Discussion Topic |
|
This thread has been locked |
BLUEBLOCR
Social climber
joshua tree
|
|
Silly Moose, trix are for kids!
The summation of the flick was; "if we can figure out intelligence, the machine will figure everything else out!"
Jus another example of man trying to build his own God/Idol.
|
|
WBraun
climber
|
|
It's never been ever nor ever will be done!
No one can build God.
One can only build poseur.
That's what modern science has done .... poseurs!
Stooopid people who don't have a clue period say stoopid sh!t like we will build a god.
|
|
WBraun
climber
|
|
Yes 100%
There's 99.9% poseurs .....
|
|
Bushman
Social climber
Elk Grove, California
|
|
'The Big Bang Reverse Singularity and the Infinite Oppositorium'
We had visitors once but we have forgotten them.
The memory of them has long since left us.
On the outer perimeter of the universe lies the Infinite Oppositorium. A great inward facing surface of immeasurable density, composed of dark matter, the surface of which opposes the center of the universe which has in the past been referred to as the origin of our universe, the Big Bang. At the surface of the Infinite Oppositorium culminate great runnels of condensed Energy, Dark Energy, Dark Matter, Matter, and Antimatter. These great runnels which are called S windings, also accelerate all visible matter outwards towards the Infinite Oppositorium.
The Big Bang is not a 14 billion year old event as was posited by scientists in the past. The Big Bang is actually a Big Bang Reverse Singularity which is an ever present and constant event at the center of our universe. The Big Bang Reverse Singularity works in harmony and opposition with the Infinite Oppositorium and trades Energy, Dark Energy, Dark Matter, Matter, and Antimatter back-and-forth through the great runnels called S Windings and R Windings. Energy, Dark Matter, Matter, and Antimatter run back from the Infinite Oppositorium in the separate runnels called R windings to the Big Bang Reverse Singularity.
Think of our universe like a humongous amoeba with a great undulating surface of dense super compacted dark matter (the Infinite Oppositorium), with the Big Bang Reverse Singularity at its center constantly radiating accelerating Energy, Dark Energy, Dark Matter, Matter, and Antimatter outwards through the S Windings to the outer surface of the universe, the Infinite Oppositorium, which draws and pulls all of the Energy, Dark Energy, Dark Matter, Matter, and Antimatter out to it through the inter dimensional S Runnels and then back through the inter-dimensional R Runnels into the infinitely dense yet constantly radiating Big Bang Reverse Singularity.
The ancients were taught all of this by the visitors. The great civilizations of ancient Egypt, Greece, Mesopotamia, and Incans also had knowledge of this but it was forgotten.
We were visited once by the mutant humanoid emissaries known then as the Denvectivans, who were genetically engineered, programmed, and sent by the Collosatroids but we have forgotten them.
The Collosatroids populate the surface of the Infinite Oppositorium in great cities built from dark matter and powered by antimatter and dark energy. Dark matter and dark energy are not truly dark but are a shadowy golden platinum type of color. Dark energy is composed of a wavelength that as yet has not been discovered by our scientists, and which remains invisible to us.
Certain individuals among us carry DNA which will someday enable them to detect the dark energy wavelength and color.
This information has been revealed to me by a humanoid agent who has time traveled from the future at great peril. For now his identity must remain confidential. I will be relocating frequently to avoid detection by clandestine organizations who would seek to silence me. My source confides in me that there will be many more fantastic and extraordinary revelations to come. Until that time be safe and au revoir.
-bushman out
|
|
Largo
Sport climber
The Big Wide Open Face
|
|
|
Topic Author's Reply - Jun 9, 2015 - 05:43pm PT
|
At the end of the Demis Hassabis vid that John referenced he said that "to truly find the theory of everything we may have to solve 'intelligence' first."
My sense of it is that Demis might be a step off base. that we may have to solve the business of sentience first.
AI is a fascinating field but many of the leading researchers (to use Sam Harris' hilarious depiction of them) are situated somewhere on the Asbergers spectrum and are sci-fi geeks hopped up on Red Bull something fierce. It is from this camp that we get promises that by 2025 a machine will fall in love etc. These folks are simply overzealous, and make the common mistake of conflating conscious human experience with mechanical data processing.
And we've seen, and as you can see by even a cursory look at the models and aims of most AI paths, the obsessioin is with data processing which can only produce a kind of blind mechanism that is no more a threat to mankind than any other machine. No matter how fast or at what depth a machine can crunch the numbers, it can have no context or self-capacity to recognize a sentient being as anything different than any other thing out there. And there is no suggestioin or evidence that sentience would aid a machine per pure computing power. So the reasons to try and replicate or "create" machine consciousness are of little to no priority.
If it were, if western scientists and psychologists and whoever has valid input really born into finding out what consciousnes IS, one wonders if it would not open a door or two towared the aforementioned "theory of everything." As is, such a theory has always been chased down as something existing separate from and standing outside of mind. Perhaps it does. This is the materialists' most fervent dream. But that dagone observer keeps popoing up unexpectedly, and like a crazy Uncle in the attict, people scramble to explain him away.
JL
|
|
MH2
Boulder climber
Andy Cairns
|
|
MikeL:
I sense an infinite set of possibilities in everything. I tend to see everything as just one thing.
One way to escape criticism is to take refuge in contradiction.
I may do that, too.
edit:
You are befuddled, JL. Better to leave AI and physics to others.
Though it is interesting to get a look into your mind.
|
|
Bushman
Social climber
Elk Grove, California
|
|
Dark matter and dark energy are not truly dark but are a shadowy golden platinum type of color.
Silence might be golden, but levity platinum, unless...
"as an attempt to introduce a note of levity, the words were a disastrous flop"
|
|
Largo
Sport climber
The Big Wide Open Face
|
|
|
Topic Author's Reply - Jun 9, 2015 - 07:05pm PT
|
MH2. "Befuddled" means confused, in normal usage. My contention is that most AI geeks have totally confused sentience with data processing. I suspect you have another understanding per these fundamental processes and it might be interesting to hear your take on them. But without specifics per what was actually said, it is hard to judge your sincerity and verity.
I advance a basic idea - that AI confuses data processing with sentience. It would be interesting to hear other people's actual ideas per this fascinating subject, as opposed to silly dismissals that we should leave such issues to the "experts," or attacking the idea sans specifics. This is hardly the tactic to advancing understanding.
JL
|
|
jgill
Boulder climber
The high prairie of southern Colorado
|
|
My contention is that most AI geeks have totally confused sentience with data processing
Dumb asses.
;>(
|
|
jstan
climber
|
|
it is hard to judge your sincerity and verity.
Wow!
Humans are pretty enamored of what is going on behind their eyeballs. So much so they even conclude that is all there is. I'll give you an example.
Invariably when a high technology line goes down, the supporting engineers who have been focussed for years on what is in their brains, do designed experiments based upon what they think is happening. They assume they know what is happening even though nature has given them gigabytes of assembly data trying to tell them what is happening. Out of our ego we ignore what nature is screaming at us. Late at night I can swear I have heard, "Listen to me, dammit!"
A week of crawling through assembly data is generally enough to point out probable causes. By then the designed experiments have had so many lot splits nothing can be learned.
We are ego driven. I don't expect we will give the real machines much competition. Our existential risk will be reduced only if we program the machines to have egos.
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
So when a photon "scatters"...
perhaps this is part of the issue with quantum mechanics, we try to bring our experience to the table.
What was mostly described in Blu's rambling question was the idea of geometric optics; that rays of light refract and reflect from the source to our eye. The optics in our eye creates an image on the retina and allows us to sense the scene.
Now if we put something that will scatter light in the path, like smoke, then there may be so much scattering that we wouldn't see, say El Cap as his example. The light on path to our eye scatters off the smoke and doesn't make it... if the smoke is thick enough all of the light scatters, the image we have is one of the smoke, not El Cap.
Einstein's light quanta makes light seem like a particle... and in fact treating light as "quanta" solves a number of interesting problems, one of which was the nature of the black-body spectrum, another is the relationship between the absorption, transmission and spontaneous emission of atoms... and on and on, in some ways Einstein's light quanta was the beginning of quantum mechanics as we know it...
So it is easy to take these light particles and have them go on straight line trajectories (or curved space-time trajectories) analogous to the light rays we're familiar with from geometric optics, but that's not correct.
The light emitted from a distant star has is emitted with an imprecise product of it's momentum and position. This is famously the Heisenberg uncertainty principle. If we are very far away, the imprecision of the momentum means that the photon could take any of a number of paths to us, paths that diverge enough to be larger than a galaxy's diameter. The gravity of the galaxy may be sufficient, at it's distance, to bend the light along any of these paths to "focus" on the Earth and be observed by us.
Quantum mechanically, the photon is in an infinite number of states representing all of the paths from the distant star to Earth. And we have to describe it as a superposition of all those states in order to calculate the probability that we detect it. However, if something happens to the light along some of the paths, the state is altered.
We define the state of the photon to be, initially, a coherent sum of all the possible states, but intervening interactions, scattering by dust and all that, causes that state to become incoherent. The photon stops behaving quantum mechanically and behaves like the light rays of our geometric optics.
There may still be enough coherence left in the state to observe the quantum mechanical aspects of the light. This is not a new idea, it was used by Hanbury-Brown and Twiss to measure the diameter of stars by observing the quantum mechanical interference of light emitted on the limb of the star. The idea works for other quantum mechanical systems, like pions emitted from a particle reaction.
http://en.wikipedia.org/wiki/Hanbury_Brown_and_Twiss_effect
If the likelihood of scattering along the paths is large, the effect is extinguished.
Using this quantum mechanical idea on a cosmological scale requires that the universe is mostly empty... as it is. Considering how much stuff is in the universe this is a startling revelation.
But back to quantum mechanics...
we'd like to think of things as being totally separate in a particle-like isolation, but quantum mechanics tells us that we have to consider these things to be some sort of extended thing. When we make a measurement of this extended thing, we measure all of it, it looks like a particle with some particular properties we can determine.
If it is born along with another thing, then life gets even more complicated, as we have to consider the coherent sum of the two...
But considering it two is already taking us down the path of confusion, as the state of the two is really only one thing...
Because this is such a foreign idea, we resort to "interpretations" of the theory to help us comprehend what is going on... we want this theory to inform us as to the "true nature" of the world. If we listen to what the theory is telling us, it might well be that it is telling us that such a thing as the "true nature" might not be attainable, at least the way we'd like it to be... it does that by providing us with the most precise way to calculate the outcome of an experiment, even when it doesn't let us know, with certainty, what is happening at each step of the way.
As odd as that sounds, it seems to be the way things are...
Einstein didn't like it one bit... but we don't always get what we want.
|
|
Jan
Mountain climber
Colorado, Nepal & Okinawa
|
|
jstan's remarks remind me of what students of mine in the U.S. Air Force used to tell me about fixing airplanes. "There's two ways to do it", they said. One, usually an enlisted guy or gal, walks around the plane, listening, looking, smelling, touching and saying to himself or herself, "If I was this plane with this problem, where, would I be hurting?" Meanwhile in their minds, the really good and experienced mechanics could see the whole hydraulic system in their mind, or the electrical circuits and panels etc. Suddenly, they would say, "I think I've got it", and go right to the problem.
Meanwhile, the less experienced problem solver (usually portrayed as a first lieutenant straight out of the Air Force Academy), would say, "We have to start at the beginning of the manual", and would still be on Page 2, paragraph 3. 2.1.a.b etc. methodically checking, while the crew chief had already solved the problem.
|
|
BLUEBLOCR
Social climber
joshua tree
|
|
^^^^ Yep!
and
Thanks for the experience Ed!
|
|
healyje
Trad climber
Portland, Oregon
|
|
Demis Hassabis isn't vaguely interested in sentience or consciousness, he's strictly about machine learning and more specifically generalized learning capabilities. He's looking to our brain / mind to try and understand how we do generalized / adaptive learning and how we do it at such a low energy budget - i.e. he's studying neuroloscience for potential models of generalized learning. He's doing that because we already do it, we have a least some ideas about the functional roles of some distinctive brain structures / regions, and because it's likely far easier to tease learning models out of our own behavior and than to just come up with them out of thin air.
Also, when he talks about dangers / ethics, he's again not talking about sentient or conscious machines; they are in no way necessary to pose grave danger to humanity, misapplied machine learning by itself is more than sufficient to that task.
Ergo your use of metaphor to explain QM - that what is being said is "like" something else, is of the mark by a few light years. If your understanding is that the experiments Down Under did not serve up literal truths, then the task is yours to explain what is literally meant by what was said.
I already did, but you're just so into your belief system you don't seem to be able to 'get it'. No conscious observer is necessary. Or, in Truscott's own words:
"I can't prove that isn't what occurs," says Truscott, "But 99.999% of physicists would say that the measurement – i.e. whether the beamsplitter is in or out – brings the observable into reality, and at that point the particle decides whether to be a wave or a particle."
Note he's talking "beamsplitter" - i.e. a 'measurement event' - not a conscious observer.
|
|
MikeL
Social climber
Seattle, WA
|
|
Jstan: Couple of complementary presentations dealing with the existential risk facing us.
There were posts below each video that suggested other things.
It’s surprising to me that these research studies don’t stop to say what they think consciousness IS, first oF all. I have to ask: how could one know what consciousness is unless one took a close look at it subjectively?
Too many theories running wild. (Correction, . . . that should be "theoreticians," not "theories.")
MH2: One way to escape criticism is to take refuge in contradiction.
I think you’re saying or implying that non-dualism presents a contradiction? If that’s so, then you present an unusual view of "One." There would seem to be no contradiction for so many reasons: there could be no movement, no comparisons, no others, no context, and no content. It’s a singularity: all is One.
Largo: AI confuses data processing with sentience.
Most likely right. Cognitive science has largely relied upon a computer metaphor to model cognition. However, there were always bugaboos that haunted those models. How does one move from an initial starting position to the understanding of the first thing? How could a computer consciousness *learn* its first language? If language is inherently inconsistent (see Derrida’s works), then how can one fill out a set of knowledge structures that would give rise to contradiction? Mind seems to work using metaphors, but mind appears strangely disembodied (mind has no need for a body). Last, how does sensory input or perception get *translated* (“traduced”) into cognition?
(There are other problems.)
Jan:
You’d like the work done on expertise by cognitive scientists. So-called knowledge representations do not yield themselves to full mapping, because people know more than they can say or articulate (See Polyani’s or even Wittgenstein’s works). Unfortunately, many in the “anti-woo” camp will not admit to any other structures than semantic, procedural, or episodic knowledge structures as content and process of “knowing.” (All of them are particularly discursive.)
There are other ways / means of knowing, as there are also different kinds of knowledge (it would seem). Emotions, instincts, narrative, archetypal, and unconscious means of knowing are all effective means of knowing. What the content of those means are, is difficult to say. What is anger, love, sadness, envy—as well as what is “fight or flight” or “the myth of a hero” anyway? Why do they compel us, why do they provide us with insights, why do they explain things to us that apparently can’t be said?
I suspect that we’re being bombarded or almost overwhelmed with input / understanding that we just don’t fully grok because we can’t assimilate or put it into familiar ways of seeing or understanding (aka, taken from different states of being).
|
|
healyje
Trad climber
Portland, Oregon
|
|
Largo: AI confuses data processing with sentience.
Crikey, as the one fellow's blog states it: 'Not Even Wrong'.
You can keep asserting this nonsense, but I can assure you you're both wrong and so hung up on your interpretation of the term 'AI' that you're clearly not going to get it. No one in 'the business of AI' confuses sentience and data processing. No one in the business other than a couple of cranks is even interested in sentience or consciousness beyond personal musings. And no one in the business is working on 'creating' sentience or consciousness, they're focused on machine learning.
It borders on a perverse attachment to a strawman you just can't seem to let go of despite the fact it has zero validity.
...instincts...unconscious means of knowing are all effective means of knowing.
Are 'instincts...unconscious means of knowing' knowing?
|
|
Largo
Sport climber
The Big Wide Open Face
|
|
|
Topic Author's Reply - Jun 10, 2015 - 09:31am PT
|
Healje, I have provided many examples of the accepted difference between machine learning and machine consciousness, just to make the distinction clear. What's more, wonky projects like Markham's human brain project
(total costs are estimated at €1.19 billion, of which €555 million would go to personnel, to compensate 7148 person-years of effort) and funded by the European Commission through its Future and Emerging Technologies (FET) Flagship grant. 113 institutions across Europe are now involved as partners and another 21 as collaborating partners and will between them receive up to one billion euro in funding over ten years. This is not small time stuff and it was Markham - a total crank in my opinion - who said his machine wold love and dance and all the rest by 2022.
This is not AI, specifically or generally, but there is no question where most AI geeks get their cues (I ride with three of them twice a week). take Christof Koch, chief scientific officer of the Allen Institute for Brain Science in Seattle. Koch believes consciousness is in the fabric of the universe itself, is an intrinsic property of matter, just like mass or energy. "Organize matter in just the right way," says Koch, "as in the mammalian brain, and voilà, you can feel." In other words, Koch is in fact a garden-variety materialist. He, like many others, conflate data processing with consciousness. When asked if a computer can be conscious, Koch said:
I gave a lecture [last week] at MIT about Integrated Information Theory, developed by Giulio Tononi at the University of Wisconsin. This is a theory that makes a very clear prediction: it says that consciousness is a property of complex systems that have a particular “cause-effect” repertoire. They have a particular way of interacting with the world, such as the brain does, or in principle, such as a computer could. If you were to build a computer that has the same circuitry as the brain, this computer would also have consciousness associated with it. It would feel like something to be this computer.
Main stream AI is not, in large part, even interested in anything but machine learning, as I have repeatedly said. When I say that the few folks interested in machine consciousness are largely conflating data processing with sentience, I am basically talking about so-called computationalism - "the theory that the human brain is essentially a computer, although pre-sumably not a stored-program, digital computer, like the kind Intel makes. AI explores computational models of problem solving and learning, where the problems to be solved are of the complexity of problems so lived by human beings. An AI researcher need not be a computationalist, because they might believe that computers can do things brains do noncomputationally. Most AI researchers are computationalists to some extent, even if they think digital computers and brains-as-computers compute things in different ways."
"When it comes to the problem of phenomenal consciousnes, however, the AI researchers who care about the problem and believe that AI can solve it are a tiny minority."
However though few AI folks are interested in machine consciousness, the recent seminar in Puerto Rico on the dangers of AI broadly assumes that A) consciousness/intentionality will somehow simply arise from a data processor who "knows" how to learn, and B) this consciousness will be vastly "better' and more powerful (can crunch more data) than a human brain could ever imagine.
What is clear is that while most AI folks are only pursuing machine learning, their trust that machine consciousness will someday emerge from a computational model is the case nearly to the man. The one point where I agree with Healje is that people believing computational models are the royal road to sentience are "cranks," not because they are not astute in their field, but rather because they are not even looking at sentience in the first instance, rater at computations.
JL
|
|
Ed Hartouni
Trad climber
Livermore, CA
|
|
|
Jun 10, 2015 - 09:35am PT
|
...conflate data processing with consciousness...
but your major criticism of this line of reasoning is that they have not yet been able to describe "experience"
you have not offered any other criticism. and while it is true that the language doesn't currently exist to offer a description, that, in its self, does not constitute a refutation of the basic idea.
do you have more to offer than that?
|
|
MikeL
Social climber
Seattle, WA
|
|
|
Jun 10, 2015 - 09:48am PT
|
healyje: Are 'instincts...unconscious means of knowing' knowing?
If people know more than they can say, if they act without thought, if they do not know why or how they did something . . . then I’d say so. But who can say?
It’s not my experience that colleagues who were into AI and cognitive science when I was at the U of Illinois were “cranks.” Many of them were beginning to think intelligence, awareness, or consciousness as something other than mere data processing. As I tried to say above, data processing relies upon a tired metaphor of brain-mind-body (a computer). Instead, most of my colleagues and professors were particularly concerned with meaning and how it gets translated into and out of knowledge representations, running various non-equivocal transformations (transductions), language, reading, etc. They were not meditators, religious, or new-age spiritualists, but they seemed more open-minded than you are about it.
If you make a machine that learns, you must model learning. It might look simple, but the work I saw was anything but. Moreover for machine learning to be useful, it must somehow interact with people. I remember a software entrepreneur telling me that if his team did not know or understand what was in the minds of people who used the system—even before they approached the machine—the system would fail.
How can you “understand” anything without coming to some strong ideas about what consciousness and sentience are? To read most of the comments here in this thread, it would appear that consciousness and sentience ARE merely input - output measurements to and from a black box called “mind.” In other words, there would appear to be NO consciousness or sentience to begin with, . . . just transformation functions.
If that’s the stand you’d take, then I’d question what you think “knowing” would be. What or who could there be to “know?”
|
|
|
SuperTopo on the Web
|
Let us know!