This proves that the laws underneath the external behaviour of the original chess pieces was internalised and abstracted...which contains all the essential ingredients of a scientific act on the unknown. You cannot do this without authentic connection to the distal external world of the chess pieces. You cannot train such an entity. The scientific act itself is the training. Neither testers nor tested can have any knowledge of the 'law of nature' or the environments to be encountered. A completely novel 'game' could be substituted for chess, for example. Any entity dependent on any sort of training will fail. You can't train for scientific outcomes. You can only build the necessities of scientific behaviour and then let it loose. You run this test on all putative theories of consciousness. If you can;'t build it you have no theory. If you build it and it fails, tough. If you build it and it passes your theory is right. You can't test for consciousness is a cultural catch phrase identical to man cannot fly. Just like the Wright Bros, we need to start to fly. Not pretend to fly. Or not fly and say we did Objective testing for consciousness is easy. Building the test and the entity...well that's not so easy but it is possible. A 'definition' of consciousness is irrelevant. Like every other circumstance in science...'laws' and physical phenomena that operate according to them are discovered, not defined. Humans did not wait for a definition of fire before cooking dinner with it. Why should consciousness be any different? cheers colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] Ethics of computer-based cognitive experimentation
Matt Mahoney wrote: --- On Wed, 11/12/08, Colin Hales [EMAIL PROTECTED] wrote: It is difficult but you can test for it objectively by demanding that an entity based on your 'theory of consciousness' deliver an authentic scientific act on the a-priori unknown using visual experience for scientific evidence. So a blind person is not conscious? -- Matt Mahoney, [EMAIL PROTECTED] A blind person cannot behave scientifically in the manner of the sighted. The blind person cannot be a scientist 'of that which is visually evidenced'. As an objective test specifically for visual P-consciousness, the blind person's failure would prove the blind person has no visual P-consciousness. If a monkey passed the test then it would be proved visually P-conscious (as well as mighty smart!). A blind-sighted person would fail because they can't handle the radical novelty in the test. Again the test would prove they have no visual P-consciousness. A computer, if it passed, must have created inside itself all of the attributes of P-consciousness as utilised in vision applied to scientific evidence. You can argue about the details of any 'experience' only when armed with the physics _after_ the test is passed, when you can discuss the true nature of the physics involved from an authoritative position. If the requisite physics is missing the test subject will fail. That is the characteristic of a useful test. Unambiguous outcomes critically dependent on the presence of a claimed phenomenon. You don't even have to know the physics details. External behaviour is decisive and anyone could administer the test, provided it was set up properly. Note that experimental-scientists and applied scientists are literally scientific evidence of consciousness. They don't have to deliver anything except their normal science deliverables to complete the proof. They do nothing else but prove they are visually P-conscious for their entire lives. cheers, colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] Ethics of computer-based cognitive experimentation
[EMAIL PROTECTED] wrote: When people discuss the ethics of the treatment of artificial intelligent agents, it's almost always with the presumption that the key issue is the subjective level of suffering of the agent. This isn't the only possible consideration. One other consideration is our stance relative to that agent. Are we just acting in a selfish way, using the agent as simply a means to achieve our goals? I'll just leave that idea open as there are traditions that see value in de-emphasizing greed and personal acquisitiveness. Another consideration is the inherent value of self-determination. This is above any suffering that might be caused by being a completely controlled subject. One of the problems of slavery was just that it simply works better if you let people decide things for themselves. Similarly, just letting an artificial agent have autonomy for its own sake may just be a more effective thing than having it simply be a controlled subject. So I don't even think the consciousness of an artificial intelligent agent is completely necessary in considering the ethics of our stance towards it. We can consider our own emotional position and the inherent value of independence of thinking. andi I'm inclined to agree - this will be an issue in the future... if you have a robot helper and someone comes by and beats it to death in front of your kids, who have some kind of attachment to it...a relationship... then crime (i) may be said to be the psychological damage to the children. Crime (ii) is then the murder and whatever one knows of suffering inflicted on the robot helper. Ethicists are gonna have all manner of novelty to play with. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
Re: [agi] Ethics of computer-based cognitive experimentation
Matt Mahoney wrote: --- On Mon, 11/10/08, Richard Loosemore [EMAIL PROTECTED] wrote: Do you agree that there is no test to distinguish a conscious human from a philosophical zombie, thus no way to establish whether zombies exist? Disagree. What test would you use? The test will be published in the next couple of months in the Open AI journal. = An objective test for scientific behaviour. I call it the 'PCST' for P-Conscious Scientist Test. You can't be a scientist without being visually P-conscious to experience your evidence. You can't deny the test without declaring scientists devoid of consciousness whilst demanding it be used for all scientific evidence in a verifiable way AND whilst investing in an entire science paradigm Neural Correlates of Consciousness dedicated to scientific exploration of P-consciousnessThe logic's pretty good and it's easy to design an objective test demanding delivery of a 'law of nature'. The execution, however, is logistically difficult++. BUT...At least it's doable. A hard test is better than no test at all, which is what we currently have. When it comes out I'll let you know. RE ETHICS..I say this in the paper: As was recognised by Gamez [35], one cannot help but notice that there is also a secondary ethical 'bootstrap' process. Once a single subject passes the PCST, for the first time ever in certain circumstances there will be a valid scientific reason obliging all scientists to consider the internal life of an artefact as potentially having some level of equivalence to that of a laboratory animal, possibly deserving of similar ethical treatment. Until that event occurs, however, all such discussions are best considered moot. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Machine Consciousness Workshop, Hong Kong, June 2009
Hi, I was wondering as to the formatwho does what, how...speaking etc etc.. what sort of airing do the contributors get for their material? regards colin Ben Goertzel wrote: Hi all, I wanted to let you know that Gino Yu and I are co-organizing a Workshop on Machine Consciousness, which will be held in Hong Kong in June 2008: see http://novamente.net/machinecs/index.html for details. It is colocated with a larger, interdisciplinary conference on consciousness research, which has previously been announced: http://www.consciousness.arizona.edu/ As an aside, I also note that the date for submitting papers to AGI-09 has been extended, by popular demand, till November 12; see http://agi-09.org/ AGI-09 will welcome quality papers on any strong-AI related topics. thanks! ben -- Ben Goertzel, PhD CEO, Novamente LLC and Biomind LLC Director of Research, SIAI [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, solve equations, analyze a new problem, pitch manure, program a computer, cook a tasty meal, fight efficiently, die gallantly. Specialization is for insects. -- Robert Heinlein *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] META: A possible re-focusing of this list
Ben Goertzel wrote: Colin, There's a difference between 1) Discussing in detail how you're going to build a non-digital-computer based AGI 2) Presenting general, hand-wavy theoretical ideas as to why digital-computer-based AGI can't work I would be vastly more interested in 1 than 2 ... and I suspect many others on the list feel similarly... -- Ben G RE: (1) OK. I'll deposit (1) in nice easy to digestible slices as and when appropriate... if that can be slotted into your original (1) for discussion every now and then on the forum ...Alrighty then! RE:(2) Well I'm just a messenger from science and other logicians who are adding to an ever growing pile labelled cognition is not computation which has had yet more stuff added this year. It seems to be invalid from so many angles it's hard to keep up with...But the 3 main existing papers I have already cited: I believe them. I also have 2 of my own in review. They are based on existing physics and empirical work...no handwavy anything. I didn't reach the position lightly, because it makes the problem about a million times harder... OK. The message delivered. I can do no more than that. The bottom line: If I am wrong and COMP is true, we get AGI. If COMP is false and I am right, we get AGI. Sounds good to me! Let's leave it there. If minimal postings to the above (1) fit into your original (1) then that makes me feel that the forum is sidling up to a scientifically sound enthusiasm for AGI. Strength in diversity. Think about it. One day a real AGI is going to read this email forum squabbling away- if those involved have the ideas that work, this discussion will part of their personal history, a gestation of sorts, and I hope the AGI will see the work of caring parents on a well informed mission dedicated to their very genesis. That's how I'd want my mum(s) and dad(s) to be. :-) I'd rather be in that lineage than not. Wouldn't you? Far more interesting. Gotta go write the AGI-09 paper, amongst 46782318 other thingsI won't be back without (1)-style deliverables. cheers Colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Hi Trent, You guys are forcing me to voice all sort of things in odd ways. It's a hoot...but I'm running out of hours!!! Trent Waddington wrote: On Wed, Oct 15, 2008 at 4:48 PM, Colin Hales [EMAIL PROTECTED] wrote: you have to be exposed directly to all the actual novelty in the natural world, not the novelty recognised by a model of what novelty is. Consciousness (P-consciousness and specifically and importantly visual P-consciousness) is the mechanism by which novelty in the actual DISTANT natural world is made apparent to the agent. Symbolic grounding in Qualia NOT I/O. You do not get that information through your retina data. You get it from occipital visual P-consciousness. The Turing machine abstracts the mechanism of access to the distal natural world and hence has to be informed by a model, which you don't have... Wow. I know I don't know what P-consciousness is.. and clearly I must not no what Qualia is.. The capital must change the meaning from the normal definition. But basically I think you have to come out right now and say what your philosophy of reality is. Let me say right away that if you don't know what qualia or P-consciousness are then you're missing 150 years of biology and things are gonna look kind of odd. I suggest a rapid wiki-googling exercise (also in a recent post I delivered a whole pile of definitions and references.) I don't have a philosophy of reality. I exist, at a practical; level, within the confines of the standard particle model, 4 forces, 4 transmitter and the particle zoo. I don't need anything else to make a cogent case for my model that stacks up empirically the normal way. I do have a need to alter science, however, to become a dual aspect epistemology about a monism, entirely consistent with all existing science. Only the options of scientists changes and the structure of knowledge changes. In that case, the objective view I use has a very simple extension which accounts for subjectivity with physical, causal teeth. If your complaint is that a robot senses are not as rich or as complex as a human senses and therefore an AI hooked up to robot senses cannot possibly have the same qualia as humans then can you *stipulate for the sake of argument* that it may be possible to supply human senses to an AI so that it does have the same qualia? Or are you saying that there's some mystical magical thing about humans that makes it impossible for an AI to have the same qualia. And if you're not happy with the idea of an AI having the same qualia as humans, then surely you're willing to agree that a human that was born wired into solely robot senses (suppose its for humanitarian reasons, rather than just nazi doctors having fun if you like) would have fundamentally different qualia. You believe this human would not produce an original scientific act on the a-priori unknown - whatever that means - or does the fact that this evil human-robot hybrid is somehow half human give it a personal blessing from God? Trent I'm not complaining about anything! I am dealing with brute reality. You are simply unaware of the job that AGI faces...and are not aware of the 150 years of physiological evidence that the periphery (peripheral nervous system and periphery of the central nervous system like retina) is not 'experienced'. None of it. I have already been through this in my original posting, I think. IO signals (human and robot) _are not perceived_, generate no sensations i.e. are Qualia-Null. Experience happens in the cranial central nervous system, and is merely projected as-if it comes from the periphery. If feels like you have vision centered on your eyes, yes? Well surprise..all an illusion. Vision happens in the back of your head and is projected to appear as if your eyes generated it. You need to get a hold of some basic physiology. So the surprise for everyone who's been operating under the assumption that symbol grounding is simply I/O wiring: WRONG. We are symbolically grounded in qualia: something that happens in the cranial CNS. Not even the spinal CNS does any sensations. Pain in your back anyone? WRONG. The pain comes from cortex, NOT your spine. It's projected and mostly badly. As you must know from my postings...qualia are absolutely mandatory for handling novelty for a whole pile of complex reasons. And robots will need them too. But they will not have them from simply wiring up I/O signals and manipulating abstractions. You need the equivalent of the complete CRANIAL central nervous system electrodynamics to achieve that, not a model of it. So I demand that robots have qualia. For good physical, sensible, verifiable reasons...Whether they are exactly like humans..is another question. A human with artificial but equivalent peripheral sensory transduction would have qualia because the CNS generates them, not because are delivered by the I/O. And that human would be able to do
Re: [agi] Advocacy Is no Excuse for Exaggeration
Oops I forgot... Ben Goertzel wrote: About self: you don't like Metzinger's neurophilosophy I presume? (Being No One is a masterwork in my view) I got the book out and started to read it. But I found it incredibly dense and practically useless. It told me nothing. I came out the other end with no clarity whatever. Just a whole pile of self-referential jargon I couldn't build. No information gained. Maybe in time it'll become more meaningful and useful. It changed nothing. I expected a whole lot more. It was kind of like Finnegan's Wake. You can read it or have a gin and tonic and hit yourself over the head with it. The result is pretty much the same. :-) colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Consciousness
John LaMuth wrote: Colin Consc. by nature is subjective ... Can never prove this in a machine -- or other human beings for that matter Yes you can. This is a fallacy. You can prove it in humans and you can prove it in a machine. You simply demand it do science. Not simple - but possible. I have written this up. Should be published this year? Not sure. We are underutilizing about 4 Billion + human Cons's on the earth today What goal -- besides vanity -- is there to simulate this mechanically ?? We need to simulate communication, if anything at all ... John L The practical mission for me is not human-level AGI. I aim merely for A-fauna that is adaptive like biology is adaptive. I target markets where natural fauna struggles or where normal solutions are damaging. The rationale for creating human level AGI will be decided by forces yet to be determined. Relentless, tireless home help for the aged might be one case. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: Hi, My main impression of the AGI-08 forum was one of over-dominance by singularity-obsessed and COMP thinking, which must have freaked me out a bit. This again is completely off-base ;-) I also found my feeling about -08 as slightly coloured by first hand experience from an attendee who came away with the impression I put. I'll try and bolt down my paranioa at tad... COMP, yes ... Singularity, no. The Singularity was not a theme of AGI-08 and the vast majority of participant researchers are not seriously into Singularitarianism, futurism, and so forth. Good, although I'll be vigorously adding non-COMP approaches to the mix, and trusting that is OK There was a post-conference workshop on the Future of AGI, which about half of the conference attendees attended, at which the Singularity and related issues were discussed, among other issues. For instance, the opening talk at the workshop was given by Natasha Vita-More, who so far as I know is not a Singularitarian per se, though an excellent futurist. And one of the more vocal folks in the discussions in the workshop was Selmer Bringsjord, who believes COMP is false and has a different theory of intelligence than you or me, tied into his interest in Christian philosophy. The only reason for not connecting consciousness with AGI is a situation where one can see no mechanism or role for it. Seeing a mechanism or role for consciousness requires a specific theory of consciousness that not everybody holds --- and as you surely know, not even everyone in the machine consciousness community holds. Personally I view the first-person, second-person and third-person views as different perspectives on the universe, so I think it's a category error to talk about mechanisms of consciousness ... though one can talk about mechanisms that are correlated with particularly intense consciousness, for example. See my presentation from the Nokia workshop on Machine Consciousness in August ... where I was the only admitted panpsychist ;-) http://goertzel.org/NokiaMachineConsciousness.ppt ouch 10MB safely squirreled away under GforGoertzel, thank goodness for the uni bandwidth.. :-) I think I rest my case. You cannot see a physical mechanism or a role. I can. Inventing/adopting a whole mental rationale that avoids the problem based on an assumption about a 'received view' is not something I can do...I have a real physical process I can point to objectively, and a perspective from which it makes perfect sense that it be responsible for a first person perspective of the kind we receive.and I can't/won't talk it away just because 'Ben said so', even when the 'category error' stick, is wielded. That old rubric excuse for an argument doesn't scare me a bit ... :-) Consciousness is a problem for a reason, and that reason is mostly us thinking our 'categories' are right. Interestingly, my model, if you stand back and squint a bit, can be interpreted as having an 'as-if pan-psychism was real' appearance. Only an appearance tho. It's not real. Anyway... let's just let my story unfold, eh? It's a big one, so it'll take a while. Fun to be had! Thanks for the 'Hidden Pattern' link... I shall digest it. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: OK, but you have not yet explained what your theory of consciousness is, nor what the physical mechanism nor role for consciousness that you propose is ... you've just alluded obscurely to these things. So it's hard to react except with raised eyebrows and skepticism!! ben g Of course... that's only to be expected at this stage. It can't be helped. The physical mechanism is easy: quantum electrodynamics. The tricky bit is the perspective from which it can be held accountable for subjective qualities. ahem!... this is my theory... ahem!... :-) A. Elk... --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Hi Terren, They are not 'communities' in the sense that you mean. They are labs in various institutions that work on M/C-consciousness (or pretend to be doing cog sci, whilst actually doing it :-). All I can do is point you at the various references in the paper and get you to keep an eye on them. Not terribly satisfactory, but...well that's the way it is. It is why I was quite interested in the AGI forum...it's a potential nexus for the whole lot of us. regards Colin Terren Suydam wrote: Hi Colin, Are there other forums or email lists associated with some of the other AI communities you mention? I've looked briefly but in vain ... would appreciate any helpful pointers. Thanks, Terren --- On *Tue, 10/14/08, Colin Hales /[EMAIL PROTECTED]/* wrote: From: Colin Hales [EMAIL PROTECTED] Subject: Re: [agi] Advocacy Is no Excuse for Exaggeration To: agi@v2.listbox.com Date: Tuesday, October 14, 2008, 12:43 AM Hi Matt, ... The Gamez paper situation is now...erm...resolved. You are right: the paper doesn't argue that solving consciousness is necessary for AGI. What has happened recently is a subtle shift - those involved simple fail to make claims about the consciousness or otherwise of the machines! This does not entail that they are not actually working on it. They are just being cautious...Also, you correctly observe that solving AGI on a purely computational basis is not prohibited by the workers involved in the GAMEZ paper.. indeed most of their work assumes it!... I don't have a problem with this...However...'attributing' consciousness to it based on its behavior is probably about as unscientific as it gets. That outcome betrays no understanding whatever of consciousness, its mechanism or its roleand merely assumes COMP is true and creates an agreement based on ignorance. This is fatally flawed non-science. [BTW: We need an objective test (I have one - I am waiting for it to get published...). I'm going to try and see where it's at in that process. If my test is acceptable then I predict all COMP entrants will fail, but I'll accept whatever happens... - and external behaviour is decisive. Bear with me a while till I get it sorted.] I am still getting to know the folks [EMAIL PROTECTED] And the group may be diverse, as you say ... but if they are all COMP, then that diversity is like a group dedicated to an unresolved argument over the colour of a fish's bicycle. If we can attract the attention of the likes of those in the GAMEZ paper... and others such as Hynna and Boahen at Stanford, who have an unusual hardware neural architecture...(Hynna, K. M. and Boahen, K. 'Thermodynamically equivalent silicon models of voltage-dependent ion channels', /Neural Computation/ vol. 19, no. 2, 2007. 327-350.) ...and others ... then things will be diverse and authoritative. In particular, those who have recently essentially squashed the computational theories of mind from a neuroscience perspective- the 'integrative neuroscientists': Poznanski, R. R., Biophysical neural networks : foundations of integrative neuroscience, Mary Ann Liebert, Larchmont, NY, 2001, pp. viii, 503 p. Pomerantz, J. R., Topics in integrative neuroscience : from cells to cognition, Cambridge University Press, Cambridge, UK ; New York, 2008, pp. xix, 427 p. Gordon, E., Ed. (2000). Integrative neuroscience : bringing together biological, psychological and clinical models of the human brain. Amsterdam, Harwood Academic. The only working, known model of general intelligence is the human. If we base AGI on anything that fails to account scientifically and completely for /all/ aspects of human cognition, including consciousness, then we open ourselves to critical inferiority... and the rest of science will simply find the group an irrelevant cultish backwater. Strategically the group would do well to make choices that attract the attention of the 'machine consciousness' crowd - they are directly linked to neuroscience via cog sci. The crowd that runs with JETAI (journal of theoretical and experimental artificial intelligence) is also another relevant one. It'd be nice if those people also saw the AGI journal as a viable repository for their output. I for one will try and help in that regard. Time will tell I suppose. cheers, colin hales Matt Mahoney wrote: --- On Mon, 10/13/08, Colin Hales [EMAIL PROTECTED] wrote: In the wider world of science it is the current state of play that the theoretical basis for real AGI is an open and multi-disciplinary question. A forum that purports to be invested in achievement of real AGI as a target, one would expect that forum to a multidisciplianry approach on many fronts, all competing
Re: [agi] Advocacy Is no Excuse for Exaggeration
' is an explanatory pariah, the needed fundamentals are missing from the toolkit...I don't expect any sense to come from anywhere. An entity with a P-conscious (occipital/visual scene) projected depiction of the external world automatically places a self (the projector) inside it and 'self' becomes the same as everything else: something about which knowledge is accrued, from which behaviour may emerge. There are heaps of papers on the self. I read them, but they tell me nothing I can build. D Free Will. An interest of mine. I noted some reference that suggested a neuroscientific attempt to explain this (or perhaps explain it away). Know any more about this? Free Will and Free Won't(my favourite!) are high level aspects which I don't have a clear story on just yet. I am focussed on the lower levels entirely. When that is consolidated I will have something cogent to say. I'd rather study it later empirically with the chips I want to build. The motivation for my AGI to do anything at all remains problematic...I have my ideas but it's early days...FW is an important idea, but I can't explicitly 'build it', so it's not an early design issue. As with most other aspects of cognition I suspect that FW is a high-level (organism level) emergent property which has its ultimate basis in quantum mechanical randomness/indeterminacy. My chip architecture will incorporate the entire causal chain, thus inheriting the same indeterminacy, so at this stage theres nothing much more for me to add. One day. - - - - - - - -- - - - - Not terribly satisfying. I know. There's no quick route through the information. The only guide I can give is that there is a 'trump card' approach that clears nomothetic dross like a hot blade through butter: /Base your AGI on an artificial scientist model/. The clarity that emerges is stunning, and it's all empirically testable. regards, Colin Hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote:
Again, when you say that these neuroscience theories have squashed
the computational theories of mind, it is not clear to me what you
mean by the computational theories of mind. Do you have a more
precise definition of what you mean?
I suppose it's a bit ambiguous. There's computer modelling of mind, and
then there's the implementation of an actual mind using actual
computation, then there's the implementation of a brain using
computation, in which a mind may be said to be operating. All sorts of
misdirection.
I mean it in the sense given in:
Pylyshyn, Z. W., Computation and cognition : toward a foundation for
cognitive science, MIT Press, Cambridge, Mass., 1984, pp. xxiii, 292 p.
That is, that a mind is a result of a brain-as-computation. Where
computation is meant in the sense of abstract symbol manipulation
according to rules. 'Rules' means any logic or calculii you'd care to
cite, including any formally specified probablistic/stochastic language.
This is exactly what I mean by COMP.
Another slant on it:
Poznanski, R. R., Biophysical neural networks : foundations of
integrative neuroscience, Mary Ann Liebert, Larchmont, NY, 2001, pp.
viii, 503 p.
The literature has highlighted the conceptual ineptness of the computer
metaphor of the brain. Computational neuroscience, which serves as a
beacon for for the transfer of concepts regarding brain function to
artificial nets for the design of neural computers, ignores the
developmental theory of neuronal group selection and therefore seriously
overestimates the computational nature of neuroscience. It attempts to
explain brain function in terms of the abstract computational and
information processing functions thought to be carried out in the brain
{citations omitted}.
I don't know whether this answers your question,I hope so... it
means that leaping to a 'brain = computation in the digital computer
sense, is not what is going on. It also means that a computer model of
the full structure is also out. You have to do what the brain does, not
run a model of it. The brain is a electrodynamic entity, so your AGI has
to be an electrodynamic entity manipulating natural electromagnetic
symbols in a similar fashion. The 'symbols' are aggregate in the cohorts
mentioned by Poznanski. The electrodynamics itself IS the 'computation'
which occurs naturally in the trajectory through in the multidimensional
vector space of the matter as a whole. Some symbols are experienced
(qualia) and some are not.
cheers
colin
.
---
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription:
https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34
Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: Sure, I know Pylyshyn's work ... and I know very few contemporary AI scientists who adopt a strong symbol-manipulation-focused view of cognition like Fodor, Pylyshyn and so forth. That perspective is rather dated by now... But when you say Where computation is meant in the sense of abstract symbol manipulation according to rules. 'Rules' means any logic or calculii you'd care to cite, including any formally specified probablistic/stochastic language. This is exactly what I mean by COMP. then things get very very confusing to me. Do you include a formal neural net model as computation? How about a cellular automaton simulation of QED? Why is this cellular automaton model not abstract symbol manipulation? If you interpret COMP to mean A human-level intelligence can be implemented on a digital computer or as A human level intelligence can be implemented on a digital computer connected to a robot body or even as A human level intelligence, conscious in the same sense that humans are, can be implemented on a digital computer connected to a robot body ... then I'll understand you. We're really at cross-purposes here, aren't we?...this is a Colin/Ben calibration process :-) OK. By COMP I mean any abstract symbol manipulation at all in any context. The important thing is that in COMP there's a model of some kind of learning mechanism being run by a language of some kind or a model of a modelling process implemented programmatically. In any event the manipulations that are occuring are manipulations of abstract representation of numbers according to the language and the model being implemented by the computer language. But when you start defining COMP in a fuzzy, nebulous way, dismissing some dynamical systems as too symbolic for your taste (say, probabilistic logic) and accepting others as subsymbolic enough (say, CA simulations of QED) ... then I start to feel very confused... I agree that Fodor and Pylyshyn's approaches, for instance, were too focused on abstract reasoning and not enough on experiential learning and grounding. But I don't think this makes their approaches **more computational** than a CA model of QED ... it just makes them **bad computational models of cognition** ... Maybe a rather stark non-COMP example would help: I would term non-COMP approach is /there is no 'model' of cognition being run by anything./ The electrodynamics of the matter itself /is the cognition/. Literally. No imposed abstract model tells it how to learn. No imposed model is populated with any imposed knowledge. No human involvement in any of it except construction. Electrodynamic representational objects are being manipulated by real natural electrodynamics... is all there is. The 'computation', if you can call it that, is literally maxwell's equations (embedded on a QM substrate, of course) doing their natural dynamics dance in real matter, not an abstraction of maxwell's equations being run on a computer In my AGI I have no 'model' of anything. I have the actual thing. A bad model of cognition, to me, is identical to a poor understanding of what the brain is actually doing. With a good understanding of brain function you then actually run the real thing, not a model of it. The trajectory of a model of the electrodynamics cannot be the trajectory of the real electrodynamics. for the fields inherit behavioural/dynamical properties from the deep structure of matter, which are thrown away by the model of the electrodynamics. The real electrodynamics is surrounded by the matter it is situated in, and operates in accordance with it. Remember: A scientific model of a natural process cuts a layer across the matter hierarchy and throws away all the underlying structure. I am putting the entire natural hierarchy back into the picture by using real electrodyamics implemented in the fashion of a real brain, not a model of the electrodynamics of a real brain or any other abstraction of apparent brain operation. Does that do it? It's very very different to a COMP approach. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: About self: you don't like Metzinger's neurophilosophy I presume? (Being No One is a masterwork in my view) I agree that integrative biology is the way to go for understanding brain function ... and I was talking to Walter Freeman about his work in the early 90's when we both showed up at the Society for Chaos Theory in Psychology conferences ... however, I am wholly unconvinced that this work implies anything about the noncomputationality of consciousness. You mention QED, and I note that the only functions computable according to QED are the Turing-computable ones. I wonder how you square this with your view of QED-based brain dynamics as noncomputable? Or do you follow the Penrose path and posit as-yet-undiscovered, mysteriously-noncomputable quantum-gravity phenomena in brain dynamics (which, I note, requires not only radical unknown neuroscience but also radical unknown physics and mathematics) -- Ben G The comment is of the kind when did you stop kicking your dog. You assume that dog kicking was an issue and any answer in some way verifies/validates my involvement in dog-kicking! No way! :-) Turing computable or Xthing-computable...is irrelevant. I am not 'following' anyone except the example of the natural world.There's no inventions of mysterious anything... this is in-your-face good old natural matter doing what it does. I have spent an entire career being beaten to a pulp by the natural world of electromagnetismThis is really really simple. Nature managed to make a human capable of arguing about Turing computability and Godellian incompleteness without any 'functions' or abstractions or any 'model' of anything! I am following the same natural path of actual biology and real electrodynamics of real matter. I have a brilliant working prototype: /the human brain/. I am implementing the minimal subset of what it actually does, not a model of what it does. I have the skills to make an inorganic version of it. I don't need the ATP cycle, the full endocrine or inflammatory response and/or other immunochemistry systems or any of the genetic overheads. All the self-configuration and adaptation/tuning is easy to replicate in hardware. When you delete all those overheads what's left is really simple. Hooking it to I/O is easy - been doing it for decades... Of course - like a good little engineer I am scoping out electromagnetic effects using computational models. Computational chemistry, in fact. Appalling stuff! However, as a result my understanding of the electromagnetics of brain material will improve. That will result in appropriately engineered real electromagnetics running in my AGI, not a model of electromagnetics running in my AGI. Quantum mechanics will be doing its bit without me lifting a finger - because i am using natural matter as it is used in brain material. Brilliant tho it was, and as beautiful a piece of science that it was, Hodgkin and Huxley threw out the fields in 1952ish and there they languish, ignored until now. Putting back in the 50% that was thrown away 50 years ago can hardly be considered 'radical' neuroscience. Ignoring it for any more than 50 years when you can show it operating there for everyone to see...now that'd be radically stupid in anyone's book. There's also a clinical side: the electrodynamics/field structure can be used in explanation of developmental chemistry/cellular transport cues and it also sorts out the actual origins of EEG, both of which are currently open problems. It's a little brain-bending to get your head around.. but it'll sink in. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: I still don't really get it, sorry... ;-( Are you saying A) that a conscious, human-level AI **can** be implemented on an ordinary Turing machine, hooked up to a robot body or B) A is false B) Yeah that about does it. Specifically: It will never produce an original scientific act on the a-priori unknown. It is the unknown bit which is important. You can't deliver a 'model' of the unknown that delivers all of the aspects of the unknown without knowing it all already!catch 22...you have to be exposed /directly/ to all the actual novelty in the natural world, not the novelty recognised by a model of what novelty is. Consciousness (P-consciousness and specifically and importantly visual P-consciousness) is the mechanism by which novelty in the actual DISTANT natural world is made apparent to the agent. Symbolic grounding in Qualia NOT I/O. You do not get that information through your retina data. You get it from occipital visual P-consciousness. The Turing machine abstracts the mechanism of access to the distal natural world and hence has to be informed by a model, which you don't have... Because scientific behaviour is just a (formal, very testable) refinement of everyday intelligent behaviour, everyday intelligent behaviour of the kind humans have - goes down the drain with it. With the TM precluded from producing a scientist, it is precluded as a mechanism for AGI. I like scientific behaviour. A great clarifier. cheers colin --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: COMP = false? (was Re: [agi] Advocacy Is no Excuse for Exaggeration)
Matt Mahoney wrote: --- On Tue, 10/14/08, Colin Hales [EMAIL PROTECTED] wrote: The only reason for not connecting consciousness with AGI is a situation where one can see no mechanism or role for it. That inability is no proof there is noneand I have both to the point of having a patent in progress. Yes, I know it's only my claim at the moment...but it's behind why I believe the links to machine consciousness are not optional, despite the cultural state/history of the field at the moment being less than perfect and folks cautiously sidling around consciousness like it was bomb under their budgets. Colin, I read your paper in publication that you were so kind to send me. For those who have not seen it, it is a well written, comprehensive survey of research in machine consciousness. It does not take a position on whether consciousness plays an essential role in AGI. (I understand that taking a controversial position probably would have resulted in rejection). With regard to COMP, I assume you define COMP to be the position that everything the mind does is, in principle, computable. If I understand your position, consciousness does play a critical role in AGI. However, we don't know what it is. Therefore we need to find out by using scientific research, then duplicate that process (if possible) in a machine before it can achieve AGI. Here and in your paper, you have not defined what consciousness is. Most philosophical arguments can be traced to disagreements about the meanings of words. In your paper you say that consciousness means having phenomenal states, but you don't define what a phenomenal state is. Without a definition, we default to what we think it means. Everybody knows what consciousness is. It is something that all living humans have. We associate consciousness with properties of humans, such as having a name, a face, emotions, the ability to communicate in natural language, the ability to learn, to behave in ways we expect people to behave, to look like a human. Thus, we ascribe partial degrees of consciousness (with appropriate ethical treatment) to animals, video game characters, human shaped robots, and teddy bears. To argue your position, you need to nail down a definition of consciousness. But that is hard. For example, you could define consciousness as having goals. So if a dog wants to go for a walk, it is conscious. But then a thermostat wants to keep the room at a set temperature, and a linear regression algorithm wants to find the best straight line fit to a set of points. You could define consciousness as the ability to experience pleasure and pain. But then you need a test to distinguish experience from mere reaction, or else I could argue that simple reinforcement learners like http://www.mattmahoney.net/autobliss.txt experience pain. It boils down to how you define experience. You could define consciousness as being aware of your own thoughts. But again, you must define aware. We distinguish conscious or episodic memories, such as when I recalled yesterday something that happened last month, and unconscious or procedural memories, such as the learned skills in coordinating my leg muscles while walking. We can do studies to show that conscious memories are stored in the hippocampus and higher layers of the cerebral cortex, and unconscious memories are stored in the cerebellum. But that is not really helpful for AGI design. The important distinction is that we remember remembering conscious memories but not unconscious. Reading from conscious memory also writes into it. But I can simulate this process in simple programs, for example, a database that logs transactions. So if you can nail down a definition of consciousness without pointing to a human, I am willing to listen. Otherwise we default to the possibility of building AGI on COMP principles and then ascribing consciousness to it since it behaves just like a human. -- Matt Mahoney, [EMAIL PROTECTED] I am way past merely defining anything. I know what phenomenal fields are constructed of: Virtual Nambu Goldstone Bosons. Brain material is best regarded as a radically anisotropic quasi-fluid undergoing massive phase changes on multiple time scales. The problem is one of thermodynamics, not abstract computation. Duplicating the boson generation inorganically and applying that process to regulatory mechanisms of learning is exactly what I plan for my AGI chips. The virtual particles were named Qualeons by some weird guy here that i was talking to one day. I forgot is name. I better find that out! I digress. :-) It would take 3 PhD dissertations to cover everything from quantum mechanics to psychology. You have to be a polymath. And to see how they explain consciousness you need to internalise 'dual aspect science', from which perspective its all obvious. I have to change the whole of science from single to dual aspect to make it understood
Re: [agi] Advocacy Is no Excuse for Exaggeration
Jim Bromer wrote: On Mon, Oct 13, 2008 at 2:34 PM, Charles Hixson [EMAIL PROTECTED] wrote: Galileo, Bruno of Nolan, etc. OTOH, Paracelsus was quite personable. So was, reputedly, Pythagoras. (No good evidence on Pythagoras, though. Only stories from supporters.) (Also, consider that the Pythagoreans, possibly including Pythagoras, had a guy put to death for discovering that sqrt(2) was irrational. [As with most things from this date, this is more legend than fact, but is quite probable.]) As a generality, with many exceptions, strongly opinionated persons are not easy to get along with unless you agree with their opinions. It appears to be irrelevant whether their opinions are right, wrong, or undecidable. I just want to comment that my original post was not about agreeableness. It was about the necessity of being capable of criticizing your own theories (and criticisms). I just do not believe that Newton, Galileo, Pythagoras and the rest of them were incapable of examining their own theories from critical vantage points even though they may have not accepted the criticisms others derived from different vantage points. As I said, there is no automatic equality for criticisms. Just because a theory is unproven it does not mean that all criticisms have to be accepted as equally valid. But when you see someone, theorist or critic, who almost never demonstrates any genuine capacity for reexamining his own theories or criticisms from any critical vantage point what so ever, then it's a strong negative indicator. Jim Bromer The process of formulation of scientific theories has been characterised as a dynamical system nicely by Nicholas Rescher. Rescher, N., Process philosophy : a survey of basic issues, University of Pittsburgh Press, Pittsburgh, 2000, p. 144. Rescher, N., Nature and understanding : the metaphysics and method of science, Clarendon Press, Oxford, 2000, pp. ix, 186. In that approach you can see critical argument operating operating as a brain process - competing brain electrodynamics that stabilises on the temporary 'winner', whose position may be toppled at any moment by the emergence of a more powerful criticism which destabilises the current equilibrium...and so on. The 'argument' may involve the provision of empirical evidence ... indeed that is the norm for most sciences. In order that a discipline be seen to be real science, then, what one would expect to see such processes happening in a dialog between a diversity of views competing for ownership of scientific evidence through support for whatever theoretical framework seems apt. As a recent entrant here, and seeing the dialog and the issues as they unfold I would have some difficulty classifying what is going on as 'scientific' in the sense that there is no debate calibrated against any overt fundamental scientific theoretical framework(s), nor defined testing protocols. In the wider world of science it is the current state of play that the theoretical basis for real AGI is an open and multi-disciplinary question. A forum that purports to be invested in achievement of real AGI as a target, one would expect that forum to a multidisciplianry approach on many fronts, all competing scientifically for access to real AGI. I am not seeing that here. In having a completely different approach to AGI, I hope I can contribute to the diversity of ideas and bring the discourse closer to that of a solid scientific discipline, with formal testing metrics and so forth. I hope that I can attract the attention of the neuroscience and physics world to this area. Of course whether I'm an intransigent grumpy theory-zealot of the Newtonian kind... well... just let the ideas speak for themselves... :-) The main thing is the diversity of ideas and criticism .. which seems a little impoverished at the moment. Without the diversity of approaches actively seen to compete, an AGI forum will end up marginalised as a club of some kind: We do (what we assume will be) AGI by fiddling about with XYZ. This is scientific suicide. Here's a start:: the latest survey in the key area. Like it or not AGI is directly in the running for solving the 'hard problem' and machine consciousness is where the game is at. Gamez, D. 'Progress in machine consciousness', Consciousness and Cognition vol. 17, no. 3, 2008. 887-910. I'll do my best to diversify the discourse... I'd like to see this community originate real AGI and be seen as real science. To do that this forum should attract cognitive scientists, psychologists, physicists, engineers, neuroscientists. Over time, maybe we can get that sort of diversity happening. I have enthusiasm for such things.. cheers colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
to finding out about the reality of human cognition and thereby gain access to AGI. Theories and inventions are two sides of the same science process. The former looks at natural outcomes seeks rules that describe initial conditions. The latter seeks initial conditions that result in a certain outcome. Being precious about failure or critique is not part of the process. At all stages choices are made and results must be critically assessed. As a nuance to the role of 'science' in 'computer science'... I would say that formally, computer 'science' is not scientific for the reason that software documents are not laws of nature the resultant machine behaviour originates from causality built into the physics of the hardware substrate. At best, computer programs are correlates, not the sought-after critical dependency which we know gets us as close to natural causality as we can get. The computer substrate, configured as per the program, causally necessitates the resultant behaviour, not the program whic, at runtime, is completely absent from the circumstance, even in 'interpretive' runtime environments. So the role of scientific behaviour in the production of an AGI seems to be in need of quite a deal of review - especially if it is to be taken seriously in a multidisciplinary approach to AGI where all the folks are used to criticism and expect it (demand it!), and are OK with being wrong. The flip side of this in technological development is that when something has a mounting pile of evidence of being based on a false premise, that continuing to fail to make choices that direct your efforts in other directions now that is something to be embarrassed about :-) I understand how tied up people can get with a particular standpoint - if you have defended it for 15 years and suddenly the whole basis of it is gone - for obvious empirical reasons that you cannot deny - strange behaviours result for example... my supervisor one had a paper rejected thus: This paper does not show that (SUCH and SUCH) is the case in an auditory cortex context and should be rejected until (SUCH and SUCH) is shown to be the case when the argument was based on empirical grounds and was quite sound! People can be so fashion-ridden and flighty and precious about their 'darlings'the editor ended up apologising for the idiot reviewer...who was too stupid to roll with the punches and was marginalised as a result. In my case I am absolutely determined to be a 1 trick scientist, like the guy who discovered the neutrino. I adopt one particular approach to AGI and *go for it - scientifically*. I either get to be the guy who actually solved it or I get to be an author who writes about failing to make my idea work. Either way is OK with me - at least i will have shown one way NOT to reach AGI, and have been seen to respond to evidence when I encounter it, not cover my eyes and ears and go blah blah blah blah. :-) The thing to be proud of is /authentic science/. Not always easy to do... but a noble goal. In my case i have a crucial experiment planned, which will sort out the basis for my new chip design. If that experiment fails - I will NOT continue with the original idea. I will have bet on a dud idea. But that's OK. Oh boy I'm blathering on again. So many words!~ Sorry! better go...I think you get my drift. cheers colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
Re: [agi] Advocacy Is no Excuse for Exaggeration
Hi Matt, ... The Gamez paper situation is now...erm...resolved. You are right: the paper doesn't argue that solving consciousness is necessary for AGI. What has happened recently is a subtle shift - those involved simple fail to make claims about the consciousness or otherwise of the machines! This does not entail that they are not actually working on it. They are just being cautious...Also, you correctly observe that solving AGI on a purely computational basis is not prohibited by the workers involved in the GAMEZ paper.. indeed most of their work assumes it!... I don't have a problem with this...However...'attributing' consciousness to it based on its behavior is probably about as unscientific as it gets. That outcome betrays no understanding whatever of consciousness, its mechanism or its roleand merely assumes COMP is true and creates an agreement based on ignorance. This is fatally flawed non-science. [BTW: We need an objective test (I have one - I am waiting for it to get published...). I'm going to try and see where it's at in that process. If my test is acceptable then I predict all COMP entrants will fail, but I'll accept whatever happens... - and external behaviour is decisive. Bear with me a while till I get it sorted.] I am still getting to know the folks [EMAIL PROTECTED] And the group may be diverse, as you say ... but if they are all COMP, then that diversity is like a group dedicated to an unresolved argument over the colour of a fish's bicycle. If we can attract the attention of the likes of those in the GAMEZ paper... and others such as Hynna and Boahen at Stanford, who have an unusual hardware neural architecture...(Hynna, K. M. and Boahen, K. 'Thermodynamically equivalent silicon models of voltage-dependent ion channels', /Neural Computation/ vol. 19, no. 2, 2007. 327-350.)...and others ... then things will be diverse and authoritative. In particular, those who have recently essentially squashed the computational theories of mind from a neuroscience perspective- the 'integrative neuroscientists': Poznanski, R. R., Biophysical neural networks : foundations of integrative neuroscience, Mary Ann Liebert, Larchmont, NY, 2001, pp. viii, 503 p. Pomerantz, J. R., Topics in integrative neuroscience : from cells to cognition, Cambridge University Press, Cambridge, UK ; New York, 2008, pp. xix, 427 p. Gordon, E., Ed. (2000). Integrative neuroscience : bringing together biological, psychological and clinical models of the human brain. Amsterdam, Harwood Academic. The only working, known model of general intelligence is the human. If we base AGI on anything that fails to account scientifically and completely for /all/ aspects of human cognition, including consciousness, then we open ourselves to critical inferiority... and the rest of science will simply find the group an irrelevant cultish backwater. Strategically the group would do well to make choices that attract the attention of the 'machine consciousness' crowd - they are directly linked to neuroscience via cog sci. The crowd that runs with JETAI (journal of theoretical and experimental artificial intelligence) is also another relevant one. It'd be nice if those people also saw the AGI journal as a viable repository for their output. I for one will try and help in that regard. Time will tell I suppose. cheers, colin hales Matt Mahoney wrote: --- On Mon, 10/13/08, Colin Hales [EMAIL PROTECTED] wrote: In the wider world of science it is the current state of play that the theoretical basis for real AGI is an open and multi-disciplinary question. A forum that purports to be invested in achievement of real AGI as a target, one would expect that forum to a multidisciplianry approach on many fronts, all competing scientifically for access to real AGI. I think this group is pretty diverse. No two people here can agree on how to build AGI. Gamez, D. 'Progress in machine consciousness', Consciousness and Cognition vol. 17, no. 3, 2008. 887-910. $31.50 from Science Direct. I could not find a free version. I don't understand why an author would not at least post their published papers on their personal website. It greatly increases the chance that their paper is cited. I understand some publications require you to give up your copyright including your right to post your own paper. I refuse to publish with them. (I don't know the copyright policy for Science Direct, but they are really milking the publish or perish mentality of academia. Apparently you pay to publish with them, and then they sell your paper). In any case, I understand you have a pending paper on machine consciousness. Perhaps you could make it available. I don't believe that consciousness is relevant to intelligence, but that the appearance of consciousness is. Perhaps you can refute my position. -- Matt Mahoney, [EMAIL PROTECTED] --- agi
Re: [agi] Advocacy Is no Excuse for Exaggeration
Ben Goertzel wrote: Colin wrote: The only working, known model of general intelligence is the human. If we base AGI on anything that fails to account scientifically and completely for /all/ aspects of human cognition, including consciousness, then we open ourselves to critical inferiority... and the rest of science will simply find the group an irrelevant cultish backwater. Strategically the group would do well to make choices that attract the attention of the 'machine consciousness' crowd - they are directly linked to neuroscience via cog sci. Actually, I very strongly disagree with the above. While I am an advocate of machine consciousness research, and will be co-organizing a machine consciousness workshop in Hong Kong in June 2009, I do **not** agree that focusing on machine consciousness would be likely to help AGI to get better accepted in the general scientific community. Rather, I think that consciousness research is currently considered at least as eccentric as AGI research, by the scientific mainstream ... and is considered far MORE eccentric than AGI research by the AI research mainstream, e.g. the AAAI. So, discussing issues of machine consciousness may be interesting and very worthwhile for AGI in some scientific and conceptual... but I really really don't think that, at the present time, more closely allying AGI with machine consciousness would do anything but cause trouble for AGI's overall scientific reputation. Frankly I think that machine consciousness has at least as high a chance of being considered an irrelevant cultish backwater than AGI ... though I don't think that either field deserves that fate. Comparing the two fields, I note that AGI has a larger and more active series of conferences than machine consciousness, and is also ... pathetic as it may be ... better-funded overall ;-p Regarding the connection to neuroscience and cog sci: obviously, AGI does not need machine consciousness as an intermediary to connect to those fields, it is already closely connected. As one among many examples, Stan Franklin's LIDA architecture, a leading AGI approach, was originated in collaboration with Bernard Baars, a leading cognitive psychologist (and consciousness theorist, as it happens). And we had a session on AGI and Neuroscience at AGI-08, chaired by neuroscientist Randal Koene. I laid out my own thoughts on consciousness in some detail in The Hidden Pattern ... I'm not trying to diss consciousness research at all ... just pointing out that the posited reason for tying it in with AGI seems not to be correct... -- Ben G My main impression of the AGI-08 forum was one of over-dominance by singularity-obsessed and COMP thinking, which must have freaked me out a bit. The IEEE Spectrum articles on the 'singularity rapture' did nought to improve my outlook... Thanks for bringing the Stan Franklin and Bernhard Baars/Global Workspace etc and neuroscience links to my attention. I am quite familiar with them and it's a relief to see they connect with the AGI fray. Hopefully the penetration of these disciplines, and their science, will grow. In respect of our general consciousness-in-AGI disagreement: Excellent! That disagreement is a sign of diversity of views. Bring it on! The only reason for not connecting consciousness with AGI is a situation where one can see no mechanism or role for it. That inability is no proof there is noneand I have both to the point of having a patent in progress. Yes, I know it's only my claim at the moment...but it's behind why I believe the links to machine consciousness are not optional, despite the cultural state/history of the field at the moment being less than perfect and folks cautiously sidling around consciousness like it was bomb under their budgets. So...You can count on me for vigorous defense of my position from quantum physics upwards to psychology, including support for machine consciousness as being on the critical path to AGI. Hopefully in June '09? ;-) I tried to locate a local copy of 'the hidden pattern'...no luck. Being in poverty-stricken student mode, at the moment...I have to survive on library/online resources, which are pretty impressive here at Unimelb..but despite this the libraries around here don't have it...two other titles in the state library... but not that one..Oh well. Maybe send me a copy with my wizard hat? :-P cheers, colin hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=117534816-b15a34 Powered by Listbox: http://www.listbox.com
OFFLIST [agi] Readings on evaluation of AGI systems
Hi Ben, A good bunch of papers. (1) Hales, C. 'An empirical framework for objective testing for P-consciousness in an artificial agent', The Open Artificial Intelligence Journal vol.? , 2008. Apparently it has been accepted but I'll believe it when I see it. It's highly relevant to the forum you mentioned. I was particularly interested in the Wray and Lebiere work... my paper (1) would hold that the problem in their statement Taskability is difficult to measure because there is no absolute notion of taskability -- a particular quantitative measure for one domain might represent the best one could achieve, while in another, it might be a baseline is solved. An incidental byproduct of the execution of the test is that all the other metrics in their paper are delivered to some extent. Computationalist AI subjects will fail the (1) test. Humans won't. A real AGI will pass. Testing has been a big issue for me and has taken quite a while to sort out. Peter Voss's AI will fail it. As will everything based on NUMENTA products.but they can try!.the test can speak for itself. Objective measurement of outward agent behaviour is decisive. All contenders have the same demands made of them...the only requirement is that verifiably autonomous, embodied agents only need apply. I don't know if this is of interest to anyone, but I thought I'd mention it. regards, Colin Ben Goertzel wrote: Hi all, In preparation for an upcoming (invitation-only, not-organized-by-me) workshop on Evaluation and Metrics for Human-Level AI systems, I concatenated a number of papers on the evaluation of AGI systems into a single PDF file (in which the readings are listed alphabetically in order of file name). In case anyone else finds this interesting, you can download the single PDF file from http://goertzel.org/AGI_Evaluation.pdf It's 196 pages of text I don't condone all the ideas in all the papers, nor necessarily consider all the papers incredibly fascinating ... but it's a decent sampling of historical thinking in the area by a certain subclass of AGI-ish academics... ben *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] COMP = false
Excellent. I want one! Maybe they should be on sale at the next conference...there's a marketing edge for ya. If I have to be as wrong as Vladimir says I'll need the right clothes. :-) cheers colin Ben Goertzel wrote: And you can't escape flaws in your reasoning by wearing a lab coat. Maybe not a lab coat... but how about my trusty wizard's hat??? ;-) http://i34.tinypic.com/14lmqg0.jpg *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modify https://www.listbox.com/member/?; Your Subscription [Powered by Listbox] http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] COMP = false
! This is supposed to be fun! cheers Colin Hales Ben Goertzel wrote: The argument seems wrong to me intuitively, but I'm hard-put to argue against it because the terms are so unclearly defined ... for instance I don't really know what you mean by a visual scene ... I can understand that to create a form of this argument worthy of being carefully debated, would be a lot more work than writing this summary email you've given. So, I agree with your judgment not to try to extensively debate the argument in its current sketchily presented form. If you do choose to present it carefully at some point, I encourage you to begin by carefully defining all the terms involved ... otherwise it's really not possible to counter-argue in a useful way ... thx ben g On Sat, Oct 4, 2008 at 12:31 AM, Colin Hales [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Hi Mike, I can give the highly abridged flow of the argument: !) It refutes COMP , where COMP = Turing machine-style abstract symbol manipulation. In particular the 'digital computer' as we know it. 2) The refutation happens in one highly specific circumstance. In being false in that circumstance it is false as a general claim. 3) The circumstances: If COMP is true then it should be able to implement an artificial scientist with the following faculties: (a) scientific behaviour (goal-delivery of a 'law of nature', an abstraction BEHIND the appearances of the distal natural world, not merely the report of what is there), (b) scientific observation based on the visual scene, (c) scientific behaviour in an encounter with radical novelty. (This is what humans do) The argument's empirical knowledge is: 1) The visual scene is visual phenomenal consciousness. A highly specified occipital lobe deliverable. 2) In the context of a scientific act, scientific evidence is 'contents of phenomenal consciousness'. You can't do science without it. In the context of this scientific act, visual P-consciousness and scientific evidence are identities. P-consciousness is necessary but on its own is not sufficient. Extra behaviours are needed, but these are a secondary consideration here. NOTE: Do not confuse scientific observation with the scientific measurement, which is a collection of causality located in the distal external natural world. (Scientific measurement is not the same thing as scientific evidence, in this context). The necessary feature of a visual scene is that it operate whilst faithfully inheriting the actual causality of the distal natural world. You cannot acquire a law of nature without this basic need being met. 3) Basic physics says that it is impossible for a brain to create a visual scene using only the inputs acquired by the peripheral stimulus received at the retina. This is due to fundamentals of quantum degeneracy. Basically there are an infinite number of distal external worlds that can deliver the exact same photon impact. The transduction that occurs in the retinal rod/cones is entirely a result of protein isomerisation. All information about distal origins is irretievably gone. An impacting photon could have come across the room or across the galaxy. There is no information about origins in the transduced data in the retina. That established, you are then faced with a paradox: (i) (3) says a visual scene is impossible. (ii) Yet the brain makes one. (iii) To make the scene some kind of access to distal spatial relations must be acquired as input data in addition to that from the retina. (iv) There are only 2 places that can come from... (a) via matter (which we already have - retinal impact at the boundary that is the agent periphery) (b) via space (at the boundary of the matter of the brain with space, the biggest boundary by far). So, the conclusion is that the brain MUST acquire the necessary data via the spatial boundary route. You don't have to know how. You just have no other choice. There is no third party in there to add the necessary data and the distal world is unknown. There is literally nowhere else for the data to come from. Matter and Space exhaust the list of options. (There is alway magical intervention ... but I leave that to the space cadets.) That's probably the main novelty for the reader to to encounter. But we are not done yet. Next empirical fact: (v) When you create a turing-COMP substrate the interface with space is completely destroyed and replaced with the randomised machinations of the matter of the computer manipulating a model of the distal world. All actual relationships with the real distal external world are destroyed. In that circumstance the COMP substrate is implementing the science of an encounter
Re: [agi] COMP = false
Hi Vladimir, I did not say the physics was unknown. I said that it must exist. The physics is already known.Empirically and theoretically. It's just not recognised in-situ and by the appropriate people. It's an implication of the quantum non-locality underpinning electrodynamics. Extensions of the physics model to include the necessary effects are not part of the discussion and change nothing. This does not alter the argument, which is empirical. Please accept and critique it on this basis. I am planning an experiment as a post-doc to validate the basic principle as it applies in a neural context. It's under development now. It involves electronics and lasers and all the usual experimental dross. BTW I don't do non-science. Otherwise I'd just be able to sit back and declare my world view complete and authoritative, regardless of the evidence, wouldn't I? That is so not me. I am an engineer If I can't build it then I know I don't understand it. Nothing is sacred. At no point ever will I entertain any fictional/untestable/magical solutions. Like assuming an unproven conjecture is true. Nor will I idolise the 'received view' as having all the answers and force the natural world to fit my prejudices in respect of what 'explanation' entails. Especially when major mysteries persist in the face of all explanatory attempts. That's the worst non-science you can have... so I'm rather more radically empirical and dry, evidenced based but realistic in expectations of our skills as explorers of the natural world ...than it might appear. In being this way I hope to be part of the solution, not part of the problem. COMP being false make the AGI goal much harder...but much much more interesting! That's a little intro to colin hales for you. cheers colin hales (all done now!) Vladimir Nesov wrote: Basically, you are saying that there is some unknown physics mojo going on. The mystery of mind looks as mysterious as mystery of physics, therefore it requires mystery of physics and can derive further mysteriousness from it, becoming inherently mysterious. It's bad, bad non-science. --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] COMP = false
OK. Last one! Please replace 2) with: 2. Science says that the information from the retina is insufficient to construct a visual scene. Whether or not that 'constuct' arises from computation is a matter of semantics. I would say that it could be considered computation - natural computation by electrodynamic manipulation of natural symbols. Not abstractions of the kind we manipulate in the COMP circumstance. That is why I use the term COMP... It's rather funny: you could redefine computation to include natural computation (through the natural causality that is electrodynamics as it happens in brain material). Then you could claim computationalism to be true. But you'd still behave the same: you'd be unable to get AGI from a Turing machine. So you'd flush all traditional computers and make new technology Computationalism would then be true but 100% useless as a design decision mechanism. Frankly I'd rather make AGI that works than be right according to a definition! The lesson is that there's no pracitcal use in being right according to a definition! What you need to be able to do is make successful choices. OK. Enough. A very enjoyable but sticky thread...I gotta work! cheers all for now. regards Colin Abram Demski wrote: Colin, I believe you did not reply to my points? Based on your definition of computationalism, it appears that my criticism of your argument does apply after all. To restate: Your argument appears to assume computationalism. Here is a numbered restatement: 1. We have a visual experience of the world. 2. Science says that the information from the retina is insufficient to compute one. 3. Therefore, we must get more information. 4. The only possible sources are material and spatial. 5. Material is already known to be insufficient, therefore we must also get spatial info. Computationalism is assumed to get from #2 to #3. If we do not assume computationalism, then the argument would look more like this: 1. We have a visual experience of the world. 2. Science says that the information from the retina is insufficient to compute one. 3. Therefore, our visual experience is not computed. This is obviously unsatisfying because it doesn't say where the visual scene comes from; answers range from prescience to quantum hypercomputation, but that does not seem important to the current issue. --Abram --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] COMP = false
Hi Will, It's not an easy thing to fully internalise the implications of quantum degeneracy. I find physicists and chemists have no trouble accepting it, but in the disciplines above that various levels of mental brick walls are in place. Unfortunately physicists and chemists aren't usually asked to create vision!... I inhabit an extreme multidisciplinary zone. This kind of mental resistance comes with the territory. All I can say is 'resistance is futile, you will be assimilated' ... eventually. :-) It's part of my job to enact the necessary advocacy. In respect of your comments I can offer the following: You are exactly right: humans don't encounter the world directly (naive realism). Nor are we entirely operating from a cartoon visual fantasy(naive solipsism). You are also exactly right in that vision is not 'perfect'. It has more than just a level of indirectness in representation, it can malfunction and be fooled - just as you say. In the benchmark behaviour: scientific behaviour, we know scientists have to enact procedures (all based around the behaviour called 'objectivity') which minimises the impact of these aspects of our scientific observation system. However, this has nothing to say about the need for an extra information source. necessary for there is not enough information in the signals to do the job. This is what you cannot see. It took me a long while to discard the tendency to project my mental capacity into the job the brain has when it encounters a retinal data stream. In vision processing using computing we know the structure of the distal natural world. We imagine the photon/CCD camera chip measurements to be the same as that of the retina. It looks like a simple reconstruction job. But it is not like that at all. It is impossible to tell, from the signals in their natural state in the brain, whether they are about vision or sound or smell. They all look the same. So I did not completely reveal the extent of the retinal impact/visual scene degeneracy in my post. The degeneracy operates on multiple levels. Signal encoding into standardised action potentials is another level. Maybe I can just paint a mental picture of the job the brain has to do. Imagine this: You have no phenomenal consciousness at all. Your internal life is of a dreamless sleep. Except ... for a new perceptual mode called Wision. Looming in front of you embedded in a roughly hemispherical blackness is a gigantic array of numbers. The numbers change. Now: a) make a visual scene out of it representing the world outside: convert Wision into Vision. b) do this without any information other than the numbers in front of you and without assuming you have any a-priori knowledge of the outside world. That is the job the brain has. Resist the attempt to project your own knowledge into the circumstance. You will find the attempt futile. Regards, Colin William Pearson wrote: Hi Colin, I'm not entirely sure that computers can implement consciousness. But I don't find your arguments sway me one way or the other. A brief reply follows. 2008/10/4 Colin Hales [EMAIL PROTECTED]: Next empirical fact: (v) When you create a turing-COMP substrate the interface with space is completely destroyed and replaced with the randomised machinations of the matter of the computer manipulating a model of the distal world. All actual relationships with the real distal external world are destroyed. In that circumstance the COMP substrate is implementing the science of an encounter with a model, not an encounter with the actual distal natural world. No amount of computation can make up for that loss, because you are in a circumstance of an intrinsically unknown distal natural world, (the novelty of an act of scientific observation). . But humans don't encounter the world directly, else optical illusions wouldn't exist, we would know exactly what was going on. Take this site for example. http://www.michaelbach.de/ot/ It is impossible by physics to do vision perfectly without extra information, but we do not do vision by any means perfectly, so I see no need to posit an extra information source. Will --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] Testing, and a question....
Dear AGI folk, I am testing my registration on the system,, saying an inaugural 'hi' and seeking guidance in respect of potential submissions for a presentation spot at the next AGI conference. It is time for me to become more visible in AGI after 5 years of research and reprogramming my brain into the academic side of things My plans as a post-doc are to develop a novel chip technology. It will form the basis for what I have called 'A-Fauna'. I call it A-Fauna because it will act like biological organisms and take their place alongside natural fauna in a chosen ecological niche. Like tending a field as a benign 'artificial weed-killer'...they know and prefer their weeds...you get the idea. They are AGI robots that learn (are coached) to operate in a specific role and then are 'intellectually nobbled' (equivalent to biology), so their ability to handle novelty is specifically and especially curtailed. They will also be a whole bunch cheaper in that form...They are then deployed into that specific role and will be happy little campers. These creatures are different to typical mainstream AI fare because they cannot be taught how to learn. They are like us: they learn how to learn. As a result they can handle novelty better...a long story...Initially the A-Fauna is very small but potentially it could get to human level. The first part of the development is the initial proof of specific physics, which requires a key experiment. I can't wait to do this! The success of the experiment then leads to development and miniaturisation and eventual application into a prototype 'critter', which will then have to be proven to have P-consciousness (using the test in 3 below)anyway...that's the rough 'me' of it. I am in NICTA www.nicta.com.au Victoria Research Lab in the Life-Sciences theme. Department of Electrical/Electronic Eng, University of Melbourne Sothe AGI-09 basic topics to choose from are: 1) Empirical refutation of computationalism 2) Another thought experiment refutation of computationalism. The Totally Blind Zombie Homunculus Room 3) An objective test for Phenomenal consciousness. 4) A novel backpropagation mechanism in an excitable cell membrane/syncytium context. 1) and 2) are interesting because the implication is that if anyone doing AGI lifts their finger over a keyboard thinking they can be directly involved in programming anything to do with the eventual knowledge of the creature...they have already failed. I don't know whether the community has internalised this yet. BTW that makes 4 ways that computationalism has been shot. How dead does it have to get? :-) I am less interested in these than the others. 3) Is a special test which can be used to empirically test for P-consciousness in an embedded, embodied artificial agent. I need this test framework for my future AGI developments...one day I need to be able to point at at my AGI robot and claim it is having experiences of a certain type and to be believed. AGI needs a test like this to get scientific credibility. So you claim it's conscious?prove it!. This is problematic but I am reasonably sure I have worked out a way So it needs some attention (a paper is coming out sometime soon I hope. They told me it was accepted, anyway...). The test is double-blind/clinical style with 'wild-type' control and 'test subject'...BTW the computationalist contender (1/2 above) is quite validly tested but would operate as a sham/placebo control... because it is known they will always fail. Although anyone serious enough can offer it as a full contender. Funnily enough it also proves humans are conscious! In case you were wondering...humans are the wild-type control. 4) Is my main PhD topic. I submit this time next year. (I'd prefer to do this because I can get funded to go to the conference!). It reveals a neural adaptation mechanism that is completely missing from present neural models. It's based on molecular electrodynamics of the neural membrane. The effect then operates in the syncytium as a regulatory (synchrony) bias operating in quadrature with (and roughly independent of) the normal synaptic adaptation. I prefer 4) because of the funding but also because I'd much rather reveal it to the AGI community - because that is my future...but I will defer to preferences of the groupI can always cover 1,2,3 informally when I am there if there's any interestso...which of these (any) is of interest?...I'm not sure of the kinds of things you folk want to hear about. All comments are appreciated. regards to all, Colin Hales --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
Re: [agi] Testing, and a question....
Hi Ben, Excellent. #4 it is. I'll proceed on that basis. I can't get funding unless I present...and the timing is perfect for my PhD, so I'll be working towards that. Hong Kong sounds good. I assume it's the Toward a Science of Consciousness 2009 I'll chase it up. I didn't realise it was in Hong Kong. The last one I went to was Tucson. 2006. It was a hoot. I wonder if Dave Chalmers will do the 'end of consciousness' party and blues-slam. :-) We'll see. Consider me 'applied for' as a workshop. I'll do the applications ASAP. regards, Colin Hales Ben Goertzel wrote: In terms of a paper submission to AGI-09, I think that your option 4 would be of the most interest to the audience there. By and large it's not a philosophy of AI crowd so much as a how to build an AI crowd... I am also organizing a workshop on machine consciousness that will be in Hong Kong in June 09, following the major consciousness conference there ... for that workshop, your option 3 would be of great interest... ben On Fri, Oct 3, 2008 at 5:01 PM, Colin Hales [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Dear AGI folk, I am testing my registration on the system,, saying an inaugural 'hi' and seeking guidance in respect of potential submissions for a presentation spot at the next AGI conference. It is time for me to become more visible in AGI after 5 years of research and reprogramming my brain into the academic side of things My plans as a post-doc are to develop a novel chip technology. It will form the basis for what I have called 'A-Fauna'. I call it A-Fauna because it will act like biological organisms and take their place alongside natural fauna in a chosen ecological niche. Like tending a field as a benign 'artificial weed-killer'...they know and prefer their weeds...you get the idea. They are AGI robots that learn (are coached) to operate in a specific role and then are 'intellectually nobbled' (equivalent to biology), so their ability to handle novelty is specifically and especially curtailed. They will also be a whole bunch cheaper in that form...They are then deployed into that specific role and will be happy little campers. These creatures are different to typical mainstream AI fare because they cannot be taught how to learn. They are like us: they learn how to learn. As a result they can handle novelty better...a long story...Initially the A-Fauna is very small but potentially it could get to human level. The first part of the development is the initial proof of specific physics, which requires a key experiment. I can't wait to do this! The success of the experiment then leads to development and miniaturisation and eventual application into a prototype 'critter', which will then have to be proven to have P-consciousness (using the test in 3 below)anyway...that's the rough 'me' of it. I am in NICTA www.nicta.com.au http://www.nicta.com.au Victoria Research Lab in the Life-Sciences theme. Department of Electrical/Electronic Eng, University of Melbourne Sothe AGI-09 basic topics to choose from are: 1) Empirical refutation of computationalism 2) Another thought experiment refutation of computationalism. The Totally Blind Zombie Homunculus Room 3) An objective test for Phenomenal consciousness. 4) A novel backpropagation mechanism in an excitable cell membrane/syncytium context. 1) and 2) are interesting because the implication is that if anyone doing AGI lifts their finger over a keyboard thinking they can be directly involved in programming anything to do with the eventual knowledge of the creature...they have already failed. I don't know whether the community has internalised this yet. BTW that makes 4 ways that computationalism has been shot. How dead does it have to get? :-) I am less interested in these than the others. 3) Is a special test which can be used to empirically test for P-consciousness in an embedded, embodied artificial agent. I need this test framework for my future AGI developments...one day I need to be able to point at at my AGI robot and claim it is having experiences of a certain type and to be believed. AGI needs a test like this to get scientific credibility. So you claim it's conscious?prove it!. This is problematic but I am reasonably sure I have worked out a way So it needs some attention (a paper is coming out sometime soon I hope. They told me it was accepted, anyway...). The test is double-blind/clinical style with 'wild-type' control and 'test subject'...BTW the computationalist contender (1/2 above) is quite validly tested but would operate as a sham/placebo control... because it is known they will always fail. Although anyone serious enough can offer it as a full contender. Funnily enough
Re: [agi] COMP = false
useless. So the words 'refutation of COMP by an attempted COMP implementation of a scientist' have to be carefully contrasted with the words you can't simulate a scientist. The self referential use of scientific behaviour as scientific evidence has cut logical swathes through all sorts of issues. COMP is only one of them. My AGI benchmark and design aim is the artificial scientist. Note also that this result does not imply that real AGI can only be organic like us. It means that real AGI must have new chips that fully capture all the inputs and make use of them to acquire knowledge the way humans do. A separate matter altogether. COMP, as an AGI designer' option, is out of the picture. I think this just about covers the basics. The papers are dozens of pages. I can't condense it any more than this..I have debated this so much it's way past its use-by date. Most of the arguments go like this: But you CAN! I am unable to defend such 'arguments from under-informed-authority' ... I defer to the empirical reality of the situation and would prefer that it be left to justify itself. I did not make any of it up. I merely observed. . ...and so if you don't mind I'd rather leave the issue there. .. regards, Colin Hales Mike Tintner wrote: Colin: 1) Empirical refutation of computationalism... .. interesting because the implication is that if anyone doing AGI lifts their finger over a keyboard thinking they can be directly involved in programming anything to do with the eventual knowledge of the creature...they have already failed. I don't know whether the community has internalised this yet. Colin, I'm sure Ben is right, but I'd be interested to hear the essence of your empirical refutation. Please externalise it so we can internalise it :) --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=114414975-3c8e69 Powered by Listbox: http://www.listbox.com
RE: [agi] Intelligence enhancement
This be Snyder... http://www.centerforthemind.com/ Tread carefully. cheers, Col --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
RE: [agi] BDI architecture
Both Peter Wallis and Mike Georgeff have a history at Melbourne University. http://www.cs.mu.oz.au/agentlab/ and they and RMIT http://www.agents.org.au/collaborate a lot. There is a mailing list from which you may launch queries. They are an active group and quite approachable. I've gatecrashed their seminars many times! BDI seems to have found a very useful placein the weak AI that is current agent technology. Most of the BDI talks that I have witnessedare all about ontologies for application of agents in different problem domains. It gets a bit sterile from an AGI builder's point of view. BDI is 'visible' in an AGI but, like so many other models (eg Schmidhubers or Baars or, yes..Pei Wang and Eliezer and my own! ), if you create an AGI based on an intelligence model you have to ask yourself is it sufficient to create an AGI. If Baars' model was complete (global workspace) then presumably Stan Franklin's IDA (a real life completed implementation of it) would be on this mailing list arguing with us! I like to think of them as analogous to building,say, the combustion process, whilst thinking all along that you're building a car. Idealised models can be very right and very mis-directing. They are all worth a look though and as I say, they are all 'right'. cheers Colin -Original Message-From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]On Behalf Of Mike DeeringSent: Wednesday, 5 March 2003 4:04 AMTo: [EMAIL PROTECTED]Subject: Re: [agi] BDI architecture Margeret Heath [EMAIL PROTECTED](CyberMeg), BDI (beliefs, desires and intentions) www.cs.toronto.edu/~mkolp/9-agents3831.pdf brief explanation of a Microsoft BDI agent named Jack. "Intention without Representation: Implementing SituatedBDI agents" Peter Wallis, The Journal of Philosophical Psychology, (submitted July, 2002) Existing agent models emphasize an intentional notion of agency - the supposition that agents should be understood primarily in terms of mental concepts such as beliefs, desires and intentions. The BDI model (Rao and Georgeff, 1991, 1995) is a paradigm of this kind of agent, although some attempts to include social aspects have been made. yallara.cs.rmit.edu.au/~ldesilva/doc/cs435.pdf This document says BDI was developed by A. Rao and M. Georgeff in 1991 based on work by Bratman 1987 and Singh 1991. ccc.inaoep.mx/~mapco/AGws-IRT.pdf Automatic Generation and Maintenance of Hypertext for the Web through Information Retrieval Techniques the construction of a belief-desire-intention (BDI) agentusing the Zeus Agent Building Toolkit www.labs.bt.com/projects/agents.htm http://www.inf.pucrs.br/~giraffa/x-bdi/Fruto do trabalho desenvolvido pelo professor Michael da Costa Móra em sua tese de doutorado no CPGCC/UFRGS - FCT/UNL. A ferramenta X-BDI foi desenvolvida com a intenção de diminuir a distância que existe entre teorias formais para especificação de agentes cognitivos e sua programação. Mike Deering, Director www.SingularityActionGroup.com
RE: [agi] Playing with fire
(1) Since we cannot accurately predicate the future implications of our action, almost all research can lead to deadly result --- just see what has been used as weapons in the current world. If we ask for a guaranty for safety before a research, then we cannot do anything. I don't think the danger of AGI is more clear and near than most of other research. (2) Don't have AGI developed in time may be even more dangerous. We may encounter a situation where AGI is the only hope for the survival of the human species. I haven't seen a proof that AGI is more likely to be evil than otherwise. So my position is: let's go ahead, but carefully. Cheers, Pei Ref: http://www.optimal.org/peter/siai_guidelines.htm www.singinst.org/CFAI.html www.goertzel.org/dynapsyc/2002/AIMorality.htm Extra credit: I've just read the Crichton novel PREY. Totally transparent movie-scipt but a perfect text book on how to screw up really badly. Basically the formula is 'let the military finance it'. The general public will see this inevitable movie and we we will be drawn towards the moral battle we are creating. In early times it was the 'tribe over the hill' we feared. Communication has killed that. Now we have the 'tribe from another planet' and the 'tribe from the future' to fear and our fears play out just as powerfully as any time in out history. --- I'm pretty sure we're screwed. I've read a lot and thought a lot about this. Whilst I disagree with Hugo Degaris' AGI (as actually constructing the worst case AGI- an Ungrounded adaptive mega-puppet of the type most likely to give us maximum grief early), I concur with his fears -there is a battle looming one way or another. Either between AGI/Human or Human/Human or both. I try to look at things very simply and pragmatically. The easiest characterisation of the AGIs we're aiming at is 'apparent causality modellers'. (Crick and Koch characterise our brains like this: Francis Crick Christof Koch A Framework for Consciousness .Nature Neuroscience, February 2003 Volume 6 Number 2 pp 119 - 126. see http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/v6/n2/abs/nn0 203-119.html ) No matter what you do, statistically we're going to be faced with scenarios commensurate with 'reality' as made 'apparent' to our AGI. This is all we have to go on. Lattitude to create abstractions (personal truths) that are 100% factual crap and then act on it comes with the package. IMO there is only one parameter that offers any hope: Cognition grounded in real phenomenology. Our AGI, if it is to survive in the reality of our laws of physics, has to be fully grounded like us. Only then will it inhabit the same universe as us and be able to understand the universe the way it needs to survive. This does not mean hooking up a few signals. It means real qualia within the 'apparent causality modelling'. This will not stop it from creating abstractions with nasty side effects for us humans. But at least we will be able to communicate with it through the common ground of existing in the same reality. For those who don't get qualia: Consider a pressure transmitter 4-20mA connected to a PC running an 'AGI'. The 4-20mA signal is around 10^15 electrons smashing into a wiring terminal. That's meaningless noise. Pressure is a label we humans attribute. The most incredible deductive AGI engine in the universe will _never_ understand pressure like this. If an outside agency tells it it is pressure the learning is grounded in the outside agency, not in the AGI. It still does not understand pressure, only how to respond to questions about pressure by an outside agency so it appears like it understands. Qualia means actual pressure_ness (intentional phenomenology in matter) in its head. Prohibition doesn't work. The current round of self censorship in the AAAS conference is a nice thing, but the Iraq's of the world will still find a way. I suspect there will be a time when people will herd around the Artificial Mind Institute with a new form of life in it, demonstrating against it. You can do all the philosophical categorisation in the universe and postulate all sort of strategies. IMHO they all mean squat. I'm pretty sure we're going to have to face this thing full on and cop the consequences. I'm with Pei Wang. Let's explore and deal with it. cheers, Colin Hales --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
RE: [agi] Playing with fire
PhilipI personally think humans as a society are capable of saving themselves from their own individual and collective stupidity. I've worked explicitly on this issue for 30 years and still retain some optimism on the subject. Colin: I'm with Pei Wang. Let's explore and deal with it.OK, if you're with Pei, what exactly is the position that you are not with?Cheers, Philip The usual paradoxes and dichotomies: -letting it happen by accident. -getting too misty eyed abouthomo sapiens sapiens when we're a last minute mayfly in the scheme of things. -not getting misty eyed enough about homo sapiens sapiensand not looking after my children's interests. -not including a moral/ethical department alongside the AGI development (like this one!) -letting an ungrounded self modifying AGI loose. -constructing a cottage industry out of second-guessing an outcome. We have enough insurance salesmen. -behaving like the down side has already happened. -behaving like the down side won't happen... Cheers Colin
[agi] The AGI and the will to live
Hi all, I find the friendliness issue fairly infertile ground tackled way too soon. Go back to where we are: the beginning. I'm far more interested in the conferring of a will to live. Our natural tendency is to ascribe this will to live to our intelligent artifacts. This 'seed' is by far the hardest thing to create and the real determinant of 'friendliness' in the end. Our seed? I think a model that starts from something like 'another heartbeat must happen'. When you don't have a heart? What - poke a watchdog timer every 100msec or die? My feeling at the moment is that far from having a friendliness problem we're more likely to need a cattle prod to keep the thing interested in staying awake, let alone getting it to take the trouble to formulate any form of friendliness or malevolence or even indifference. If our artifact is a zombie, what motivation is there to bother _faking_ friendliness or malevolence or even indifference? Without it Pinnocchio the puppet goes to sleep. If our artifact is not a zombie (.ie. has a real subjective experience) then what motivates _real_ friendliness or malevolence or even indifference? Without it Pinnocchio the artificial little boy goes to sleep. Whatever the outcome, at its root is the will to even start learning that outcome. You have to be awake to have a free will. What gets our AGI progeny up in the morning? regards, Colin Hales --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
