How Brain Architecture Leads To Abstract Thought (umass.edu)

catchblue22 writes: UMass Amherst scientists have analyzed fMRI data to link brain architecture with consciousness and abstract thought. "We momentarily thought our research failed when we saw that each cognitive behavior showed activity through many network depths. Then we realized that cognition is far richer, it wasn't the simple hierarchy that everyone was looking for. So, we developed our geometrical 'slope' algorithm," said neuroscientist Hava Siegelmann (abstract). "With a slope identifier, behaviors could now be ordered by their relative depth activity with no human intervention or bias," she adds. They ranked slopes for all cognitive behaviors from the fMRI databases from negative to positive and found that they ordered from more tangible to highly abstract.

"'Deep learning is a computational system employing a multi-layered neural net...the brain's processing dynamic is far richer and less constrained because it has recurrent interconnection, sometimes called feedback loops.' Her lab is now creating a 'massively recurrent deep learning network,' she says, for a more brain-like and superior learning AI."

Comments

Re:Obligatory Archer

[Anonymous Coward]

As long as we can get it through the 'terrible twos', it doesn't have to be all bad...

Re:Obligatory Archer

[Anonymous Coward]

Do you want power to be based on blood-lines? Because this is how ignorance breeds oppression.

Strong AI claims another researcher! .

[WOOFYGOOFY]

FTA:
"'Deep learning is a computational system employing a multi-layered neural net...the brain's processing dynamic is far richer and less constrained because it has recurrent interconnection, sometimes called feedback loops.' Her lab is now creating a 'massively recurrent deep learning network,' she says, for a more brain-like and superior learning AI.""

But this is not new. This is connectionism and all its descendants. Anyways...the larger point is, that there is no reason to believe that consciousness or experience falls out of machines, no matter how wet or complex they are. Consider that anything which can be modeled with a computer can be modeled with something much more primitive, albeit in a cumbersome way, for example, a Turing tape or a even a very fancy abacus made of wood, wires, beads. Yes you definitely want to keep that fact in mind before you pin that Strong AI Booster pin on your lapel.

If we want to get down to brass tacks and go really hardcore the one thing we have absolute irrefutable evidence for is just experience itself. Everything else is possibly a chimera based on a re-construction of experience-as-thought; material substance is real the same way a false belief or an illusion is real.

If anything, consciousness gives rise to matter (as an illusion or particular way of thinking) and also all the laws of physics and everything that flows therefrom. Given the absolute non-reputability of the existence of experience it's more likely that consciousness is the "fundamental" stuff of the universe than matter or that consciousness arises from sufficiently complex arrangements of non-conscience matter.

Anyway, it is overwhelmingly likely that at this stage of our knowledge and scientific inquiry we just radically and fundamentally misunderstand the nature of matter and consciousness. People still think about densely packed protons and neutrons but in fact there are only quarks (at least so far) and a quark is a "thing" oh, about the size of a virus, which is contained in a sphere of emptiness- just nothing- the diameter of Neptune's yearly orbit around the sun. So there's your "material stuff". Last seen receding ever further into purely mathematical constructs....

We specialize because we have to with the result that we just don't think big enough or more accurately closely and precisely enough about what the Big Picture is telling us. Science slowly grinds down, or more charitably "hones", your imagination in a particular way so as to make you a successful researcher or current theorist.

Whatever framework we are using to try to understand consciousness - call it reductionism to material- is most likely as wrong as voodoo for reasons which will later be shown to have been hidden to us now. If that doesn't just immediately strike you as most likely correct, then consider the argument has inductive support: that's been the historical tale of the tape when we are in the same state of ignorance with respect to a subject matter as we are with respect to the brain.

AI researchers who start quacking on making big, definitive statements about consciousness and the brain are just marking themselves as lesser lights and can always count on a round of ridicule, at least from me, on /.

Re:Strong AI claims another researcher! .

[Hognoxious]

Very interesting.

Now, about those fries...

Re:Strong AI claims another researcher! .

[Anonymous Coward]

You have a very fancy way of saying "I don't know it, therefore it's unknowable". It's not a logical position to take.

Re:Strong AI claims another researcher! .

[AthanasiusKircher]

Anyways...the larger point is, that there is no reason to believe that consciousness or experience falls out of machines, no matter how wet or complex they are.

Sure there is, if you actually believe in the empirical world. And that's all you're "evidence" is that you're discussing -- the physical stuff you see and can measure and dissect. There's no way currently to "prove" that consciousness emerges from the physical stuff of the brain, but the burden of proof is on your side to prove that there's "something else" there, because the rest of physical reality as we sense it seems logically consistent with materialism.

Consider that anything which can be modeled with a computer can be modeled with something much more primitive, albeit in a cumbersome way, for example, a Turing tape or a even a very fancy abacus made of wood, wires, beads. Yes you definitely want to keep that fact in mind before you pin that Strong AI Booster pin on your lapel.

What kind of crappy philosophical argument is this? Just because you can model the physical atoms of the brain with a sufficiently large "abacus," materialism must be false? Have you ever heard of "emergent" behavior? Have you ever seen apparent complexity emerging from apparently simple systems? People have been convinced that ELIZA was a conscious person. Now you have the brain with many orders of magnitude more complexity than the simple ELIZA algorithm, and you're telling me that you couldn't possibly be fooled by emergent behavior in such a complex mechanical system into thinking there was consciousness there -- or even strongly believing that beings with such mechanisms (like other humans) might possess consciousness... even if it's based only on a hugely complex material basis??

If we want to get down to brass tacks and go really hardcore the one thing we have absolute irrefutable evidence for is just experience itself. Everything else is possibly a chimera based on a re-construction of experience-as-thought; material substance is real the same way a false belief or an illusion is real.

Sure, but this is the path to solipsism. If you can't prove the material world exists, you can't even prove that other people or other minds exist. So, actually, it's inaccurate to say that the "one thing we have irrefutable evidence for is just experience itself." Rather, you should have said, "the one thing *I* have irrefutable evidence for is just my own experience -- I don't know that anything outside of my own thoughts exists, and I could be participating in a pseudo-masturbatory exercise of pretending that there is an external world and other people out there and this imaginary place called Slashdot where I pretend to communicate with these figments of my imagination."

That's where your argument leads ultimately. If you accept the material world exists, and you accept that there is actually a "we" in a meaningful sense of other people who might also be conscious, then you need to explain how they might be conscious in the way that you are. And short of some other explanatory method, materialism seems a pretty good explanatory tool for most progress in understanding and making consistent sense out of our perceptions of this apparent sensory world.

If anything, consciousness gives rise to matter (as an illusion or particular way of thinking) and also all the laws of physics and everything that flows therefrom. Given the absolute non-reputability of the existence of experience it's more likely that consciousness is the "fundamental" stuff of the universe than matter or that consciousness arises from sufficiently complex arrangements of non-conscience matter.

On what basis could you possibly judge what is "more likely" in this argument?? Is this possible? Sure. It's also possible that we're all brains floating in vats, or that we're in the Matrix, or whatever. But you have no empirical (or even a pr

Re:Strong AI claims another researcher! .

[swillden]

there is no reason to believe that consciousness or experience falls out of machines, no matter how wet or complex they are

And there is no reason not to believe that our experience of consciousness arises out of anything more than the complex interactions of our neural structure. If you had any evidence that there is something other than matter involved, you would have stated it. In the absence of non-material causes, I'll stick with the Occam's razor assumption and not require the invention of new non-physical constructs.

Your post is just mysticism dressed up as reason.

Re:Strong AI claims another researcher! .

[rgbatduke]

Let me see, religious much?

Berkeley rides again. Seriously.

I now cease giving you your subjective existence by no longer thinking about you. Sorry about that.

rgb

Re:Strong AI claims another researcher! .

[rgbatduke]

Rather, you should have said, "the one thing *I* have irrefutable evidence for is just my own experience -- I don't know that anything outside of my own thoughts exists, and I could be participating in a pseudo-masturbatory exercise of pretending that there is an external world and other people out there and this imaginary place called Slashdot where I pretend to communicate with these figments of my imagination."...

...in which case you should be wondering, "Why am I bothering to pretend to argue with myself about this?" and return to pretending that if you jump out of a tall building over hard pavement that those illusory "laws of physics" won't erase your brain function and self very much like a squashed bug.

Or that a stroke or simple old age won't do the trick, like it or not, over time.

rgb

Re:Strong AI claims another researcher! .

[swillden]

Excellent response.

As an aside, here's my answer to solipsism. It's almost certainly not novel, though it is original:

If I'm all that exists, damn I'm smart.

But, weirdly, I always attribute my deepest and most interesting thoughts to other parts of myself, which I call other people. I mean, I think I'm reading a book on statistical inference, or poetry, or listening to music, and there's all this enormous complexity and elegance which I don't understand and doesn't fit together... until I work hard to understand it, then it does. So if I'm all that exists, I somehow invented, say, non-Euclidean geometry, linear cryptanalysis, Bayesian inference, number theory, General Relativity, and a thousand other deep, sophisticated and -- most importantly for this argument -- intricate yet self-consistent ideas, but then convinced the one part of myself that I recognize as me that I don't know them and have to learn them. And then in the process of "learning" I go through a series of misunderstandings, where stuff I imagine I'm reading doesn't quite make sense, until I do get it, at which point everything falls neatly into place. Perfectly into place, actually. So perfectly that it's clear I had to have worked it all out in advance.

If I'm all there is, I'm some bizarre sort of split personality, and some of my other personalities are crazy smart. Heck, it's not even unlikely that one of my other much-smarter-than-me personalities will reply to this post and explain why it doesn't make sense, providing yet another iteration of me explaining what I already worked out to myself.

My conclusion is: The existence of the massive body of intellectual work apparently produced by other people demonstrates to me that other people do, in fact, exist, because I'm demonstrably not smart enough to have done it all. Unless I'm insane.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

I said:

Anyways...the larger point is, that there is no reason to believe that consciousness or experience falls out of machines, no matter how wet or complex they are.

You said:

Sure there is, if you actually believe in the empirical world. And that's all you're "evidence" is that you're discussing -- the physical stuff you see and can measure and dissect. There's no way currently to "prove" that consciousness emerges from the physical stuff of the brain, but the burden of proof is on your side to prove that there's "something else" there, because the rest of physical reality as we sense it seems logically consistent with materialism.

No I only affirm what I know with certainty- that there IS experience. You know there is also, and so does everyone else. The "something else" is something your side owns completely. You believe you know there's a physical world and you believe you know of it's characteristics only because of experience- the experience of looking in a microscope or out the window or reading a book or thinking. But the experience here is absolutely primary and cannot be disputed; you cannot say it does not exist.

Now that we know with 100% absolute certainty that experience exists and no skeptic can meaningfully deny it, let's take a good look at material substance. We have chased that stuff for thousands of years and each time we have been certain we have gotten to the irreducible and final form it, except for currently. While we may believe in quarks, in whatever sense a wriggling bit of string theory exists, researchers, chastened by experience, also allow that this might not be the final stopping point. So Experience: 1 matter: TBD, possibly eternally.

I am not making the above up, that is actually the way things are. I am not distorting the record, not in the least.

So it is on this basis that I claim that something which IS is more likely to be the basis for something which MAY BE than the other way around.

It is also the basis by which I conclude that it's more likely to be the fundamental stuff of the universe than, say, the equations of String Theory, even more so since, uh actually, equations are purely the product of a kind of experience, an experience we call logical thought.

So yes, I am saying that the world you take to be "out there" may be nothing but experience and all the things you think are properties of that world- time, quarks, space, all of it is possibly not real, possibly real in the sense of being a compelling illusion of thought and possibly "composed" of experience in some way no weirder than a cat is "composed" of the non-corporeal wriggling strings of string theory- which, please keep in mind, it literally true.

Observe that I am not playing games of any kind here. I am only stating the facts as they are.

The Turing tape remark of mine is meant to jar the reader out of what I call "wetworld handwaving" that Strong AI gets into. That's a trick of the mind whereby complex systems, if they're made of wet biological stuff, get infused with a magical quality which permits the leap from "complex machine" to "consciousness".

On this point, you can take a lesson from Marvin Minsky, who admits that if very complex systems give rise to consciousness because of "ether", er, I mean "emergent properties" then it must admit of degrees. So it follows that, and he says this explicitly, simple feedback loops like oh, your house's thermostat, actually possess a rudimentary form of consciousness.

It's fun to think about the implications of this imminently reasonable and completely defensible corollary of Strong AI and the 'emergent property" theory you espouse. What else might be conscious by dint of complexity- the weather? Perhaps we the Ancients were, perhaps crudely, really onto something with their Gods of the Wind and Fire and War and the Sea.

Once you open the door to the "emergent properties of complex systems" you're going to find yourself working overt

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

You said:

Sure, but this is the path to solipsism. If you can't prove the material world exists, you can't even prove that other people or other minds exist. So, actually, it's inaccurate to say that the "one thing we have irrefutable evidence for is just experience itself." Rather, you should have said, "the one thing *I* have irrefutable evidence for is just my own experience -- I don't know that anything outside of my own thoughts exists, and I could be participating in a pseudo-masturbatory exercise of pretending that there is an external world and other people out there and this imaginary place called Slashdot where I pretend to communicate with these figments of my imagination."

Well this is exactly backwards. I am not saying I actually exist- in words that's a contradiction but being a fair fellow you will admit my meaning- it's just what I said. The "I" is one of those things which could be an illusion like any other form of matter. Whose illusion? *Just an illusion owned by no-one.* Just an experience, which is an illusion which exists and stop right there.

I am not saying "people have experiences" I am saying "there is experience". Out of cogito ergo sum take away cogito and replace it with experience. Experience, ergo experience. Anything else is an assumption.

When I say (all material) things could exist the way illusions exist, there IS an existence to illusions, but it's not a "full existence" if you get my meaning. And you do get my meaning. For example, if you go crazy and believe you see the ghost of Blackbeard telling you to quit this science thing, find yourself a seaworthy ship and start living life, there is a sense in which you DID see the specter and another one in which the specter was an illusion. This is a state of affairs, assigning different levels of reality to phenomena, that is not confusing or controversial in the least.Certainly it's not less weird than Minsky assigning degrees of consciousness to increasingly complex cybernetic feedback loops or you assigning increasingly "emergent properties" to increasingly complex systems.

So my statements are clear and any contradiction involving the word "I" ( i.e. I doubt the ultimate reality of my personal identity ) is an unfortunate byproduct of language, and this little quirk of language is a separate issue from the truthfulness and accuracy of my statements, which are of course fairly dubitable, but I don't think , wrong.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

You said: How is it "overwhelmingly likely"? Because you say so? Do you have some way of estimating that probability? Or do you really just mean, "I think it is thus!"

No, it's for the reason I said- our deliriously sparse knowledge of the brain itself. See my post.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Your post is just mysticism dressed up as reason.

You can say that but it's not true. Literally, you know there is experiences. Literally, everything else is in doubt. Literally.

As to your "not any reason to believe" comment, this is just an unsupported assertion on your part. You have a BIG problem- how a Turing machine composed of tape and a reader can display every form of consciousness possible. It so absurd on the face of it that you have to resort to phlogiston -like theories of "emergent properties". You know what "emergent properties" are? A noun phrase without a referent. They're a whatever-you-want-to-to-put-in-here trashcan of hand-waving bullshit

Literally here is your reasoning:

The brain gives rise to consciousness because it manifests (or contains or supports) complex networks of electrical impulses which have "emergent properties" one of which is consciousness.

Do you know how how much that \smells like classic bullshit "I don't know what's going on, but here's a noun phrase" scienci-ness. Did you take a class on the theory of science? I mean do you know what does or does not qualify for a scientific theory?

It's not what I am saying that has got your panties in a bunch; it's what I am not saying. I am not saying I buy into department approved, career advancing, conformist philosophical bullshit. I am not saying anything more than what I know. and certainly I have not said anything as weird and unsupported as "emergent properties".

Emergent properties is what you say when you don't know what to say about what you don't understand. You can apply it to literally anything to frame as actual the existence of "things" which do not and never did exist. Get a hold of yourself man, it's the magical pixie dust of science.

It's more dignified by orders of magnitude to just say "I have no fucking idea; it's a mystery to me." when your theories lead you to the edge of ridiculousness.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

I agree with you in spirit. I am also I don't know-er, but in a specific way I won't get into. Anyways I am not saying that the physical world doesn't exist at all in any sense; I am saying it's exact nature and relation to "consciousness" or what I would rather call "experience" is not explained by strong AI and I am making the case for experience as the preferred fundamental "stuff" of the universe, which resolves the so called "mind-body" or "consciousness-brain" "experience-matter" problem in favor of the existence of the thing we know exists, experience; this is an approach which seems nothing but reasonable to me.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

See my reply to the solipsism charge below. I am not saying anything like this.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

You are talking about direct experience. While irreftuable, it is also unprovable (like most everything else really).

I understand the meaning of irrefutable to be essentially "provable". You may wonder as to the"reality"of what you experience, but the fact that you have an experience requires no proof since a thing which is (and is in the proof-theoretical sense, true) requires no (further) proof. If by proof you mean "elaboration", then sure, experience may be something which cannot be "reduced" or elaborated on any further. It just is.

Logically, science can really never settle the matter. Yet if you ask someone if it's OK to end their experience, they magically start to appreciate it somewhat, regardless how little respect they have for life in general.

It is possible that some true "facts" cannot be represented by logical thought but are not outside the realm of knowability. That is not an unreasonable hypothesis for someone to hold.

It must be unsettling, for the dogmatic scientific mind, that what they seek, can only be obtained through direct experience of second-hand phenomena. It's like shadows in the cave, only worse: to the n-th degree at all times. Yogis call them worlds within worlds, but it might just be an excuse to keep all options open in an uncertain world too.

Our entire history is largely defined by a certain "kind" of thinking and the various systems of belief that kind of thinking gives rise to. That's an encompassing truth so vast that almost no one sees it.

The only refuge for the unsettled and smallish mind, is to start believing that nothing is truly real, alive or great, so as to protect its already established convictions and conditionings. Because, it's usually the prejudice and ego that feels threatened, history shows it always will have to yield to more illumination, knowledge, and ironically enough: science.

Reductionsim as a form of escapism.

Maybe everything is just clockworks all the way down. If it is, it'll truly always remain the most magnificent clockwork of them all!

Do you believe that? Is that what you're deepest instinct tells you is true? It's your life. Are you going to let someone separate you from your own knowledge?

Re:Strong AI claims another researcher! .

[ceoyoyo]

"Strong AI are the ones hypothesizing a relationship and capability of a thing - the brain- here. The burden falls squarely on you, not me. Prove it."

People actually trying to build AI ARE trying to prove it, by making one. You're sitting in your armchair saying "it's impossible because magic!"

If you want to get down to it, your position is that there is something quintessentially different about intelligence that we cannot model. That's an extraordinary claim, with absolutely no evidence backing it up. On the other hand, the idea that intelligence is something complicated but based on ordinary principles is not extraordinary at all.

Re:Strong AI claims another researcher! .

[ceoyoyo]

How do you know you have experiences? Your memory could very well be an illusion. You could simply exist as a fleeting quirk in an infinite universe of random arrangements that think it's sentient.

If we're going to get all mystical, we might as well not pull punches.

Re:Strong AI claims another researcher! .

[AthanasiusKircher]

No I only affirm what I know with certainty- that there IS experience. You know there is also, and so does everyone else.

Yes, but how do you make the epistemological leap from the fact that "there is experience" (i.e., whatever you experience) to the reality that I am having "experience" and that "everyone else" has experience?

How could you possibly know "with certainty" that I or "everyone else" exist and have experience without accepting that there is some reality to the world as generated by your senses? Even if the only thing you believe about the reality in the world is that other consciousnesses exist, the only way you have evidence of them is through the medium of your sensory experience.

It's a huge logical leap to go from "I exist" and "I experience" to the "certainty" that "we all have experience" without making an assertion to the validity of SOME of your sensory experiences, i.e., those about the existence of others and their behavior.

Now that we know with 100% absolute certainty that experience exists and no skeptic can meaningfully deny it, let's take a good look at material substance. We have chased that stuff for thousands of years and each time we have been certain we have gotten to the irreducible and final form it, except for currently.

And I would counter that I am NOT "certain' that we have reached any irreducible or final understanding of the material world. Any decent scientist who has any basic understanding of epistemology should also agree with that. But just because we haven't made some sort of final determination about the properties of something doesn't mean that it's ontologically defective. There are LOTS of things we don't understand about the nature of experience -- I've spent a lot of time reading literature about phenomenology and the philosophy of perception. We hardly have complete and final understanding of how that works either. Yet you're convinced that that truly has real existence, while you're iffy about the material world because you don't think you have complete knowledge of it. That's a weird argument.

On this point, you can take a lesson from Marvin Minsky, who admits that if very complex systems give rise to consciousness because of "ether", er, I mean "emergent properties" then it must admit of degrees. So it follows that, and he says this explicitly, simple feedback loops like oh, your house's thermostat, actually possess a rudimentary form of consciousness.

You seem to find this disturbing, as if somehow admitting "consciousness" to a thermostat creates some fundamental problem in your universe or ideology.

I don't freakin' care. "Consciousness" to me is just a name for an emergent behavior. Perhaps it has degrees. Perhaps a dog is "conscious." Perhaps not. Does it matter? Apparently a great deal to you. But to me, not so much. To me, it's all about some sort of arbitrary definition of an emergent behavior. To you, it seems to be endowed with a fundamental ontological distinction. How you come to that distinction is still a mystery to me. And I have no idea why you think it MUST exist. How do you KNOW for certain that your thermostat isn't "conscious"? Why? Is a dog "conscious"? How do you know? Are you certain? How about a "brain-dead" person? Are they "conscious"? How do you know? What if their body still responds to certain stimuli, apparently a byproduct of reflexes and certain nerve pathways? Where do you the draw the line? More importantly -- how are you SO CERTAIN where you can draw the line about who or what has this special kind of "experience" you seem to think the universe is made of?

Once you open the door to the "emergent properties of complex systems" you're going to find yourself working overtime gerrymandering the legal boundaries of *which* complex systems should qualify for "consciousness" and the more things you find you have to gerrymander, th

Re:Strong AI claims another researcher! .

[AthanasiusKircher]

Man, that's harsh. Very cold. I wish I had thought of something that pithy instead of trying to argue. :)

Re:Strong AI claims another researcher! .

[AthanasiusKircher]

People actually trying to build AI ARE trying to prove it, by making one. You're sitting in your armchair saying "it's impossible because magic!"

To be fair to him, he's a little more sophisticated than your average religious nut who believes that people have "souls" or something and computers can't. Instead, he has a somewhat complex metaphysical argument about the nature of reality, and he ultimately is happy to deny the existence of the material world. (Why he would bother arguing so long with people hear about such issues then, is beyond me -- from his perspective, we're all lunatics who have been taken in by some illusion.)

From my perspective, the philosophical argument is a bit muddled. But it's not like he's sitting in his armchair saying "it's impossible because magic!" -- rather, he isn't even sure his armchair or anything else exists -- except perhaps some version of disembodied "experience," which he somehow has "100% certain" knowledge exists not just for himself but for "everyone else." So it's not "impossible because magic!" but rather "it's impossible because the material world, including all robots, is basically ontologically inferior to 'experience' and since robots might all be an illusion anyway, there's no way they could have true 'experience' or 'intelligence' or 'consciousness' or whatever."

It's wacky, but I give him slightly more credit for claiming more than "It's magic!" as a religious wacko might. He's vaguely trying to make a rational argument, even if it seems pretty kooky.

Re:Strong AI claims another researcher! .

[swillden]

You should learn something about emergent properties. They're extremely common. In fact, nearly all phenomena we perceive are emergent properties of underlying processes that are rather different from what we see. In some cases the emergent properties are much simpler than what's really going on underneath, in other cases they're vastly more complex. To provoke some thinking about higher-level emergent properties, I recommend that you read Hofstadter's "Godel, Escher, Bach". It's an old book, and decidedly overly optimistic about AI in some areas, but very thought provoking.

Anyway, your argument is, basically, "I don't buy that this common thing is happening in this more complex case that I don't understand, so I'm going to throw science out the window and assume magic."

Note that I'm not actually saying you're wrong. We don't know. We don't understand intelligence, or consciousness. It is possible that thought is the stuff of which the universe is made, rather than vice versa. But that's a pretty radical departure from what we do understand, and there's no clear rationale for going that way. I mean, the clearest argument you have stated thus far is summed up in your sentence "Do you know how much that smells like classic bullshit?" (I added the question mark, since you forgot it). You claim it smells like bullshit, and then rant about how you're a non-conformist... but offer nothing to refute the simpler, more straightforward interpretation.

My bet is that research and experimentation like the above will move us incrementally toward actually understanding how our brains work, and interaction between that work and work on building ever more sophisticated (and effective![1]) "thinking" machines will eventually lead us to figure out how to build machines that rival or even surpass our own abilities. We won't build strong AI by accident, though; we'll build it by developing sophisticated theories of intelligence and applying them. Along the way I expect we'll learn some things about the hard limits of such computation, which will prove to apply to our brains just as much as they apply to anything we can build. I suspect that one of the things we'll learn is that fuzziness is a necessary attribute for computational universality, and that when we build really smart machines we'll also have constructed machines that make many of the same kinds of thinking errors that we do.

Those are just my guesses, though. I'm in no way claiming to have some special insight.

[1] If you doubt the effectiveness of the work that's been done, you haven't paid attention to the phenomenal advances in the last few years in areas like voice recognition, natural language processing and image recognition.

Re:Strong AI claims another researcher! .

[khallow]

there is no reason to believe that consciousness or experience falls out of machines,

A brain is a machine with consciousness and experience. So we're trying to construct in a different way, a machine to do something for which we already have an existence proof.

Re:Strong AI claims another researcher! .

[rgbatduke]

Sorry, I let a reference to Bishop Berkeley, who made all of these arguments long ago back in the heyday of the materialism vs idealism wars. The idealism argument lost to everybody who wasn't religious a couple of centuries ago:

http://www.iep.utm.edu/berkele...

I just didn't have the patience to walk through the entire tired argumentation associated with Berkeley and his tree falling in the quad and god and David Hume's empiricism that ultimately won the argument and ended the era of bullshit philosophy. But I can, if you like.

Several posts in the thread do a perfectly fine job of it, though. Solipsism cannot be refuted, but neither is there any good reason to think it is true. The same is the case for Berkeley's arguments. Scientific materialism, OTOH, is supported empirically to the extent that nobody seriously doubts it compared to the alternatives simply because there is no evidence to oppose it and every minute of our existences constitutes experience that tends to support it.

rgb

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Muy claims are not about intelligence per se (whatever thatis) but about experience. My TI -80 is intelligent by a lot of measures. More elaborate TI-80s are not going to save the empty assertion that consciousness arises from complexity of any sort, unless you strip consciousness of it's defining characteristic- a thing having experience - and reduce it to "clever compuation which would have formerly required a human being to do". A calculator qualifies for that, as does a chess playing program.

Your side slides greasily between A) consciousness-meaning-clever-computation and B) consciousness-meaning-a-thing-posessing-experience.

You like to achieve things from column A and use them to imply victories in column B.

Note that victories from A are interesting enough, I have some of my own, but not Interesting in a metaphysical or deep way. There are no basic mysteries being resolved here.

So if you claim I am claiming we cannot model intelligent activity, you're wrong. if you're claiming we cannot model or bring into being a thing posessed of experience, you're right.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Should also say "bring into being through sheer computation" because sometmes you are arguing with eople who justike to argue ad win instead of argue in a proper, disinterested way.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Come on, you hear me ranting because you want to hear it that way. There is no rant here.

Regarding emergent properties, you're not getting clear in your own mind what they refer to. Either they refer to a property which is reducible to the thing which subtends them- like the heat of a fire or the color of the grass, or (it is claimed that) they don't.

The first "kind" are unremarkable- this is just straighforward material reductionism.

The second kind are magical, as in unicorns- i.e. they don't exist. These would be emergent properties which are physically dependent on the thing which subtends them but ontologically independent (for some reason). What THAT kind of emergent property requires is a break in causality between the necessary substance which "gives rise to" the "emergent property" and the emergent property itself. That's a joke. The only reason anyone even entertains that idea is because they're either not thinking clearly or think material things can be deepely related in a non-causal (however indirect that linkage is) relationship, i.e. related but not related, which is just voodoo.

We won't build strong AI by accident, though;

You don't know what Strong AI is. Strong AI isn't "very effective AI" or "big project AI" or "amazing results AI". It's the specific assertion that computation IS consciousness and visa-versa. This is not my personal defiinition, this is THE definition. It's helps if we're talking about the same thing.

When someone says that consciousness is an emergent property of the brain (Strong AI- this paper) or of a computer or of a Turing machine they should know they're not saying anything more than "I don't know how experience falls out of material substance, but have some bullshit instead".

Re:Strong AI claims another researcher! .

[iMactheKnife]

Actually, matter gives rise to consciousness, not the reverse. Any computational substrate, meat-engine or silicon, is made of matter. In the creation of any simulation of the physical world, the same laws of physics must ultimately apply. I am paraphrarsing Dr. David Deutsch, Oxford University.

Re:Strong AI claims another researcher! .

[ceoyoyo]

Your posts are pretty much content free. You haven't even defined what you mean by "experience." What you consider experience is just memories of the past. Your TI-85 has those, and they're much more reliable than your own.

You're still using a lot of words to basically say "it's magic" (whatever you choose "it" to be). Maybe it is. Until someone shows some evidence that it is, science proceeds based on the assumption it's not.

Re:Strong AI claims another researcher! .

[ceoyoyo]

Since you brought up religion, you should read some of the arguments philosophers and theologians have invented in order to support the idea. They're very sophisticated. Solipsism in particular, which is what the OP seems to be heading into, has been well explored. If you're honest though, all those arguments rely on one of two basic theses: "we can never know because magic" or "we can never know because we can never know anything!" Both are... unproductive.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

OK in the context of this conversation that qualifies as "assuming the consequent".

Re:Strong AI claims another researcher! .

[khallow]

OK in the context of this conversation that qualifies as "assuming the consequent".

Ok, what's being assumed and why is that a problem?

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Stay focused on the thing to be proved. Thething to be proced is not "can we build a clever machine" or "can we model how the brain learns or reacts ina computer" (I assume this is possible). It's not even - can we learn something true and new about actual brains from models we create in computers. None of that is what's in question.

What's in question is- can consciousness arise from computers (of any type, wet, vacuum tubes Turing tapes it's all the same thing) or, equivalently is consciousness a byproduct of brains (alone).

Lots of interesting things to do and contemplate in this space. We're talking about one of them.

Re:Strong AI claims another researcher! .

[khallow]

I am saying that computers can't auto-generate a class of (real, actual) thing via "emergent properties" by dint of being "sufficiently complex". That's pholigston-level hand waving and appeal-to-noun-phrase.

And do you have a reason for saying this? I'm seeing the same pattern again. Assertions made with absolutely nothing to back them up.

My view is that brains already generate experience and consciousness. We have an existence proof that it can be done. Even if consciousness is an external property/deeper reality that somehow attaches to some or all human brains, there's still no reason to expect that it won't similarly be able to attach to similarly complex human-made machines.

Whatever hand waving you apply to rationalize why human brains are a seat of experience can similarly be applied to any other system with the same attributes.

Re:Strong AI claims another researcher! .

[khallow]

Solipsism in particular, which is what the OP seems to be heading into, has been well explored.

Solipsism may also be a real thing. Modern physical theories remain stubbornly attached to a bunch of arbitrary parameters. Those parameters may have the values they have because you observed them that way.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

So again if you read my rebuttal to solipsism you'll see I am explicitly not saying that.

Lelt me turn the tables on you for a second and examine YOUR beliefs (or someone who accepts that consciousness can arise from brains or computers) and let's see how weird they are.

First you have to admit that basis of consciousness is not tied to any particular material- biological or silicon or anything else- it's an issue of systems and computation and any suitable substrate capable of doing the computation, even if the computation is entirely in a chip, will do. Given the *right* computation, consciousness is created.

So if I just run the right program with the right simulated inputs I can and will produce a thing having experience on a chip. That's what you believe.

I am not mocking this idea since every road we take from here for any distance goes to a strange place, including mine. I just want you to see how strange your own landscape becomes, because I don't think you are aware of it.

Then you have to admit, because we all agree it's true and it is true, that that said same experience-having chip-creature could be reproduced completely, including the experience-having part, in a very long tape and a tape mover and tape reader, i.e. a Turning machine.

Now this really sounds like I am making fun of you, but I am not deliberately doing that. I am just a passive passenger looking out the window and commenting on the passing scenery in the car you're driving down the road you chose.

In addition to this your idea of consciousness coming from computation has some curious properties. For one, it admits of degrees. This is Minsky's thermostat but realize also you have the unenviable task of deciding when the blessed event arises in arbitrary machines. Even if your philosphical opponent doesn't make you explicitly define what is and is not to qualify as experience-bearing, still you ought to be able to give a thumbs up thumbs down (although how you would do that without an explicit checklist is anyone's guess) to a given set of machines. Just how conscious is my calculator and is there any truly principled reason to assume the answer is zero? It needs a bucnh of feedback loops? It needs to do something fancier than calculate integrals? Then experience will blossom forth or slowwwwly starts to accrue or how does that work exactly?

These are just some of the things implied by the computation-as-consciousness Strong AI claims.

Really the quandry that the existence of experience puts us in is not one we can extricate ourselves from casually. Your road is plenty weird. Mine is weird also. I own that weirdness with the claim that the essential incoporealness and ontological primacy of consciousness and experience is not very much weirder and in fact is much closer to the ontological weirdness that the things of physics has become and moreover ,it's consistent and doesn't hypothesize the existence of things we don'tknow with 100% certainity actually exist.

. Physicists publish papers all the time wondering if we're all in a hologram and such like things and no one gets out the woo-woo cannon and points it at them. By making the claims String AI makes, researchers have to accept that those claims come with huge unresolved issues that they can't just handwave their way past.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

That's not quite solipsism. Observer bias is a benign problem which can and does still admit of an objective, shared reality. Even an inescapable observer bias, even an inescapable, undetected and undetect-able observer bias still admits of a real, if never-to-be-known objective reality. So that's still not solipsism.

I too admit of an objective reality which exists independently of myself. It is the reality that experience reveals, the reality that experience is. I simply doubt the reality of the existence of things and all the other stuff of thought.

If I do bad math and come to bad conclusions, in what sense are my beliefs real (about the things of my bad-math universe) ? I am totally convinced of them because of the way I used my mind. But I used my mind wrongly. They don't exist except as figments of my wayward thinking, and then in some sense, OK they exist.

I claim the same thing for notjust the products of my bad mathematical musings but all thought-things. They have a kind of existence, but that existence is not what it appears to be by my mis-thinking mind.

Re:Strong AI claims another researcher! .

[khallow]

That's not quite solipsism.

And it's not quite not solipsism.

Even an inescapable observer bias, even an inescapable, undetected and undetect-able observer bias still admits of a real, if never-to-be-known objective reality. So that's still not solipsism.

Unless, of course, that objective reality exists solely because you observe it and the other observers. It's not as strong as if you're making everything up, but there is a possibility here that reality exists as it does because of characteristics of an observer of that reality.

in what sense are my beliefs real

In a very strong sense, since they have a physical representation as knowledge in your brain.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

Sure I have a reason. For the same reason no other phenomena has ever had an "emergent" property which also posessed ontological independence. Because reductionism is how things ARE. Things are made up of their parts, and how those parts interact with the world and other parts determine completely how the thing behaves.

Your second point is related to what I am saying, but not a direct attack on it (which is what you think). What I am saying is the claim that consciousness arises from brains or complex systems is wrong. Whether there can be a brain imitating computer is another matter. If the computer imitates the brain perfectly and is not forced into being an exact replica of the brain down to its constituent parts- ion channels , physical neurons etc, then that would be interesting. But experience isn't going to magically issue from that brain anymore than it does from our own. The brain is not the seat of experience. I am claiming that experience exists independently of brains and gives rise to the perception that there are brains.

You're doing a lot of interrogating and accusing but sliding by without answering the troubling questions which issue directly from your own theory. Can a Turing tape machine be conscious and if not then why not? Are some suitably complex machines already conscious to some degree and if not, then why not? Start there.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

In a very strong sense, since they have a physical representation as knowledge in your brain.

What I am obviously saying is they do not correlate to the things in the external world I thought they did. They are real the way illusions are real illusions. They're real (an illusion is a real phenomena), but their designated referents are not real.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

>>What you consider experience is just memories of the past.

Yeah, actually that's not true. You are having an experience right NOW. At any given instance you care to check it you are having an experience. You don't need to refer to past time or memory or any othersuch thing. Just look NOW and see for yourself.

It's amazing and a measure of the dogmaticism and rigidity that abounds in some circles that in the course of this debate you've actually now tended an argument which rejects the reality of experience.

I also enjoy it when people act bewildered about what this crazy thing "experience" is I speak of. Because that's all it is, an act.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

This is what's being assumed:
A brain is a machine with consciousness and experience.

And that's a problem because "consciousness" or it's more precise definition "experience" even if you allow that it is dependent on brains or complex machines to manifest (and I am not saying I do) is a phenomena which nevertheless is not a brain or complex machine and is has no credible reductionist (scientific) pathway from either of those things.

We don't assign experience to a watch; it "knows" nothing even though it "tells" time- a complex task. And so also with any even more complicated machines. At some undefined point of complexity, you claim a ghost enters the machine, and lo, the machine is having experience. But do one computation differently and no experience is being had.

It's just enough that any computer you can build, I can match the functioning of given a long tape and a scanner , a Turing machine. That is enough to give any impartial person pause at what exactly it is being claimed by Strong AI. Consciousness and experience "comes out of" Turning machines. Why not go all the way then and just say the universe, the grandest Turing machine of all is conscious and has experience? Some distinguished physicists actually DO say this. No one scoffs at them. Really, looked at in this context, your claim that consciousness is found only in brains and computers looks like a piece of special pleading. Maybe the whole universe is conscious, given your criteria for consciousness.

The ideas espoused by Strong AI really depend on the enforcement of a kind of incurious parochialism with respect to science at large and a certain overhwelming of the credulous by complexity and the magic of seeing something done (the Jeopardy! computer comes to mind here) which previously would have required a human.

 

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

OK some people are not familiar with the philosophical problem called the mind-brain duality. It goes like this.

Even if brains are necessary for minds (experience, consciousness some how needs them to manifest- not something I myself believe ) they are not sufficient.

They are not sufficient because we can imagine a brain working and causing all the effects it does without a shred of conscious experience. The reason this is not just plausible, but the preferred explanation is because the atoms and molecules which constitute the brain and cause it to do what it does don't NEED conscious experience to react with each other in any other system and there is no plausible reason they should in this one. Brains are not a special form of matter to the laws of physics. They operate in brains and to make brains the same way they operate everywhere else, and that is without "experience".

No other physical system created and governened by the forces of physics and atomic particles NEEDS this new TYPE of thing -conscious experience -to function.

Put baldly, the laws of physics control the atomic (and sub-atomic) particles which in turn control and determine the interaction of molecules which control and completely determine both the structure and the functioning of the brain don't need jack shit from "experience", are not thought to have it, and the functioning of these experience-free parts can completely and deterministically explain the functioning not just of brains generally, but any given event in any brain there ever has been or ever will be.

Those same laws of physics don't make any other system- no matter how complex- "conscious" or causeit to have "experience".

If we let the quinessential deterministic mechanism be a clockworks then we can express this idea of mechanisitic determinism governing all phenomena both organic and inorganic, alive and inert, all biological systems included with all their parts -including brains - a clockwork universe,. Clocks dont' need experience and we dont' assign it to them because Occam's Razor and all that and "experience" is entirely superfluous to the clockwork.

BTW this idea is where the title of A Clockwork Orange comes from. The seemingly friendly biological thing is really just a mechanism at it's core and nothing more.

Experience is an phenomena the universe just doesn't need . I could create a parallel, stone cold dead no-experince universe with animals and people and everything and it would run just like our own. The only difference is, ours has experience and the other one doesn't.

When we dispose of the dead weight of uneeded concepts in science- like "life force" or "elan vital" or all forms of immaterial vitalism and spirtualism in all its guises, it's always a process of showing that what we took for evidence of the (ontological ) existence of some unseen and inferred "thing" was not evidence for this things existence at all, and the thing could be done away with with no loss of explanatory power to our theories.

In the case with experience, it's just the opposite. I can show that we don't need it, it's just fucking dead weight , I can acount for allphenomena without including it as a concept and yet ...unlike with every other dead weight concept, we're not "inferring it's existence from evidence; we know it exists. In fact, it's the only thing in the entire universe we can truly say we are absolutely, positively out-Descarting Descartes, full on skeptical-unto- pathology certain of, without harboring even a shread of doubt no matter how small. There is experience.

If I can account for the all the workings of brains without experience then how are you, asa good scientist going to add it back in? Where are you going to do this in the process? Anywhere you try, I can say "and here you come with your ghost in the machine". I will be scientifcally rigorous and you will be injecting mystical entities into biochemical, or molecular or electrical events governed and account for 100% by the

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

OK so I see you as mixing up the concept of "abstract concept about a thing" with "emergent property".

The top speed of a car is a concept about the car, not a property. This is seen by seeing how many other emergent properties - as you're using the term- the car might have. The answer, in short is an infinite number. Here are some of them:

The chick magnet strength of the car.

The most miles the car will go without needing a repair of its transmission if I drive it only to church on Sundays.

The best price the car will fetch under any circumstances whatsoever.

and on and on.

What these are are not properties of the car, but descriptions of how the car will interact with other parts of the world in a potentially infiniite number of artificial circumstances.

These are not emergent properties as that term is usually meant. Emergent properties are properties of a thing that
1) posess casual power independent of the thing
2) are nevertheless somehow dependent upon the underlying "thing" in order to emerge
3) the effects of that casual power are not reducible to the the determiistic effects of the underlyig thing upon the emergent thing (as is the case everywhere else i the universe).

So it's a break in the reductive chain of causal linkage which otherwise exists between different "levels" of material substance. (I. e. the behavior and properties of atoms cause and determine the behavior and properties of molecules which cause and determine the behavior and properties or organic compounds and so on up the chain of causal linkage to brains.. aka "reductionism").

So to be clear, emergent properties are said to be dependent upon (in some way, so they can emerge from) a lower level of in the causal chain but their effects are NOT reducible or determined, even in theory, by those lower level's properties.

This kind of thing just doesn't exist and anyway would be magical.

So all that's really going on is that we do not know how to map e.g. insight, or knowledge through the various layers of brain architecture , function, chemical processes, neuron count, connectedness

OK this is something different than the above now. Your problem here is how do you account for the fact of experience if you dno't need experience to account for the behavior of brains - and you absolutely don't - see here:

  http://slashdot.org/comments.p...

Re:Strong AI claims another researcher! .

[khallow]

And that's a problem because "consciousness" or it's more precise definition "experience" even if you allow that it is dependent on brains or complex machines to manifest (and I am not saying I do) is a phenomena which nevertheless is not a brain or complex machine and is has no credible reductionist (scientific) pathway from either of those things.

Let's think about what you just wrote. You admit that consciousness is a phenomena and hence, observable. You then admit that the only place you've observed it is in brains. I'm not seeing an argument which I can rebut here. But I'll look at your next, more extensive reply to see if you have anything there.

We don't assign experience to a watch; it "knows" nothing even though it "tells" time- a complex task. And so also with any even more complicated machines. At some undefined point of complexity, you claim a ghost enters the machine, and lo, the machine is having experience. But do one computation differently and no experience is being had.

Telling time is as complex as intellectual tasks, like abstracting ideas from patterns? Do tell. And no, I do not make the "ghost enters the machine" argument. I see no reason to expect consciousness to be a bit you set, but like so many other things, a gradual attribute.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

OK I will answer this post again since I missed parts of it in my reply.

I did read and re-read GEB a long long time ago and everyone would benefit from your recommendation. + ! on GEB.

Regarding so called "emergent properties" I wrote to another poster here

http://slashdot.org/comments.p...

but I'll clean it up a little more:

Emergent properties are properties of a thing that

1) posess casual power independent of the thing

2) are nevertheless somehow dependent upon the underlying "thing" in order to "emerge"

3) the effects of that casual power the emergent properties posess are not reducible in any deterministic way to the effects of the underlying "thing" from which the emergent property emerges.

(regarding 3 above, if 3 were NOT being claimed by the "emergent crowd" then just everything at the next higher level of composition would qualify as an "emergent property" - heat from a fire, light from the burning of a tungsten filament etc etc. )

So it's a break in the reductive chain of causal linkage which otherwise exists between different "levels" of material substance. That reductive chain and causal linkage you know very well- the behavior and properties of atoms cause and determine the behavior and properties of molecules which cause and determine the behavior and properties or organic compounds and so on up the chain of causal linkage and deterministic causality all the way to brains.. aka "reductionism").

So to be clear, emergent properties are said to be dependent upon (in some way, so they can emerge from) a lower level in the causal chain but their effects are NOT reducible or determined by, even in theory, those lower level's properties. The way everything else is in the known universe.

This kind of thing just doesn't exist and anyway would be magical.

Emergent properties sound good -until you think about what's actually being implied by the concept. Then the concept just falls apart.

It is possible that thought is the stuff of which the universe is made, ...But ...there's no clear rationale for going that way.

OK I do have a clear rationale. It is that "experience" (consciousness) is 100% superfluous to the working ofthe brain. It's a concept we don't need, like we don't need "life force" to explain the difference between live things and non-live things. Yet there it is.

To see how this state of affairs amounts to a real genuine quandry for science, I answered another debater with a fuller explanation which is clear enough (given the hour I am typing) as it is:

http://slashdot.org/comments.p...

but I will shorten and clarify it into this:

We can imagine a brain working and causing all the effects it does without a shred of conscious experience.

The interaction of molecules which control and completely determine both the structure and the functioning of the brain don't need jack shit from "experience", are not thought to have it, and the functioning of these experience-free parts can completely and deterministically explain the functioning not just of brains generally, but any given event in any brain there ever has been or ever will be.

If we let the quinessential deterministic mechanism be a clockworks then we can express this idea of mechanisitic determinism governing all phenomena both organic and inorganic, alive and inert, all biological systems included with all their parts -including brains - a clockwork universe,. Clocks dont' need experience and we dont' assign it to them because Occam's Razor and all that and "experience" is entirely superfluous to the clockwork.

Experience is an phenomena the universe just doesn't need . I could create a parallel, stone cold dead no-experince universe with animals and people and everything and it would ru

Re:Strong AI claims another researcher! .

[khallow]

So that's where we are. But since there IS experience, where do you go? You go with the obvious other possibility, although it's counter-intuitive to the limit of that word. You say "I cannot deny experience. Any construction of the universe which excludes it ontologically (as ours does) must be in error, no matter how compelling it seems to me. That is what I am arguing for.

The odd thing here is that our models of the universe don't exclude "experience" ontologically or otherwise. I don't see anything here to rebut either. Bad priors, bad conclusions (or GIGO). Too bad since you've given this a lot of thought.

Even if brains are necessary for minds (experience, consciousness some how needs them to manifest- not something I myself believe ) they are not sufficient.

They are not sufficient because we can imagine a brain working and causing all the effects it does without a shred of conscious experience.

No, actually we can't imagine that assertion of the second paragraph. My view is that if we couldt he above, we would just be agreeing with my original stance, not have this particular argument at all, and nuanced disagreement would be elsewhere.

Experience is an phenomena the universe just doesn't need . I could create a parallel, stone cold dead no-experince universe with animals and people and everything and it would run just like our own. The only difference is, ours has experience and the other one doesn't.

Then do so. Create this sterile universe, rather than merely say that you can. Then we can compare the two and see if there actually is anything different about them.

If I can account for the all the workings of brains without experience then how are you, asa good scientist going to add it back in? Where are you going to do this in the process? Anywhere you try, I can say "and here you come with your ghost in the machine". I will be scientifcally rigorous and you will be injecting mystical entities into biochemical, or molecular or electrical events governed and account for 100% by the dead-unconscious no-experience here folks laws of physics, of things just bumping into things, or more modernly, of forces acting on particles..

You haven't actually accounted for what you claim here. And if you can observe something, then it is no longer mystical.

There are quoted here, three spectacular and completely unfounded assertions. I realize that this is just the traditional abuse of a poor rhetorical tool (rather than say, a serious claim that you can create universes), but I think it's another symptom of rather sloppy reasoning.

Moving on, strong AI would also be able to exploit each of your objections. After all, we can keep adding things until we do get consciousness (the brain indicates this process is not that hard once you have a sufficiently sophisticated system in which to manifest consciousness. Also, I don't consider complexity to be a directly ordered set BTW.). Second, consciousness in strong AI is no more or less precluded by our models of reality than the brain is. Since we're asserting consciousness exists, then the models are just as right or wrong for strong AI as they are for the human brain).

Third, we don't rule out phenomena in science just because it doesn't fit models. Instead, the model changes (or more accurately, new models are created to account for the new phenomena). Experience exists therefore, if we want to model it, we include it. "Ghost in the machine" is one such way to attempt that. Whatever models we develop would have to account for what we think happens in human brains. And thus, would allow for the same phenomena in similar systems which are not human brains.

Re:Strong AI claims another researcher! .

[khallow]

Sure I have a reason. For the same reason no other phenomena has ever had an "emergent" property which also posessed ontological independence. Because reductionism is how things ARE. Things are made up of their parts, and how those parts interact with the world and other parts determine completely how the thing behaves.

No. Reductionism doesn't explain initial conditions (going beyond the anthropic principle). To be blunt, I think it quite reasonable that the fact that consciousness exists in the universe is an imposed precondition of WOOFYGOOFY the observer and any other observers whose preconditions are compatible with those of WOOFYGOOFY.

I am claiming that experience exists independently of brains and gives rise to the perception that there are brains.

And I can claim that the Moon is made of green cheese. We need more than claims here.

You're doing a lot of interrogating and accusing but sliding by without answering the troubling questions which issue directly from your own theory. Can a Turing tape machine be conscious and if not then why not? Are some suitably complex machines already conscious to some degree and if not, then why not? Start there.

Sure, why wouldn't an appropriately programmed Turing tape machine be conscious? Or for that matter, not just conscious, but encoding the full extent of our universe and all the conscious beings who dwell therein. The infinite computational machine is a really big thing. Who knows, we may all be running on a Turing machine equivalent (or perhaps a semi-quantum version of it)?

I don't know if there are already human-made machines which achieve some degree of consciousness, but it's not something I'd rule out.

And honestly, I don't see these questions as being troubling. It's not any more troubling to say that a Turing machine can be conscious (or much more than that), than it is to say that a few kilograms of goo can be conscious.

Re:Strong AI claims another researcher! .

[khallow]

What I am obviously saying is they do not correlate to the things in the external world I thought they did. They are real the way illusions are real illusions. They're real (an illusion is a real phenomena), but their designated referents are not real.

Note that I did, fully answer your posed question. The applicability of the belief doesn't change the reality of the belief. That's a different thing.

And beliefs are experientially highly dependent on how closely and extensively you test those beliefs. For example, if I believe that the Moon is made of green cheese, and no one can be bothered to test my assertion at even the simplest of levels (even to the rudimentary observation that the Moon is not green in color, which would require at least some elaboration of my belief in order to explain that) nor do I ever try to go to the Moon to get some of that delicious cheese, then the belief is real enough. It doesn't attempt to explain anything that I experience and hence, is just as real as any other similarly untested belief that I should happen to have, no matter how correct that belief happens to be.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

What kind of crappy philosophical argument is this? Just because you can model the physical atoms of the brain with a sufficiently large "abacus," materialism must be false?

No I am not arguing against materialism here, I am saying that the idea that computation equals cognition- Strong AI's basic claim- is silly on the face of it.

It's not that I can model a physical system and therefore it doesn't exist (I guess this is what you are accusing me of saying?). It's that Strong AI is making it's Strong AI claim- consciousness IS computation. Well, here's something that can compute anything- does this now make it conscious?

Neither am I saying that other things can't be conscious. What I am saying is if they are, the source of that consciousness is not their brains. Turing machines are good models of computers and computation. IMHO the Turing test lacks a lot for being a test of ... whatever consciousness, AI.. basically because people are gullible in the short term.

You're going after solisism and I am not that. I am only willing to defend myself against so many things I didn't say. This is one of them.

from here:
http://science.slashdot.org/co...

It's not solipsism because solipsism means the only thing I know for sure is my own self and my own thoughts. That is subtly (depending on whose thinking about it) a very different statement than "there is only experience", or experience is ontologically superior or the only "real" or "ontologically most real" thing. Hopefully those two things are clearly distinguishable from each other.

If you conclude that I must think this or that about other people or their minds or their existence or non-existence of whatever, well... I don't. I just say- the only thing that is certain is experience. Full stop. This by itself has big consequences of its own it turns out, sice experince is totally superfluous to the functioning of brains - we can imagine a brain which behaves exactly the same as any other, but has no experience since brains are controlled by their underlying biochemical processes - which have no experience- they're underlying electrical activity- which has no experience- and so on and so on and it is these things which totally determine the brain's function.

So anytime you feel like pointing to a process and inserting "experience" into it, just remember that *no one needs you're mystical thing to account for the behavior of brains* and inserting it any scientific theory is itself unscientific since it contributes nada to the explanatory power of the theory.

The only issue then is, *experience does exist*.

So how are you going to be a good scientist? If you inject experience into brains at some level- the biochemical, the electrical, the computational, via so called emergent properties, you're doing bad science or rather not doing science at all, since "emergent properties" all scientific bullshit on a stick:

http://slashdot.org/comments.p...

  If on the other hand you choose to ignore experience, well, you're avoiding doing science on a category of thing , which is not *really* doing science in the fullest sense since science is willing to account for al lphenomena and this is one we can't pretend doesn't exist (although, to be fair, BF Skinner did make exactly this claim).

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

>> To be blunt, I think it quite reasonable that the fact that consciousness exists in the universe is an imposed precondition

Imposed by whom? Your prof? Yout think just exists as a kind of fundamental matter, like the Weak Force? You are arguing against that. Anyway this thing "consciousness or experience" (same thing) has deeply troubling properties. If you say it's emergent, then that's scientific bunk. If you say no, it's a part of the reductionist chain, then it's a thing without purpose , one our explanations can do without. It has neither mass, nor volume nor does it interact with anything causally. Sounds like Life Force level bullshit to me.

You're tlaking out of two sides of your mouth without realizing it.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

never understood -1 insightful comments. They're comments which have a high rating and people like them but have been modded down by a one or a bunch of people determined to put thehated poster in his or her "place" by jacking the rating system.

Good to know.

Re:Strong AI claims another researcher! .

[WOOFYGOOFY]

OK khallow you're the climate change denying troll that is the ONE person I actually did mark as a troll in my control panel. I answer posts without reading the name because it's not important to me who you are but in your case, actually, I amnot arguing with a person who is legitimately confused over nad over again I am arguing with someone whois a known to me troll.

Suffice it to say that that each of yoru asertions are wrong. Wrong that we can't imagine a world without consciousness for exactly the reasons I delineated. Wrong about "emergent properties" being bull shit , a point I won't prove again and wrong about it not being anuncessary but extant 5th wheel to any theory which tries to make use of it.

I am pretty sure anyone else reading my posts does not have your objections and I am pretty sure you're the single person interested enough in this argument (with me, you're hated climate change opponent with the signature you hate) to downvote its (my) posts days after it's been moved 3 or 4 pages back from the front and for which there are only 100 psts, fully a quarter of them mine.

You're the worst kind of liar. An insincere, time-wasting trolling system-gaming liar. That, sir is what you are. Goodday.

Re:Strong AI claims another researcher! .

[khallow]

OK khallow you're the climate change denying troll that is the ONE person I actually did mark as a troll in my control panel.

Here's some background on that. I think no further comment is required.

Suffice it to say that that each of yoru asertions are wrong. Wrong that we can't imagine a world without consciousness for exactly the reasons I delineated. Wrong about "emergent properties" being bull shit , a point I won't prove again and wrong about it not being anuncessary but extant 5th wheel to any theory which tries to make use of it.

Obviously, I disagree. Let's start with the first assertion. Do you know what consciousness and experience are so that you know when you are imaging their absence? I doubt it. I certainly don't have that knowledge.

Further, you have yet to establish why consciousness appears. I believe presence of a sufficiently complex (in various ways) physical system is necessary and sufficient. Merely imaging a copy of our universe without consciousness then becomes futile since it would have consciousness as well in contradiction of your desired imagining.

Wrong about "emergent properties" being bull shit

Let's look at your definition of emergent properties:

Emergent properties are properties of a thing that

1) posess casual power independent of the thing

2) are nevertheless somehow dependent upon the underlying "thing" in order to "emerge"

3) the effects of that casual power the emergent properties posess are not reducible in any deterministic way to the effects of the underlying "thing" from which the emergent property emerges.

1) and 3) aren't true. Emergent properties are dependent on the underlying constituent parts and their properties. For example, the many physical states and properties of matter such as a variety of solids (eg, crystal, glass, conducting and superconducting, transparent/translucent/opaque, color, etc are all well defined and easily observed emergent properties. And none of these properties has causal power beyond that of the constituent parts.

And 3) is wrong simply because there's numerous examples of many different constituent parts with many different properties yet resulting in overlapping emergent properties. The classic example is vibration. The average orchestra generates vibration in many different ways, such as vibration of a reed, of lips, or the skin of a drum. If you were to start with the vibration, you would have difficulty determining what generated the vibration. I think for example, it would be possible for a specially manufactured violin and trumpet to sound alike including overtones even though the modes of vibrations are different in each.

Thus, it would not be possible in that last example to distinguish the underlying constituent parts because the emergent property of vibration didn't have sufficient information to resolve the underlying cause.That in turn implies your assertion that the emergent property is "not reducible in any deterministic way to the effects of the underlying "thing" from which the emergent property emerges". It turns out you are saying less than you thought you were.

And if one thinks about emergent properties, point 3) is an expected outcome. You are discussing collapsing the properties of a many bodied system into a few emergent properties. There is a necessary loss of information and hence, a necessary inability to fully reconstruct the original system just from the description of the emergent properties.

and wrong about it not being anuncessary but extant 5th wheel to any theory which tries to make use of it

The obvious rebuttal is that if consciousness physically is an emergent property, then it doe

Re:Strong AI claims another researcher! .

[khallow]

Literally, when I see it's you posting, I ignore it. Sorry I didn't check before answering this time. You're the worst persistent troll on /. You malignantly anddeliberately engage in pretending to be genuinely confused and pretending to genuinely misunderstand the conversation just to antagonize your opponent in a debate. You're a broken person and I will not spend any time in conversation with you.

That's too bad. I don't believe you gave my arguments and discourse the same courtesy I gave yours.

Re:Strong AI claims another researcher! .

[RespekMyAthorati]

Well, I choose to believe that you do not exist.
You are just a just a component of my brain that exists only to amuse me.

Re:Strong AI claims another researcher! .

[RespekMyAthorati]

"I have no fucking idea; it's a mystery to me."

I agree that I don't. But unlike you, I don't try to pretend that I do.

serviscope_minor is shopping for shoes today

[Hognoxious]

If cows have different brain architecture to bulls, then this research should be suppressed and the perpetrators hounded out of their jobs in case it discourages heifers from embarking on STEM careers.

Re:serviscope_minor is shopping for shoes today

[Zontar The Mindless]

What does San Francisco Samoyed Rescue have to do with this? I'm not seeing the connection. Thanks!

stretching conclusions from false positives

[argStyopa]

Let's not forget that same fMRI technology successfully identified brain function in a DEAD SALMON.

http://blogs.scientificamerica...

Re:stretching conclusions from false positives

[ceoyoyo]

You post a link to an article about a scientific paper warning about the dangers of amateurs doing fMRI analysis and somehow think that's evidence?

JIHAD!

[Spy Handler]

Thou shalt not make a machine in the likeness of a human mind!

Re:JIHAD!

[dsmatthews9379]

OK wise guy first define, exactly, what a human mind is. Then when you have done that define what exactly is and is not God. At least I know what a bowl of spaghetti is when I see it!

Re:Obligatory Archer

[penguinoid]

Waiting is how you get Skynet.

Re:Obligatory Archer

[U2xhc2hkb3QgU3Vja3M]

I’m game. We’ll see who rusts first.

Re:I wonder...

[slashname3]

... if there are computers trying to understand what makes them work... Is there some computer out there that is pondering the question are humans capable of intelligent thought...

Actually that is a great question for all of us: Are humans capable of intelligent thought?

Based on data collected so far the odds of that are slim to none.

Intelligence in the universe is a constant. The population is growing....

I HATE the overhyped Story Title

[IceAgeComing]

"How Brain Architecture Leads To Abstract Thought"

Really, now that sounds super interesting, I've been waiting my whole life to read this story!

Oh, wow, the story is actually not about that at all. Could it be someone cynically posted this story with an overhyped title to get a few extra clicks?

Can someone please tell the story poster that there are people trying to find a new website with accurate information. Seeing lazy, clickbait headlines makes me want to hurl.

Re:Obligatory Archer

[ShanghaiBill]

Do you want Skynet? Because this is how you get Skynet.

You get Skynet anyway. The only question is who gets there first: The Western democracies, or China.

Neuronal Firing Rates and Implications for AI

[catchblue22]

Depending on sources, the firing rate of neurons is on the order of 1 to 200Hz. That would imply that any particular short term thought that we have must necessarily involve relatively few neurons, at least in terms of the depth of the tree. Of course there is a huge parallel dimension to particular waves of neuron firing associated with particular thoughts. However, at an intuitive level this does give me hope of the future of AI.

At the cellular level, our neurons are not that different from the neurons of fruit flies. In terms of evolution, once nature "solves" a problem, say creating functional networks of neurons, it largely recycles that solution in future generations. It may provide variations on the original theme, say by increasing the size and complexity, but the basic chemical and biological systems remain unchanged. To me, this suggests that AI is in fact "solvable", at least in terms of having it perform discrete tasks.

Re:Feedback Loops

[ceoyoyo]

Recurrent ANNs don't have to be trained to have feedback. It's built right in. The original scientists are probably talking about special kinds of feedback between intermediate levels, which the brain does seem to use in some cases (not all). That kind of thing was also tried in the past, but it's extremely difficult to train.

The article messed up their description.

Re: Obligatory Archer

[xrobertcmx]

As long as it has a wall plug I'm good.

That's my design

[iMactheKnife]

Hah! Told you so! (recursion is the key to self awareness, also see "I Am A Strange Loop" by Hofstaeder) http://tinyurl.com/h8dww8n.

Re:Obligatory Archer

[Kazoo the Clown]

To be governed by Skynet couldn't be any worse than what we've got now-- at least we'd have SOME kind of intelligence in government, even if it's artificial. It's clear there's no natural intelligence in government as it currently exists...