Will machines become conscious someday?

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

  1. ElectricFetus Sanity going, going, gone Valued Senior Member

    Don't be a semantic nit-picker, any non-conscious interaction can change a quantum mechanical state, superpositions collapse by interaction with the rest of the universe, not simply because someone observed them, and slit experiment and photon interaction occur the same way regardless if anyone is directly observing them or not, we know this because those quantum mechanical effects leave and impression on the universal long after non-conscious interactions destroyed them. For example photon jumping in chlorophyll happens, has to happen for photosynthesis, no one observing it directly. Likewise what makes a photon or an electron choose a slit or break an interference pattern is not a conscious observe but interaction with the slit or a detection mechanism its self: if information is retained of which slit the photon or electron passed through. No conscious observer is need retain or know of this information, the detector it self does not need give that information to a conscious observer, just the unconscious detector is all that is needed.

    Occurs razor, we need proof it is not a function of the physical world (the brain) not the other way around, we can assume spirituality once we can rule out the simpler physical answers.

    I would say that being able to alter brain activity will asleep or conscious is evidence for consciousness in the brain, not against. Alter states of consciousness are matched with alterations of brain activity, alternatively the consciousness of dreaming has been well documented with brain activity of sleep. What your suggesting is the consciousness changes brain activity and not that brain activity changes consciousness, all that you present can't prove one comes before the other. Consider meditation, attention focusing, hypnoses, all of these alter consciousness as well as change activity in the brain and rightly so if consciousness is activity in the brain! What you need is proof consciousness can exist independent of the brain, say for example an experiment in which people in near death experiences who have no EEG reading later are able to recount a series of numbers screamed at them while they were brain-dead.

    On the other hand we know through the very extensive history of brain damage and brain disorders that consciousness can be altered and manipulated in predicable manner via neuroscience and in ever greater detail as neuroscience gets more advance. If consciousness exist independent of the brain how come we can alter consciousness by altering the brain?

    This is rather philosophical but I believe the who Chinese room ins conscious, not merely any one part, consciousness is not simply there or not there, but has gradients, levels of awareness. Remove a person occipital lobs and they are not simply blind but can't even imagine sight anymore, they are no long conscious of vision beyond an academic knowledge of it. Likewise damage to a front lobe can utterly change a person personality, if there is some kind of soul does it too change with the brain damange, then what use is it?

    A green tiger can pop on top of your from another reality at any moment... prove me wrong. Technically anything is possible, science though is about predicting what is probable and improbable. A soul beyond the physical workings of the brain is improbable by present scientific evidence, sure maybe it exist simply by the fact that it is impossible to prove something does not exist, but what value would the soul have if we could explain all that it supposedly imparts via the organization of analog spiking neural networks? Sure we are not there yet, but there is no reason to believe we won't be, unless you can prove this soul really exist and is necessary for consciousness.

    I'll let Billy T answers stand for those last questions.
  2. Google AdSense Guest Advertisement

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

    A better continuation of this sentence would be:
    ... The interference pattern can fall on sheet of photographic film. That is soon "developed" by an automatic processing machine as 99+ percent are now days. If by chance it falls behind the machine and no one sees it until 15 years later when the machine is replaced - the interference pattern is there in the form of stabilized silver compounds that have not changed during those 15 years.
    Permit me to clarify this as it reflects / implies / too much normal human POV. The electron or photon does not chose which slit to pass thru. It goes thru both if one must speak in the human terms.

    Also if some knowledge is gained (even by just a recoding machine) that shows it went thru slit "A" and not slit "B" as if a classical particle, then the interference pattern is destroyed. I.e. a somewhat Gaussian intensity pattern will strike the photographic film or screen, not a set of parallel bright and dark lines.

    It may not stretch human belief too greatly to admit the photon goes thru both slits is correct as normally the slits are close together. So I'll tell of an experiment I have done to measure the length of photons. My lamp's "Sodium D" photons were abut 30 cm long, if interested but some from other sources are more than a meter long. (Those forming the green light of the Northern Lights, are at least 2, probably more than 3, meters long!) None are "little balls of energy" like 99% of people with any opinion think they are.

    It is well established that each photon ONLY interferes with its self - That is why more than 100 years go, when no coherent light sources, like LASERs existed Young could do the two-slit interference experiment. To show each photon interfere with its self one uses very long exposures and a very weak light source. So weak that most of the time there is no photon in flight between the source and detector yet exactly the same interference pattern will slowly be collected.

    I used a "two path" interferometer to measure the photon length, but did not need to use low intensity light source as I knew each photon ONLY interfers with it self. In fact different photons don't interact, not even to just scatter one off the other as you can demonstrate with two flashlights* in a dark room with their beam crossing each other. Each photon in my setup first came to a 50/50 beam splitter (lightly silvered glass sheet). Those that passed thru were on path "A" and those that reflected off the beam splitter were on path "B" and the "B" photons soon hit a mirror at 45 degree angle. When they left it they were traveling parallel to the "A" photon, which soon thereafter also hit a 45degree mirror.

    I.e. with two 45 degree mirrors the photons were going around a rectangle and came together again at another 50/50 beam splitter at the diagonally opposite corner of the rectangle. Some "B" photos passed thru it and with the "A" photon that were reflected by it were both on the same path to the screen, I simply looked at to observe their interference pattern.

    Let me call the length of the travel leg (edge of the rectangle) "B" photons first traveled on by themselves B1 and that of the path they traveled on after reflection off the mirror at 45 degree angle B2. Like wise A1 is then length the "A" photon traveled after just passing thru the first beam splitter to a different 45 degree mirror, and A2 is the length of their path from that mirror to the second beam splitter they reflect off of to join the "B" photons on a common path to the screen.

    Note that A1+A2 = B1 +B2 for this rectangular light path = I.e. both photons travel equal distances to the detector. This produces the best interference pattern. I.e. there is no illumination at the center of the "dark lines" on the screen.

    Next I slightly rotated the first beam splitter which made no difference in the length A1 +A2 path, but caused the "B" photons reflected by it to not even hit the mirror the had reflected off of when they followed the edge of rectangle. Say I had to move that mirror 1 cm more distant from the second beam splitter and slightly rotate it too so that after the "B" photons reflected off it they were again on path parallel to that of the unchanged path of the "A" photons, but now B1 +B2 is slightly longer than A1 + A2. This path change weaken the interference pattern - I. e. there was now some light even at the center of the "dark" lines.

    Thus to speak falsely in normal human terms so you will understand: The first part of "A" photons, traveling a shorter path arrive at the detector before any part of the "B" photons do. This causes a less well developed interference pattern. I.e. there is some illumination even at the center of the still quite dark lines. With path A and path longer by 30 cm the screen was uniformly illuminated - no observable interference.

    With a series of set ups, of ever increasing path difference, I discovered that the interference pattern was complexly washed out when B1+B2 = A1 +A2 +30 cm. I. e. the illumination on the screen was uniform in intensity but only half the intensity it had been ate the center of the zero path difference bright lines. (Conservation of energy at work but I'll not say more on that except to note: sin^2 + cos^2 = 1, always and that the AMPLITUDES interfere.)

    My lamp's photons were 30 cm long. Note the when photon "P" arrived on a 30 cm shorter path the were "zillions" of other "B" path photons arriving at the same time, but P would have nothing to do with them - P could on interfere with its self and part of its self (speaking falsely again) was still in the interferometer - not yet to the screen.

    Now My point was that although the single photon going thru the double slits never was more than a few mm for itself, every one of my P photons was at times in the interferometer more than a meter away from itself! Nature does as nature does and does not care if that is impossible for humans to truly understand. With more than 60 years of knowing all this, it does not even seem strange to me any more.

    * If you don't have two flash lights, just look up at the stars some night. The light that has been traveling many years to fall on your retina and die as a photon is part of a very well defined "beam line" from the star to your eye. light from a Zillion, zillion other stars has crossed directly thru that "beam line" while those photons were traveling to your eye. If there were any momentum transfers, you would not see stars, but a nearly uniform "night glow sky." It is highly improbable that man would ever guess that we live in a huge universe when it is so obvious that we live under a strange "Night Glow Dome" - No doubt there would be struggles to explain where the "Night Glow Energy" came from, just as there were prior to any under standing of the source of solar energy. That ignorance was a great comfort to the anti-evolutionist of Darwin's era - Impossible for the sun to still be shining if the dinosaurs etc. lived million so years ago - Ergo God must have made it all perhaps ~5000 years ago as the Bible states.

    Nature has done the "two flashlight" experiment for you - only you did not understand what seeing stars implies. (A defect in your "consciousness" I hope now corrected as were others above. - Especially the false, but common, idea that photons are little balls of energy.)
    Last edited by a moderator: Feb 2, 2014
  4. Google AdSense Guest Advertisement

    to hide all adverts.
  5. Walter L. Wagner Cosmic Truth Seeker Valued Senior Member

    not to quibble with your analysis, but visible light wavelengths are quite short, not many centimeters long as you seem to imply. see: http://en.wikipedia.org/wiki/Fraunhofer_lines for the sodium D lines at D1 at 589.592 nanometers and D2 at 588.995 nanometers. This limits light microscopes to a resolution in the micrometer range, circa 1600X magnification.
  6. Google AdSense Guest Advertisement

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

    Quite correct, but I don't imply that. Yes the wave length is very small but I measured not it*, but the length of the photon. Some tables of wave length give them with 8 significant numbers. If you know any Fourier Analysis, then you know that this precisely defined wave length or almost "pure frequency" implies an extremely large number of EM cycles. How many can be estimated for my Sodium D 30 cm long photons by dividing 30cm by the wave length in cm of the Sodium D lines.

    The delta T * Delta E uncertainty principle is another way to understand (without any Fourier analysis) I.e. with precisely defined wave length Delta E (the uncertainty) of the photon's energy is very small so the Delta T (time when emitted or duration of the emission process) is relative large. I.e. to speak in crude classical terms, the Sodium atom that emitted one of my photons did so for a relatively long time - "pumped" out a "zillion" cycles, one cycle at a time, to make the long "photon string" of cycles for the whole duration of Delta T.

    * To measure wave lengths, one use a spectrometer, not an interference set up normally, but with a Fabry Perot interferometer one can measure the relative wave length, and I have done that too, in my Ph. D. experiment on Argon line widths and shift in a high density plasma. - The first to do so for lines from ionized Argon. I used the Fabry Perot to get the high resolution needed to separately measure within the line profile at different wave lengths in cascade with a regular grading spectrometer to separate the different FP orders.

    Argon has relative few strong lines or colors in the visible. Some times n wave lengths of one add to the same total length as m wave lengths on another and both come out of the FP together as a color that argon does not actually emit. My FP plates were quite close to each other and I notice a unique color when adjusting the optical separation (easily with constant physical separation by varying the gas pressure inside the box they were in)
    I.e. I saw that 6853 cycles or wave lengths the strong yellow line (5875.62 angstroms) passed by the FP with two others (blue and green) to make a unique color. I could , just by looking, know the optical separation* between the FP plates to seven significant figures!

    I.e. 6853x5875.62 = 8028x5015.68 = 9005x4471.62 are all are same length so pass all together thru the FP.

    * if I wanted to, but I never bother to, I could known the physical separation between my FP plates as the gas pressurizing the FP's box was very precisely measured (so I could precisely and very slightly move to another slightly different wavelength within the width of the line) and use the known refractive index to convert optical to physical length.

    I don't think many have ever been able to know a length to more than seven significant figure accuracy just by looking! But this was just interesting, not important for my Ph.D.
    Last edited by a moderator: Feb 3, 2014
  8. Walter L. Wagner Cosmic Truth Seeker Valued Senior Member

    my apologies. I misread your statement to measure the 'length' as meaning the 'wavelength'. interesting concept to measure the 'length' of a photon; i.e. the uncertainty in forward position using the speed as c.
  9. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    Just out of curiosity, were you among the 99% who, if they have any mental image of a photon, think of it as a little ball or packet of energy?

    If I correctly told someone, even most well versed in physics, that some photons are 3 meters long, they would ask: "What have you been smoking?"
  10. Walter L. Wagner Cosmic Truth Seeker Valued Senior Member

    No, I 'visualize' them as a 'packet' of vibrating electromagnetic radiation moving forward at c, with the electric field oscillating, inducing a magnetic field, in perpetuity until interaction with matter. The 'length' that you ascribe would actually be the region along the path of high probability of potential interaction (if you were to, for example, rapidly insert an electron along the path) , and the 'length' would technically extend beyond the distance you assert, is that not correct?
  11. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    Not much wrong with that but there does seem to be some confusion between our "lengths.". My "length of the photon" was an instantaneous POV, not the long path it will travel until interaction with matter.

    I am quite certain that the EM field is not a set of uniform cycles that abruptly start at one point of space (for T some fixed instant) and abruptly end some photon length along the travel path X cycles behind the start.

    Likewise I'm very vague on how wide the EM field is say be half amplitude lateral points but believe it is orders of times more narrow than it is long for most photons - nothing like little spherical energy balls.

    I should note that you seem to be making the same, very logical error I did for years - I.e. thinking that the collapsing magnetic field's dB/dt created the rising E field, but in fact the E & B field are in phase with both becoming zero at the same time (and peak at the same time)
  12. Walter L. Wagner Cosmic Truth Seeker Valued Senior Member

    well, you could start a whole new thread on the topic of 'length' of a photon. but yes, if the 'width/highth' is the wavelength, and visible light is on the order of a half micron in wavelength (0.0005 millimeters or thereabouts), and their 'length' is on the order of 50 cm, then it follows that they are much longer than wide. on an aside, did you check out the great pyramid thread? the idea of interior tunnels is intriguing. and lots of great vids of 3-d animation that was prepared, if you check out all the vids.
  13. ElectricFetus Sanity going, going, gone Valued Senior Member

    jesus nitpicker, try to simply explain how quatum mechanics does not need a coscious, sentient detector, and *boom* off-topic banaza.

    Back to the point, there is no evidence consciousness is not a physical process that can be replicated artifically, either brute force in a gigantic digitial turing complete machine, or via artifical analog spiking neurons made on silicon chips with an artifical neural network of synapses made of CMOS, memristors, spintronic transistors, or even half way inbetween via a turing complete quantum computer (if one is every built). If we do find consciosuness requires somethign which can't be replicated or understood, we will have found the soul and proved the spiritual exists, if on the other hand we one day create artifical consciousness we will have evidence that the soul either does not exist or has no important function. Of course it may always remain possinble to deny a strong AI or artifical general inteligence is truelly conscious, but merely behaves as if it is, of course it is equally possible to deny anything or anyone else is trully conscious as well and thus fall into solipsism.
  14. spandrel Registered Member

    Hi peeps, me again. Isn't life fun?
    I understand your frustration EF, but I would just say this. If you believe that brute force processing is all there is to conscious machines then by all means get a bunch of supercomputers and hook them together. In fact it is perfectly possible now to connect a few million computers over the internet (with appropriate neural connection simulations) and replicate a brain. Each computer could easily simulate a few thousand neurons. It would be slow but that doesn't matter in a simulation does it?

    I'm still griping about the assumption that a suitably fast and complex information processing machine will inevitably exhibit consciousness. Or that any sufficiently intelligent machine will inevitably develop consciousness (the Skynet thesis). It would seem to me to be a good idea to find out whether this is likely or whether consciousness is a necessary prerequisite for intelligence. What is intelligence? We haven't even discriminated between consciousness and self-awareness. I know some pretty stupid dogs but I wouldn't deny that they are conscious. According to Prof. Gordon Gallup there are only three self-aware species, all of them primates. Is it necessary for intelligence? Some crows and squirrels are pretty smart. Is there a machine which can compete? Well we're back to the interaction with environment problem. Is intelligence a gestalt of entity and environment? Aha! Got you there eh?

    Now back to quatumology stuff. Why? Because if quantum mechanics is involved in the operation of the brain I think it would be helpful to know about. First off, I don't believe in photons. There seems to be a circular argument going on. Electromagnetism is carried by photons yet they are packets of electromagnetic waves. I may be ignorant, and I hope I will be corrected by BillyT or others, but I don't know that it is possible to detect a photon without destroying it. So from my point of view all you can say is that there is a loss of energy from one point in spacetime and a gain in another. Not very helpful perhaps but highlights the problem with modelling nature.

    So anyhoo...there are pro and con arguments about quantum effects in the brain. A review of these can be found here -
    It will be noted that the thermal swamping argument is not so clear cut. This is not to say that the quantum effects play any functional part in neuronal activity, just that it is a possibility. You'll also notice that the Penrose effect (microtubules) isn't demonstrable either.

    Next, the possibility of consciousness affecting quantum wave collapse. Again, there are differing views. Either there is a wave collapse or there is a multiverse of possibilities. Take your pick. Neither are satisfactory.

    Finally, cosmology. I don't have the physics or maths to deal with this so...
    Quantum Cosmology and the Hard Problem of the Conscious Brain
  15. ElectricFetus Sanity going, going, gone Valued Senior Member

    You can stop there, because I don't believe that: brute force processing could (and does make) conscious machines, but it needs to be programed, wired, even designed specifically for that task. A gigantic supercomputer is nothing but a huge overprice paperweight without programing. Likewise an infant may have 100 billion neurons in it little skull, but it can't yet think past: suckle, scream and shit.

    No I don't believe that either.

    No it not, as IBM's Watson can show you. I guess the question would be does it need to be conscious to have GENERAL intelligence: to solve any problem presented to it as well as or better then a human? I don't know, I guess we will find out eventually, it may very well not think beyond "task received, solving task, task solved, awaiting further tasks" which few would consider consciousness, and frankly I would be very pleased if that was as far as it's 'consciousness' got. I would think it would be possible to give it complex conflicting goals, emotions and such, but I think that an extremely bad idea, akin to discovering skynet can't simply come into existence by accident, so you intentionally make it!

    An artificial general intelligence that that can solve any problem as well as or better then a human is good enough for me, it does not matter if it is conscious or not. All are labor problems would end with mass production of those machines, we would become functionally obsolete except for the one task they would be specifically designed not to be able to do: assigned new tasks.

    it isn't.

    What your asking for is that somehow there is something beyond our detection that will bless us with un-replicable souls. All I can say to that is time will tell.

    I've gone over this, there does not need to be conscious observer to affect quantum wave collapse, so your choices are false.

    Cute set of hypothesises, but unproven, and frankly a clear attempt to try to make scientific theory and even conjecture fit a set of philosophical beliefs, rather then facts.
    Last edited: Feb 13, 2014
  16. Billy T Use Sugar Cane Alcohol car Fuel Valued Senior Member

    Not as much as many; but yes it is sort of (at least conceptually) possible to detect a photon with only change (energy reduction) of it, not its destruction caused by it being absorbed. Done via measurement of the Compton scattering atom's recoiling energy and momentum change. You can learn a lot about the photon, as energy and momentum are conserved but actually doing the measure is very tough - Perhaps possible by using "time of flight" to get its speed and which detector pixel the atom subsequently hits to learn the momentum vector direction.

    I have Compton scattered X-rays off a small carbon block but measured the photons before and after to infer the carbon atom recoil, not the other way round. Actually theory predicts the angular distribution of the scattered photon when the incident beam is well collimated - testing that prediction was standard lab experiment I did in a very special 5-year program at Cornell - Too tough so was discontinued in a few years, but IMO was the best physic undergraduate program ever offered in the USA. (Half my entering class transferred out to easier Cornell departments like Electrical or Chemical engineering. The now reduced to 4-year residue still claims first or second place in US News's raking of under graduate physic/engineering courses every year.) I had a "full needs" scholarship, washed dishes for my meals, and never paid a dime for my education thru the Ph.D. later at JHU but worked like "two dogs" to keep the scholarship at Cornell (GPA of 80 required) as I was very badly prepared by my West Virginia public schools education.

    You can think of this as sort of reflecting the photon by a mirror which has impossible to actually make sensitive pressure gage attached to it to measure the mirror's recoil. You probably know that "solar sailing" has been demonstrated. That of course get a observable acceleration of the space craft by reflecting "zillions" of photon.
  17. ElectricFetus Sanity going, going, gone Valued Senior Member

    Would not a very light mirror and a very powerful laser count, pressure on the mirror could be detected without destroying the photons, just sucking energy off them. A solar sail works via this principle... Oh god dam it now you have me doing it, none of this has to do with consciousness!
  18. Waiter_2001 Registered Senior Member

    The only way to know would be to create a conscious machine.

    Please Register or Log in to view the hidden image!

  19. Nataliee Registered Member

    how can you create a consciousness? I just can't even begin to comprehend the concept of it, I don't even know how I got my own consciousness... but of course that's just me.
  20. wellwisher Banned Banned

    One preliminary task, that is needed to set the stage for creating conscious computers, would begin with a new approach to memory design. Computer memory is based on stability, so memory does not change, allowing long term storage. Living memory of the human brain, is more based on plasticity and instability, where change is part of the game. For example, if we look back and try to remember an event, we lose data with time, and may even remember the event in a biased way. Time heals all wounds because memory is not static but changes with time. There is a logic for this.

    When neurons are in their rest state, they are at highest energy potential. The cell expends ATP energy, pumping and exchanging cations creating a concentration gradient of cations across the membrane, which contains potential energy in terms of energy and entropy. In more specific terms, we get a millivolt level potential, as well as segregated cations that prefer to mix, to increase the entropy.

    The active state (as defined by science) of the neuron; action potential, is when this highest energy state is discharged by neuron firing, and neuron energy heads toward lower energy by moving cations around. The beauty of this design, with respect to consciousness, is the movement from higher to lower energy, also involves lower to higher entropy, meaning new paths will appear; memory changes with time.

    Picture a computer using this memory design, where memory is potentiated and subject to changing. This is not good for repeating anything, but it is useful for creating spontaneous variations, such as MP3 songs changing notes and words. Normal programming logic will be now be using memory that is subject to spontaneous change. The results might come out unique each time. This is only the beginning; get memory hardware working. The next step requires software design based on the 3-D programming model of the personality firmware. Consciousness is 3-D not 2-D, which is why logic can't fully define it, yet we can sense it with intuition; third variable of 3-D.

    This personality firmware is based on both hardware and software to create 3-D.

    Picture this; The 3-D firmware sets an energy grid, with the discharge of the variable memory (moving to lower energy and higher entropy) being led down logic built into hardware, that can make the change in the memory closer to the needed result. Picture a fountain pumping water upward. Once the water reaches its highest height it temporarily stops (rest neuron). After this short rest, it then begins to fall downward (action potential). The cascade looks consistent at some level, but there is also variation because it is dynamic. To make variations in the cascade more specific, we can add hardware to the fountain, such as obstructions and diverters so the path is more defined in space. There is still variation but narrowed down. If we really funnel the cascade, it may flow in a tube to the bottom, so there is little variation.

    Step one is variable memory that is subject to change to see how this behaves and how we can set parameters.
  21. Waiter_2001 Registered Senior Member

    By dividing zero by zero: the current algorithm for computing division is insufficient.
  22. kmguru Staff Member

    I was told...you need Spirit Universe to develop consciousness...hence in 8 Billion years...it is impossible...I could be wrong...
  23. cluelusshusbund + Public Dilemma + Valued Senior Member

    If you thank machines wont be considered conscious (as in curent human conscuousness) someday... you surely thank consciousness is somptin other than biological... as in magical/a soul... an if thats the case... you'r view of machine consciousness will be based on unfounded beleifs you make up as you go... or on unfounded beliefs others have made up... insted of reality.!!!

Share This Page