Will machines become conscious someday?

Discussion in 'Intelligence & Machines' started by Magical Realist, Sep 19, 2012.

  1. spandrel Registered Member

    That's why I mention the Culture space ships. They're a good case study. They seem to know all about us squishy organics and share our sensibilities to some extent. How realistic is this? Would they not have a completely different form of consciousness. not being squishy organics and living for the most part in space? Hard to say.
    Universe cannot exist without consciousness. A nice conceit because it's not provable, by definition. 'The minimum number of Universe is two' said Buckminster-Fuller. The experiencer and the experienced. So maybe it doesn't matter what form it takes. In fact the Culture space ships would seem to be one step ahead of us, their brains exist in some sort of hyperspace.
    I don't seriously think the Internet routers-as-neurons could be conscious. As I've mentioned elsewhere I am a Penrosian in that respect. No algorithmic machine can be conscious, in which case our own brains are not just neural nets. Something else is going on, maybe quantum s**t. And as the latest quantum computers are just extensions of algorithmic computing models I can't see any chatty iPads appearing any time soon. It's not so much the technology as the paradigm. A sea change at MIT? No chance. Too many vested interests.
  2. Google AdSense Guest Advertisement

    to hide all adverts.
  3. ElectricFetus Sanity going, going, gone Valued Senior Member

    The thread question is "will machines become conscious someday" not "Will computers become conscious somday" we know as a matter of fact that a "machine" with nearly 100 billion analog spiking neurons and a quadrillion synapses operating at a pitiful 1000 spikes a second tops can produce consciousness.
  4. Google AdSense Guest Advertisement

    to hide all adverts.
  5. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    The two aliens on fact finding mission to Earth, did take a few life forms on board their spaceship to learn how they worked. After completion of the study, finding no electronics were used in the creatures one said to the other: How is thought possible in eatable meat? Other said: It isn't, so if we put that in the report, they will put us in the loony bin and our careers are over. First replies: Ok, So we agree to report there is no intelligent life form on Earth only an abundance of eatable protein.
  6. Google AdSense Guest Advertisement

    to hide all adverts.
  7. ElectricFetus Sanity going, going, gone Valued Senior Member

    Yeah I read the short story too.
  8. spandrel Registered Member

    Despite that fact that definitions of machine assume artifact, let's accept that life-forms are also machines, or at least mechanical. I sense that you are a strong-AI believer, in that if we could replicate the electrochemical activity of the brain then it would be conscious. This position presupposes that consciousness is an emergent property of this activity. There is no evidence for this. The brain is active when we are unconscious and can be inactive when we are conscious. Neuroscientists are having a hard time pinning down consciousness in the brain and generally prefer not to deal with it.

    If your belief makes you happy then I wouldn't dream of challenging it, but this forum is a place to discuss hypotheses so it doesn't help to state articles of faith as if they are obvious truths. (Smiley face.)
  9. ElectricFetus Sanity going, going, gone Valued Senior Member

    There is no evidence it is not, so far we have managed to prove just about everything operates by physical laws limited to this universe, no evidence of anything supernatural, ergo replicating the brain should be possible, if not then we prove the existence of the soul, either way: discovery!

    By all means can you cite a study that has details this phenomena you describe of the brain being inactive while conscious.
  10. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    No, not evidence for a soul, but proof that the brain develops it powers ONLY with a body that is in developmental interaction with an environment.

    For example (1):
    Two cats, taken at birth and placed in baskets which hung from different ends of a beam with central pivot, but one basket had holes in bottom so that cat could walk and propel the “cat merry-go-round.” Both looked at (note I did not say “saw”) exactly the same design on the nearby circular wall surrounding the “cat merry-go-round.” In some experiments, there were only regular vertical stripes, others with horizontal only and still other more complex patterns.
    In cat, there is a quite short period (less than 3 weeks, as I recall) after their eyes open in which their visual cortex can become normally organized. The cat that did not walk, did not ACTIVELY interact with its visual environment was to some extent blind for life. (Cat brains are not as "plastic" as human brains.)*

    For example (2): Bach-y-Rita, with whom I have exchanged some Emails,** is the world's foremost expert on “sensory substitution / neural plasticity.” Here is a list of 135 of his papers: http://tcnl.bme.wisc.edu/library/articles/paul-bach-y-rita
    He ALWAYS places great emphasis on the need for active, not passive, interaction with the environment for brain development. One of his main areas of study was with the congenitally blind. Could their brains learn to see, but via tactical stimulation (usually on their backs as a large array of skin contacting electrodes was required for even very crude image resolution.)

    Weeks of passive stimulation (sitting in a chair with the electrode array in the chair's back contacting the bare skin) were always perceived as only tactical stimulation of the back, but after even only a few hours of freely exploring a room with head mounted images source camera driving the electrode area, changed that tactile perception into a visual one. This is not just the subject's report. Bach-y-Rita could test it and did, via remote control of the camera's magnification. I.e. Un known to the subject who had developed “crude tactile stimulation of back vision” when he was looking directly at some moderate sized object the image (stimulation pattern static on his back) quickly, but steadily grew larger as if the object were hurling towards the subject's head. Subject did exactly what you or I would do – automatically ducked to not be hit!

    * Congenitally blind (from birth) human adults who have an operation restoring vision are also never able to get completely normal vision. They NEVER get even reasonable depth perception for objects far beyond their reach. - May initially think it OK to step out of a 14th floor window to the "nearby" ground but quicky consciously learn that the "tiny toy cars" parked are real full size cars far away. Some never recover color vision, but many have some color vision.

    Some other animals, including monkeys, have had their eye lids sewn closed at birth but the eyes must remain closed much longer (many months, as I recall) for them to end up blind for life. Humans too have defined periods for brain development, but after 20 years of living here in Brazil, I can speak Portuguese about 20% as well as a Brazilian 2 year old. Like him, I never studied - just picked up what I heard. Interestingly I don't translate from English. The words, bad pronounced, just flow out of my mouth and I have no idea what they will be until I hear them!

    ** I had some interaction with NIH researchers working on directly electrical stimulation of the V1 with micro-electrode arrays. (A program lead by a Dr. Hambrick, if memory serves me after ~40 years). They did not understand why the stimulation of nearby electrodes "bloomed" - destroyed their resolution. I did as I was at time stimulating rhesus monkey brain in a study related to epilepsy and knew about "Jacksonian epilepsy." Their "blooms" decayed (after stimulation turned off) in a fraction of a minute, just like Jacksonian epilepsy does (and how all epileptic seizures do, I think, - by "metabolic exhaustion" of the self activating cells). I wanted to encourage Paul in his approach and assure him that the NIH, sort of his much better funded competitor, was "barking up the wrong tree."
    Last edited by a moderator: Jan 12, 2014
  11. ElectricFetus Sanity going, going, gone Valued Senior Member

    That a silly conclusion as we could easily test for that by giving a machine an artificial, cybernetic or virtual body. No if we can't replicate human level consciousness of better despite having ALL the parts, then we must conclude their must be some missing part beyond the physical realm.
  12. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    No it is a demonstrated fact and not only by the two experiments I mentioned, but by many dozens. Some of which I referred to briefly. The brain and the body's sensors MUST co-develop co-processing the environment and in the case of even advanced brains (cat or human in my two detailed examples) there are critical periods, built into genetic controlled development, when various aptitudes / abilities can be normally acquired and in lesser than human, some are ONLY possible then.

    At birth, your brain is sort of wired up very locally with random connections. In the first few years of life you will REDUCE then number of synaptic connection drastically (by ~40% I seem to recall) - The ones that were not useful are eliminated.

    So if they are not useful because your eyelids were sewn shut during the period when you normally would have been learning how to process visual data, you will eliminate those connections that should have been refined. I.e. in lesser primates where these experiments have been done - They never learn to see even close to normal.
  13. ElectricFetus Sanity going, going, gone Valued Senior Member

    And I was not challenging it, we can make a body for a machine, so honestly I don't know what your arguing about.
  14. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    I'm not arguing, just trying to correct what seems to be your POV that if a brain could be perfectly replicated to duplicate every neural connection it would be a "mental clone" of the original person, if also placed in an identical body. etc.

    It would not because the sensory input history would not be the same as for real brain in its body naturally learning. Real brains have CO-DEVELOPED WITH THEIR SENSORY IMPUTS ON A GENETICALLY DETERMINED SCHEDULE.- have a learning history. - That is "what is missing" (not a non-physical part / soul, etc.) from the new robot with body produced in the 21st century advanced factory.

    In the first year of life, when something like half an entire life time of learning takes place, a large fraction of the learning process is the ELIMANATION of not useful locally random and thus UNIQUE to each individual neural input signals to brain cell connections (and brain cell to cell connections).

    Perhaps knowing something about neural networks (I hate that false term, for "connection computers") will help you understand. Typically they are simulated in a digital machine and the initial weights of the connections are set by the computer's random number generator. After the connection computer has been "trained up" on the sample set and the connection weights are stable, it may do very well on closely related problems - much like a baby in first year getting rid of many initially random neural connections in it brain.

    And just like the baby if you start over again to train up a second connection machine for the same problem types, etc, every thing the same EXCEPT FOR THE RANDOM INITIAL WEIGHTS, the second connection computer will do well on closely related problems OFTEN WITH AN ENIRELY DIFFERENT SET OF CONNECTION WEIGHTS. I. e. like a baby learning how to walk or control the acid flux in its stomach after eating, etc. the particular set of surviving random neural connections (on very local level) is UNIQUE as it CO- DEVELOPED WITH THE HISTORY OF THE SENSORY INPUTS.

    SUMMARY: If you want to build a human like robot, not only do you need to give it a body with sensors for brain input, but it must grow up learning how to modify the initially locally random connections between brain cells (to other brain cells) and to the sensory input network. Again becoming human is a CO-DEVELOPMENT, MULTI-YEAR (IN HUMANS AT LEAST) AND UNIQUE DEVELOPMENT PROSCESS, mainly initially one of ELIMINATION of not very useful connections.

    Almost all have the idea that learning is the formation of new neural connections, and that is true in later life, but mainly false in the early years when half of all brain leaning takes place.* On a detailed local level there is no resemblance between brain tissue connections either to other brain cells or to the sensory imputs fibers even in identical twins who were raised almost exactly the same. In wiring up a brain / learning there are many more than one way to "skin that cat."

    * What you consciously know / have learned / is a very tiny fraction of what your brain learns.

    BTW I say nothing about consciousness, as can not even define it, except I strongly suspect evolution has selected for it.
    Last edited by a moderator: Jan 12, 2014
  15. ElectricFetus Sanity going, going, gone Valued Senior Member

    I didn't think I was saying such a thing, rather that if we made a machine based on the principles that we know works in the brain we, a spiking neural network with that many neurons and synapses we should be able to raise it to consciousness, via a body, real world interaction, virtual, what ever. If you want to talk about the concept of mind-uploading I would say your jumping the gun, we need to prove an artificial construct first.
  16. spandrel Registered Member

    Billy T, I sincerely hope that machines do not become conscious as they may decide in their arrogant superiority that it is all right to strap you to a bench and poke electrodes in your brain, or put you in a basket and force you to walk in circles.

    Apart from that you are still talking about the brain and have declared that it is not a separate entity to the body. I agree. I think ElectricFetus would have a job creating a machine body that would grow and develop as we did. Can you remember being children? It is all about interaction with the environment so I would suggest any machine will only gain intelligence through this process.

    As to consciousness being possible without brain activity it is quite possible and even recommended by some. It is possible to switch off the brain, although the mystics of many religions would call it stilling the mind. A short course of meditation would demonstrate this. Unfortunately it is not a phenomenon that lends itself to objective study, except perhaps by fMRI scans and the like, but it certainly improves the efficiency of the brain if it is given a rest now and then. What is curious is that it is possible at all. I would guess that there is an evolutionary advantage to shutting the damn thing off now and then. It does get quite bothersome sometimes, don't you find?
  17. ElectricFetus Sanity going, going, gone Valued Senior Member

    Yeah, yeah aah you have any links to scientific studies on this, if not is pure new age BS conjecture. Now mind you I enjoy meditation and sensory deprivation but my consciousness most certainly does not stop.
  18. spandrel Registered Member

    Time is an illusion, lunchtime doubly so.

    Oops. It's clear from this that you have a strong belief that consciousness is an emergent property of brain activity. I didn't say that consciousness stops but that brain activity stops. You believe the two are the same, thus the mistake. I have no beliefs one way or the other. Belief has no place in any scientific investigation. I'm happy to accept that consciousness might associated with brain activity, but as I said, it is absent during busy activity in sleep and present when the brain is quiescent.

    I'm interested in the concept of consciousness as a quantum phenomenon (see http://www.quantumconsciousness.org/ ) but that's a whole nother kettle of bicycles, as they say. There is enough leeway in our current understanding of the quantum field potential in the universe to posit a distributed model of consciousness which might chime with the ancient teachings of oneness with the universe. New Age? Nope, consistent throughout recorded history. It's churlish to reject a concept because it's not popular or challenges your belief. Again, it's unscientific. You must at least be a bit curious about the commonality of such ideas across time and cultures, and the ever-increasing convergence with concepts in physics. It is at least demonstrable that the intervention of conscious observation can change the state of a particle. If you need references for that I'll dig some out.
  19. ElectricFetus Sanity going, going, gone Valued Senior Member

    Do you have any proof that the brain's activity stops?

    The present empirical evidence for what you describe.

    Oh anything possible, but what you describe is not proven or even likely.

    Sure and there are simpler occam's razor answer for that: Human spirituality is a consistent property through out time and cultures because it is the product of common human psychology trying to make up answer for unknowns.

    A non-conscious observer can change the state of a particle, any interaction with a particle (and all observation is an interaction with particles) changes it state. Consciousness is nothing special in quantum mechanics.
  20. Read-Only Valued Senior Member

    This is just pure BS. Brain activity does NOT stop - ever - until life ceases! There's absolutely no point in reading any more of your childish posts.
  21. spandrel Registered Member

    You are quite right, childish nonsense, but at least it prompted you to post here, Read-Only.

    Please Register or Log in to view the hidden image!

    Could we perhaps hear of your opinions on this subject?

    As to your points ElectricFetus (all quotes), in reverse order -
    There is no such thing as a non-conscious observer. By any definition (and I've looked at a few) it requires a person.
    Fair enough. I apologise for using the word 'intervention' in a broad sense. I was meaning 'quantum-intervention' before an observation is made. A photon, for example, cannot be observed, only the results of its interaction with some form of detector, a screen in the two-slit experiment for example. (This spins off into a discussion of whether a photon exists). In that sense there is no direct physical intervention. The decision to see the event one way or another has an effect on the particle before the observation. This is an intervention albeit at a distance, and not of a classical form. There are many examples of this, here is a good one. quantumenigma.com/wp-content/uploads/rosenblum-kuttner-consc-obsvr-q-exp.pdf This effect is verifiable by anyone.

    You are confusing spiritual experience with the dogma of organised religion. But I concede that spiritual experience could be a function of the brain, but we'll have to wait for the evidence before jumping to conclusions.
    (cosmic or distributed consciousness)No proof of course, but an interesting idea. I refer you to current understandings in the science of cosmology.https://www.math.auckland.ac.nz/~king/Preprints/pdf/tuz6.pdf
    As to your first comments, and in answer to you as well Read-Only, I have to admit I went too far. Obviously brain activity does not stop. It was provocative of me. However it cannot be denied that brain activity continues when we are asleep. As to whether it is possible to suppress activity when conscious, then there is only limited empirical evidence. It is a difficult experimental problem.

    Many studies have been done on subjects during meditation which show that consciousness can change the activity of the brain directly. Most research is funded either by organisations wishing to promote the benefits of meditation or those involved in identifying the function of particular areas of the brain, including psychiatric institutions. Most of the studies use subjects who practise mindfulness meditation (often used as part of cognitive therapy) and compassion meditation. These types are concerned with identifying and monitoring different thoughts and feelings, and learning to either control them or ignore them. It is no surprise therefore, that the relevant areas of the brain are activated. However some areas are de-activated. The results are varied across studies but some commonalities are observed. Here are a few

    http://www.sciencedaily.com/releases/2011/04/110405174835.htm about focused attention
    http://www.cmiv.liu.se/research/current-research-projects/fmri-meditation about compassion meditation
    hms.harvard.edu/sites/default/files/assets/Sites/Longwood_Seminars/ScienceofEmotion3.20.12.pdf about the science of emotion
    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2944261/ about open monitoring meditation
    http://psychcentral.com/news/2011/1...itation-tunes-out-some-brain-areas/31770.html about the default mode network
    http://www.frontiersin.org/Journal/10.3389/fpsyg.2014.00099/abstract about eeg reductions
    http://www.bme.ufl.edu/labs/sitaram...xperiments-on-Meditation-Erb-Sitaram-2010.pdf posterior cingulate de-activation

    There are problems inherent in MRI technology. It does not show actual neuronal activity. There is no baseline state possible for comparison so results are always comparative w.r.t. other subjects. The variability of subject brains, due to plasticity, limit precision and certainty in results. Despite these limitations it provides the only method to investigate brain activity. I would rather these machines were used for diagnostic work rather than to satisfy my idle curiosity. Maybe there'll be a handheld version soon.

    What is clear is that consciousness can control the activity of different area of the brain. This is problematic when trying to identify where consciousness resides. More research is needed. Present theories identify the thalamus as being involved in promoting consciousness by cyclical stimulation of the cortex, but it is not suggested that it is the seat of consciousness.

    And from way back
    'No evidence' is the sort of thing we hear from politicians. If a reconstruction of a brain does not exhibit consciousness it doesn't prove anything. It certainly doesn't show anything supernatural is going on. Supernatural just means it hasn't been explained yet. Magic is only magic until you know how it works. But your comments demonstrate that you are willing to go ahead with a certain line of enquiry despite the absence of evidence or proof. Great! That is true scientific enquiry. Nothing should be rejected until proven to be wrong.

    My purpose has been to draw attention to some of the problems associated with consciousness. It is these that need to be discussed here rather than the mechanics of computers. The three main points are:
    1. Is the brain purely algorithmic or is there a quantum mechanical component?
    2. How does conscious intervention in quantum events work?
    3. Is consciousness necessary for a universe to exist?
  22. Write4U Valued Senior Member

    IMO, it depends on how far down you want to go. At some point yes QM applies

    Look at Cern.

    IMO, only shared information is necessary for a universe to exists. This implies connectedness, not consciousness.
  23. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    On (1): The two are not mutually excusive - people are hard at work trying to develop algorithms for quantum computers (as well as those computers themselves)
    On (2): I doubt it does. You said in earlier post you would dig some support for it doing so up - please do.
    Just so we understand each other let me tell design of simple test: Two identical radiation detectors (1 & 2) are positioned near their separated identical radioactive sources with exact separations adjusted so that on average (done with hour-long test of each separation setting for days to get good statics and essentially identical count rates) detectors give say 50 detections per minute by each detector.

    After this "set-up" the consciousness enters and starts to interfere to make the rate of detection by 1 great than 2 much more than happens by chance during the test's one hour duration say by 5 sigma and the "consciousness" can repeat the performance, but reversed. I. e. in the next hour make 2's decays greater than 1's by 5 sigma more than chance would.

    On (3) No, not needed. The universe existed with no planets even for at least a 100 million years. - Much longer than human level consciousness is likely to, in my opinion.

Share This Page