Nobel Laureate - Sir Roger Penrose OM, FRS - Plenary 11 TSC2020

the title of my talk as you i hope can see is that well orco are and it's really about new developments on or or so i should explain that things have happened during the lockdown period because normally during uh most of my time i would be gathering about the world ruining my uh carbon footprint and so on or at least enhancing it i suppose which is the wrong way around and uh unable to do this i was able to think about things which i've been sort of leaving aside for a long time and i thought it was worth thinking about more about the or our part that is to say physics and uh i want to talk about some of the things that i've come upon since the beginning of lockdown which are basically new um not all to do with the gravitational uh reduction of the state i should explain what o.r is and i hope you can read the bottom of this picture is that can that be read what i want to say is it stands for the objective reduction of the quantum state and of course the acronym or says one or the other rather than the superposition of the two because quantum mechanics just leads to these more and more messed up superpositions and you don't resolve out one thing or another people often say that quantum mechanics is the most wonderful theory we've ever had in physics the trouble with it is that it's self-inconsistent theory and i don't think the most wonderful theory ought to be inconsistent with itself it's self-inconsistent because well let's go to the uh well before going to that yeah no let's say about a little bit more about or see it's self-inconsistent because it consists of two features the most uh detailed one if you like is the schrodinger equation this is an equation which tells you how the state of the universe evolves and the that schrodinger equation does not tell you what happens in the world and i let me say a bit more about that but before coming to that i thought i would mention a paper which came out just last week it came out on monday in uh nature physics and it's i don't know i'll put it up here i don't know whether the uh this is i hope that's more or less in the middle of the screen there the paper is about an experiment which has been just done recently about putting a crystal that's a gun with its germanium or something and they try to see whether you get radiation from it now i should explain that there are many theories of oh they don't necessarily call it o-r the idea is that well some people don't worry about the reduction of the state at all they just go in and do their quantum and well i'm talking about people who work in quantum theory they may uh not worry about it they just use the rules and forget about the fact they're self-inconsistent the schrodinger equation does not have a collapse in it the collapse is something which is to be introduced outside there are various proposals for sort of classical ones that somehow when the environment get mixed up with the state then you might as well say that collapse has happened i should say these arguments are not correct they're very very prominent and people somehow believe okay when you don't preserve your quantum state very precisely and the environment gets mixed up with it then somehow uh one thing happens or the other well that's not true it's just the environment gets entangled with the one thing and with the other and it's just such a mess that uh there is a procedure which people adopt often which is a bit of magic i mean you wave your hands around it's not a consistent procedure it doesn't really explain why one thing happens with the other there are other models like many worlds as we had earlier mentioned today which means that all these things do happen all at once that doesn't really explain why we don't experience more than one thing so it means in the sense that once consciousness maybe follows one of these but then why does why does somebody else's consciousness follow the same thing or do i lose all my friends because they've all gone different down different tracks i've never i should say it's probably a stage in life in physics which you go through there was a time when i went through this stage i don't remember how many weeks it lasted but i certainly did contemplate the idea of a many worlds picture before coming to the conclusion that really we need something much more drastic than that it doesn't explain anything um okay now among those people who think that something should be done about it there are very well broad categories of theory um and most of these involve either heating see if you have an object let me just explain it i this is an article i i didn't want reading to go into and i hope you don't see it very clearly because well i'm you won't be able to see the small print but it's basically disproving what they call the the diyoshi penrose model and uh they come to a conclusion that the diyoshi penrose model or you see here's the second page of this article and at the top there's a sort of cartoon of what happens you have a an atom in one place superposed with an atom in another place and the sum of those is what happens they're both at once and the idea is that if you have any theory which collapses the state it becomes one or the other then you sort of jump from this state to that and this jump is a sort of cause it causes heat because you have some motion it and if you have a jump in your system then it's likely that you would expect well things get jiggled around and so it provides heat or in this case what they're saying is it produces radiation and so they won't consider putting these crystals down in mind to see whether they actually radiate down a mine because there's all sorts of other radiation going on and they want to shield all that out just to make sure that they see this particular radiation they don't see anything and so they conclude this is a disproof of this model it probably is a disprove of this model but it's not my model even though it has my name attached to it i should explain why it has my name attached to it because the criterion that this hungarian physicist yoshi put forward several years before i did which is a proposal for the rate at which the reduction takes place it has to do with if you have a superposition between two states here we have a something in one place and something in the other and you look at the difference between this mass distribution and that one and you look at the gravitational self energy of that and the reciprocal of that is the rate of decay and that formula is the same as one i used i came up about it from quite a different direction but that is the same formula so that's why they link our names but in my model and i'm going to try and explain it because it's it's it's a bit outrageous what i want to say but then quantum mechanics is outrageous and the reduction of the quantum state is pretty outrageous so you need something i think fairly outrageous to explain what's going on and it's more outrageous than i expected at first but the a lot of things i want to say are not specific to my model they apply to other models like this well this one yeah i'll say that it does apply to it in a sense let me before going to that say just something about quantum mechanics as it's normally used and here we have schrodinger's equation written at the top the main point about it which i want to make is it's a linear equation you just have a derivative and you've got size on both sides psi is the quantum state and it's because well you have a the hamiltonian which is a linear operator everything is linear what does linear mean well you see if you have one solution of the equation you call it psi usually this letter is used for the wave function if you have one solution and if you have another solution because phi then any linear combination alpha psi plus beta phi is also a solution that's the whole point of linearity alpha is a constant beta is a constant these are complex numbers and they are related to when you make a measurement if you make a measurement to ask the question is it a or is it b or is it alpha is it psi or is it phi should say then the squared modulus of these numbers give you the relative probability so that's the bourne rule now the problem with linearity as schrodinger was very clear to point out people often point to the schrodinger's cat and say well if you had a really elaborate experiment you could put a cat in the state was alive and dead at the same time and isn't that wonderful schrodinger was really pointing out the absurdity of his own equation to have a state where you have a cat which is alive and dead at the same time he regarded as an absurdity as i do nevertheless uh you it's it's an irrational conclusion from this linearity here i have a laser and here we have a beam splitter that's half silver mirror if you like it it emits single photon this single photon goes along it splits into two so the photons location is partly going along here and partly going here the the state of the photon is the superposition of these two different routes this route goes and it fires a gun i should say that there's a cat at the other side of the room here if it goes this way the cat is unharmed if it goes this way it fires a gun kills the poor cat and the state if the photon is in the state of the linear superposition of these two which it is as the photon goes along its state becomes a linear superposition of here and here so the reality of the world if you like is this together with that on top of each other this one fires a gun and therefore if you if the linearity continues right up to this level then you must have a cat in a state which is alive and dead at the same time so that is a completely correct conclusion of the formalism and schrodinger was basically pointing out that his performance he's really pointing out a flaw if you like in his own theory he's saying roughly speaking my theory can't be right because it predicts such absurdities people often don't read it that way let me um i could move things sideways i'm not sure whether if i put both together that can be seen um let me do that if you like don't you can you see both of them at once yes we can okay now i'm going to raise this question is is the state vector is the wave function is it real now you see people often say well it's not really real it's something which is it tells you probabilities it's really telling you something about the probability of one thing happening or of another thing happening it's not really probabilities because these numbers alpha and beta are not probabilities they're amplitudes they're more subtle quantities you have to take the squared modulus which is a number to get the probabilities the ratios of these squared money but there's more information in it these these amplitudes are very subtle quantities and they're not probabilities they're certainly not probabilities there's something from which you can get your probabilities if you have a measurement which uh distinguishes one from the other the the size state from the five state now is it real now my eyesight is not good enough to read this down here but this is a quote from a famous paper by einstein and colleagues of what he calls an element of reality what he says more or less is that something has an element of reality if you can predict with certainty result of an experiment which does not disturb the system so if you have an experiment which you perform on that system and it doesn't disturb the system and it provides the answer yes that you expect with certainty then there is an element of reality to be attached to that thing and i think it's a very good criteria it has very strange consequences one of the consequences however is that the wave function is real there's a little confusion of terminology here because real is not in the mathematical sense the state is a complex thing not not real in the real number sense but real in the sense of reality well in a sense yes because there's a theory always a theoretical experiment you could perform whether you can actually perform it there's a theoretical experiment which could tell you that the state it gets you answer yes if it is in this state with certainty and that's the only state which will give the answer with certainty yes i won't go onto that because my arguments don't really depend upon it but i say it does have a kind of reality so i'm going to say that the quantum state is real now i'm going to say something more about this namely this is what one of the things that came about from my worrying about well in the lockdown trying to worry about the reduction of the state there really are two kinds of reality and this didn't occur to me for quite a long time the reality here is what we're going to call quantum reality and i'm calling this einstein's dictum if you can well let me say the two kinds of reality in classical reality you can ask the state of the system what is your state and it can say my state is x you can go and explore it and measure it and all sorts of things and yes you can find out what it is you can find out what its state is you don't tell it you don't put to it what it might be you say what is your state and it says my state is x that's classical reality many experiments are of that nature quantum reality is the einstein dictum thing you can't ask what its state is but what you can do is you can make a little calculation you think oh i think the state by now ought to be x and then you say can you confirm whether your state is x and it says yes with certainty you can repeat the same experiment many many times every time it will come up with the answer yes so einstein says this is a real thing but the point i'm trying to make it's a different kind of reality and i think this is an important point so i want now to say a little bit about the classical reality here let me consider a situation like let me not take that away for the moment i'm going to consider many of my diagrams will be space time diagrams this is a space-time diagram time is going up a picture now i'm considering a lump of a lump here and you're hitting it with a beam split photon so it's not the cat which could be there alive um i can find that again but what it does is if the photon goes one way it gives the object a little push if it goes the is that the cat no i'm looking the one i'm afraid i have to apologize my eye size is very bad and i can't see these things here we go here we have a laser and instead it goes beam split instead of firing a gun it pushes a lump so if it goes this way it pushes the lump if it goes that way it doesn't so we have a lump at the superposition of two locations i'll say a bit more about this later but here we have a history as time progresses think is it going up here the thing is put into a superposition of these two different locations now after a while it reaches this criterion to say well it's likely that it's going to reduce to one or the other and let's say it does the dotted line means it's the superposition of these two different histories as time progresses and then it becomes this one not that one so this one disappears so that is the sort of history of this evolution now the problem i'm raising here is that i'm not a relativistic picture that is to say you're trying to consider that simultaneously the lump becomes one over here and not over here however let me consider an observer moving at great speed to the left in the picture and the line here represents simultaneous events and as that time progresses that line is tilted with regard to the other line and so if this one becomes the lump location here and this one disappears when you get to this point the lamp is here there and partly there there and this is a terrible thing because if it was partly there it means that if you perform a measurement there's a 50 chance you might find it here in which case you'd find it in both places so that's a nonsense what about the other you can consider now the green person's measurement is the other way around now the simultaneous lines are this way and then here's where you reach the problem this one is still maybe there maybe not there this one is gone so there's a chart 50 percent chance that it might disappear from here and 50 chances might disappear that's absolute nonsense that's clearly wrong so it can't be the picture cannot be correct that somehow it disappears from one location not from the other now you might have a model in which it sort of phase out gradually it doesn't help there's even a reason why models where where the where the superposition sort of phase out and it becomes more and more one and less and less the other and there i have a strong argument why that can't work um i'm not going to go into that argument uh i think i'll give a hint as to what it is and if anybody wants to know i can talk about it but what i really want to say is that a picture in which it disappears in one place after being in a superposition then it's totally in the other place doesn't make relativistic sense in fact the only sense seems to me you can make of this picture which is consistent with relativity is to trace everything right back to here so if it disappears here it's as though the reality was this all the time now that's an absurd idea you might think it's uh what i call the retroactive perspective and i have to hear another picture of that i'll move this over to here and what i'm saying is that we think now the space times here i've just had these paths in spacetime if you like but you see each of these lumps deforms the space-time so as we go along here we have at the bottom of the picture i hope you can see that uh a laser here a beam splitter there the photon goes through partly here and partly is displaced out that way so the proton is in the superposition of coming this way and that way so when it hits the lump if it hits the lump it moves it it displaces it backwards a little bit but if it doesn't hit it it doesn't displace it so the lump is in the superposition of being well as as time evolves it moves a little bit so here it is a superposition here and here superposition here and here but what we have to bear in mind is that the space time now becomes a superposition in the sense of two slightly different space times now the criterion that deoshia and i formulated i can phrase it a different way and that is that the time up to the collapse which is here this is where it becomes one or the other the the the difference between the space time so that's a four dimensional distance is one unit in planck units now planck unit is the unit that you get when you try to combine gravity with quantum mechanics and you get this absolute unit and it usually puzzles people because they say if you get to that small distance then space doesn't make sense or something but this is something rather different they say that your space time it can still be a continuous picture but when you get a superposition which in space-time terms becomes one unit then one of them has to drop off now it's an average time it doesn't happen exactly then you have to think of this like an unstable particle which has a half-life and this is basically the half-life after which it becomes one or the other incidentally it's slightly well i on the right hand side i have the quantum reality now the quantum reality preserves both superpositions so that persists as a supervisor position but the classical reality says it's got to be one all the time so when this thing makes its choice so the choice is made here in a sense it affects the classical space time back here now this is what is a sort of outrageous thing to say it looks as though your in the future influencing something in the past but in a certain sense i think this is what happens you have to be careful of all sorts of paradoxes and i'm worried about this of course i worry about paradoxes uh but i think it's safe from paradox it's it's it's a i should say i did actually talk about this i think it was two years ago at a tucson conference when right at the end of the conference i mentioned this sort of idea i hadn't really thought it through very much but i did talk about it before that was the first time that i ever put this particular idea forward but uh i've sort of filled it out quite a bit but uh into and i think we have to think of the quantum reality in the quad classical there's a bit of an irony here because if you use einstein's dictum it doesn't quite give you this because it would say does it say with certainty it's this superposition or not well you see there's a little chance that it might reduce the state so it's almost certainty if your way down here say it's almost certain that you're in this superposition but not quite because it might reduce so it gives a little bit of fuzziness to the quantum uncertainty which is fine but i thought i would mention that point now let me give you another picture this one i did get before it is in color so you might like to see a colored picture of the same thing so i'll move that over here and put my colored picture in here we have the oh yeah i've dropped something i think i've got it yes here we have in color the same thing which is going on here but it was space time you see the lump becomes in to a superposition of two locations and then after a little while that's too much of a strain on the system one of them has to die and the other one lives and this is where the reduction happens and now i have a cartoon of um i think it's this way up here we have where the or takes place so we have objective reduction this is the moment of proto-consciousness i like to call it proto-consciousness not consciousness because well if everything is conscious you can't even walk down the street without killing something and i think uh proto-consciousness is a safer thing it's the proto it's the building block out of which conscious actual consciousness takes place it has to be orchestrated in some way which is a much more subtle problem i'm sure to give something which is we could call genuine um something that can really feel something in a genuine sense but in a certain sense it's the or our choice which maybe gives us so scope for free will i've never been quite sure what i thought about free will um and certainly the view is that if you have a deterministic universe like newtonian picture when you introduce maxwell's equations it does it's still deterministic if you introduce um maxwell's demons even it's still still deterministic if you introduce the quantum wave function it's just a schrodinger equation it's still deterministic it's not deterministic or something when you have this reduction of the state and maybe that's where one's consciousness actually does come in and makes a decision which goes back in time in some sense now this relates to in fact this was the sort of thing which was worrying about before in the tucson conference then was experiments due to benjamin limit in the um in the late 1900s and uh i'll give you a picture of one set of these experiments which is that i mean two classes of these experiments these which involve where you make a movement so there's the active effect of consciousness i will my hand to move this way and there's the passive aspect of consciousness when you actually sense something and there are these experiments which seem to indicate that moving something or other there's a precursor of that and and people worried about this i was never quite sure about that uh because i'm never quite clear whether when you actually decide something the precursor may be gearing yourself up to decide something i don't know but the other one is to do with the sensation and this is also a limit experiment and i've got a cartoon of this here well not this is actually another cartoon it's a picture i don't want to talk about all these things but i think let me explain the situation is there is a stimulus to the hand and this is i think an electrode which is touches the hand at a certain point and this is the same time as this patient is having an operation on the brain for some other reason and the part of that brain which is concerned with the feeling of that p part of the hand being stimulated is being stimulated with an electrode and it's a very very interesting phenomenon i think i'll describe this one particular one here which is where you first of all stimulate the skin and then a little bit later you start to stimulate the brain now if you don't stimulate the brain this patient thinks that the stimulation of the skin is pretty well similar simultaneously i mean it takes place at a certain moment and the patient feels it only slightly after that so it's more or less simultaneously whereas if you stimulate that part of the brain up to about half a second later then the patient doesn't feel the stimulation of the hand it fuels the stimulation of the brain but it can distinguish the patient can distinguish the difference between those two things it feels like the skin but not quite the same you say no no it's not what actually stimulating the skin feels like i can tell the difference but it's the same part of the skin that's involved and if that stimulation takes place after the stimulation of the skin within i think half a second or something then you don't feel it the actual simulation of the skin at all which is very remarkable now it seems to me that this may tie in with some of the ideas i don't think stewart mentioned particularly but he did mention the pyramidal cells and this may well be i think he may have said where the consciousness takes place now it takes a while before the the touching of the skin makes its way up to this part of the brain and then you've got to if that's when you become conscious of it you have to refer it backwards to when you think the stimulation actually took place now you see if you can also create an action like as in this picture here which moves your hand or whatever it does you feel you feel it up here but then you start to do something a little bit earlier you see i've often worried about this delay that people say well you don't feel something until half a second later and this theme always seemed to me a little bit strange because if you play ping pong i used to play ping pong quite a bit and this is something you i mean half you know it's got to be much much less than half a second to do something you've got to make a decision to do with the ball and is this entirely unconscious or or actually are you doing something consciously it always seemed to me that you're doing something consciously maybe not these it may be pretty well pre-prepared the kind of thing you might do but the choice may come later so it struck me as these things are things which do need explanation and i don't think um we can disregard experiments of this nature how much time do i have left well we're technically a little bit over do you think you can bring things to a nice tidy conclusion all right i'll bring it to a tidy conclusion which is what about an experiment which could test the retroactive perspective here he is this is i had experiments in my notes which were like this but i stretched it out so it's a stretched experiment i hope you can see this here we have a beam splitter and a laser here the laser is there hits the beam splitter the photon goes either this way or that way if it goes this way it hits a lump starts it moving if it goes this way it is a mirror and there's a mirror on the lamp which reflects it back and if they're coherent it goes straight back here and detector over here won't feel anything now you see i've did it here so the beams come back after the lamp has started moving but it hasn't moved far enough for the state reductions to take place it moves and moves and then finally it does one or the other now when it does one or the other any other model i think of state reduction would say well it's too late the photon has already got back here it's reassembled itself and there's no way this detector would be activated but in the retroactive picture it would go back to this this is where the moment of reduction takes place and it would have to go back to there and therefore you would have 50 percent chance of this detector actually receiving the photo it might be a pretty difficult experiment to do it's not too far beyond the kind of experiment the dirt bar mister has been doing to test it's an experiment this one like this here at the bottom where you have um cavities and so on and i won't go into that and if anybody asks me um but it could be done it's certainly not on the cards yet yeah i think it should be experiments should be geared to seeing whether this is possible or not thank you very much thanks so much roger well done so over if i did yes so let me take a few minutes to to think about this because i i'd like the audience and just all of us to to consider and really get our minds around how radical a picture you're proposing and in the same uh sentence also a picture very much in the mainstream of physics this is an idea we have in the least action principle for example that we have to consider the entire trajectory from beginning to end of a of a particle so in a sense there's a teleological aspect to physics already and i'm wondering if you could just comment almost philosophically or just reflectively on the significance of your retroactive interpretation well it's radically different from all these models i mean go back to newton and uh la pass and people who uh who worried about uh you know where's free will even if you don't know what makes you make your choice or whatever it is um well if it's all predetermined and somebody i mean i think was laplace who made this particular point very strongly that some somebody could in principle or calculation could exist which would calculate what everything is going to do once you know the initial state i've lived with not really making up my mind about this for a long time and uh the schrodinger equation is completely deterministic the only place that there could be a difference there is in the reduction of the state now as i think it was mentioned earlier there is this point of view that maybe it is the conscious being that reduces the state now i find this very difficult to swallow particularly because you could think of an experiment like following there is a space probe going out to look at a very distant planet several light years away it's an earth-like planet and so it's got an earth-like atmosphere on it and this probe goes out there and may take several years to go out there and it sends a signal which takes quite a long time to get back to an observer looking at the signal now it's pro is having it takes a photograph of the atmosphere now the atmosphere is as we know a chaotic system and people talk about the butterfly effect and things like this well even slight quantum effects are going to change the atmosphere so you're not going to see you know if there is no state reduction you're not going to see a particular atmosphere or you're going to see a miserable superposition of all possible atmospheres it's just going to be one blur this signal comes back and and an observer is sitting in front of the screen and finally this signal comes back to him or her of what this space probe is signal and just at that moment the consciousness comes in blip it becomes one atmosphere doesn't go all the way back to the signal i mean it doesn't make any sense to me at all so i just don't see how people i mean i can see why they did because quantum character works so well and people talk about observations and von neumann was no fuel no fool it was many people call him the greatest mathematician of the 20th century i don't think i would say that i think herman var was a greater but never mind um he was a very good mathematician roger how does the retroactive view that you've now put forward and developed in the past few months help with the problems that you had identified earlier for instance the heating problem yes where you see there is no heating because it comes the heating comes about because you have these gas molecules starting to move away from each other so the superposition is now they start it's a superposition of the molecules being here and being here now when that happens the state changes from this to this that's a big change now that will produce a heat or radiation as in this experiment but in my scheme what it does is it traces back to where there was very very little um change in one or the other so that the two the molecules were in the same place i mean you could you can have a you can have a well there's no heating because it goes back to where they i mean they just did i mean that there is there is no jump the classical space time evolves continuously smoothly and you see this is the other problem but in the all these other models where you have a jump there you violate well you violate the conservation of energy it may be a very very tiny violation but that's bad news it's bad news for general relativity i suspect that it might be bad news even for celestial dynamics i i think the you can track the motions of planets so accurately now i mean you know the motion of the planet venus was one of the great uh the very first great achievement by the bending of them i don't know which way you say because they the perihelion and mercury was known before but they um i don't remember well calculations when was that noun i don't know about the history of it very well but anyway that there were two the perihelion mercury in the bending of light but the perihelion of mercury uh i mean the planetary emotions were known extraordinarily precise before that so that you would be able to distinguish einsteins from newton's theory now i should have thought that if there was any change of mass of any significance in a planet which would be really quite large if this state reduction has taken place all the time that this would have been shown up by now now i'm just guessing whether that's true so roger the if i'm following you uh this new view the retroactive view maintains the principle of conservation of energy but on pain of just introducing a key theological element to to our physics that's correct yes but it's a it's a very slippery kind of teleology which is hard to pin down you see i i have worked hard to make the experiment i mentioned at the end i had to work hard to think of any experiment where a model of the standard kind of i mean you have an experiment the first one i had which i i uh it was developed from was one here which i considered it but you see you consider that the photon might be entangled with the object here so you lose your coherence because it's entangled but this is so long afterwards and the thing has already cohered so it took me a while to think of an experiment where you could conceivably see the difference well i say the difference between well actually see the retroactivity directly i mean that's what i'm trying to say whereas this particular experiment that i'm putting forward here it's only in a very uh primitive form i think i mean maybe one of some of these experiments using bose einstein condensates that you get fuentes is collaborating with me on and i think there's a lot of a lot more scope there because you could there a lot of flexibility but how you do this experiment with them i i don't know yet the balmiest one is a little closer to that you could modify it by putting some more cavities in you put a couple more cavities here's one you have one cavity cavity here and one cover here and you put another one all right so i'm looking forward to seeing how this idea develops and the paper you've already shown me a draft of a paper that's something like 80 pages long so i'm curious i'm gonna have to use that yes so let me try to rob the discussion a little bit um by the way audience members the organizers have allowed us to go a little bit long on this session uh just to make sure we get some of your questions in but using my prerogative as a chair for the for a moment i wanted to ask stuart if you could just bring us quickly up to speed on the status of empirically testing and intercomparing the different theories orco r iit i know you've got a whole initiative to try to bring some kind of sense to all this this zoo of theories and where does that stand now well templeton has the project accelerating research and consciousness templeton world charity foundation and uh we initially attempted a uh uh adversarial collaboration with iit but we couldn't come to an agreement and agreed to fund our uh attempts to falsify oracle r without uh bringing iit uh into the mix i personally don't see how iit is testable or falsifiable i know that they are they're in a project with a global neuronal workspace something about if the mri is active in the back of the brain iit wins if it's in the front of the brain global workspace wins but i've seen studies uh where the activity supposedly correlating with consciousness is sometimes in the front like with the executive actions or if you're just watching a film mindlessly it's in the back of the brain so i think consciousness could be enough first of all we don't know that mri uh correlates with consciousness it's blood flow and there are examples where we know it they they don't correlate uh so i don't see how uh activity bold activity in the front of the back will prove or disprove iit i don't think it's it's testable or falsifiable i actually don't think it has any spec it's very general it applies some uh non-linear i'm not even sure what it is when i heard first heard about phi i thought ah the golden mean that's cool fibonacci microtubules interesting it's not that it's something else and i still don't know what it is after all these years i also have to say that i spent a lot of years with a non-linear dynamicist very good friend of mine alwyn scott who wrote stairway to the mind and started the cnls at los alamos and went to a lot of chaos meetings up there and whatnot and tried to apply non-linear dynamics to consciousness and microtubule activities and it just didn't seem to do anything for it and then i read roger spoke and met roger and decided that the quantum approach was much better so um iit's main main claim to fame seems to be that they explain that the way the cerebellum is wired up it doesn't have enough fire which is why the cerebellum isn't conscious but the cerebellum doesn't have pyramidal cells i think consciousness occurs in pyramidal cells primarily it can occur in any microtubule bearing uh material or or anything else really if it's proto-conscious um but uh uh you don't need to invoke the you know how the cerebellums wear it up if if you need uh pyramidal cells so uh for a while uh they were trying to say that uh if we could show uh quantum vibrations in microtubules uh that uh that uh iit would a phi would apply to that and i said okay well let's do it and then we'll see which uh which applies better uh orco r or iit and then they backed out they didn't i called their bluff and they didn't go for it so uh we're going on our own we have funding to do uh and i'm not involved in the experiments roger's not involved it's being done by experts who are basically neutral uh in fact uh some of them are are skeptical uh of of what we say so we'll see and we're you know i as i said in the title of my talk uh it would be very easy to falsify work or at least or by showing that there's no none of these quantum vibrations in the microtubules as our mod as our computer modeling study showed basically we're trying to do that computer modeling sh that i had a blitz through too fast uh do it experimentally to find these quantum vibrations and microtubules in the terahertz and if they're there at room temperature or biological temperature then see if they go away with anesthesia proportional to the anesthetic potency i think that would be a very strong claim certainly stronger than any other theory of consciousness yet proposed but i have no idea how to disprove iit right so looking at some of the questions that are coming in there's actually a common theme has developed a few of the questions between your talk and peter and hartman's talk and maybe actually pose this question to peter and hartmut therefore and that is the the role that you're giving to pleasure seeking as opposed to other uh evolutionary adaptive advantages for example or subjectivity so can you flesh out a little bit why you give pleasure this central role peter uh hard better me um maybe maybe hartman and then peter can chime in and then stew you two yeah so we definitely didn't want to propose a pure pleasure seeking robot um it's actually both ways um what pleasure and displeasure and all we meant to say it is a good place to start um when you try to reconcile the first person perspective and third person perspective as i said it's a good place to start with these perspectives are correlated and it seems like they follow antonio de marzio's who points us out very eloquently that feelings are basically about checking in how the homeostasis is doing and if your homeostasis is doing well then you seem to be in a good state it feels good and if your homeostasis is threatened or becoming unstable then it feels unpleasant so it's this correlation it's not just about pleasure it's about both signs peter did you want to chime in as well yeah yeah i mean uh inevitably i suppose as an investor i look at this with a slightly more commercial angle so for me um you know trying to sort of tease that out and i think what's interesting for me is the where we got to right at the end which is what happens when you integrate this homeostatic engine that works in the way that harmful just describes with a learner and because i think that is where i see commercial application for this it as as we said it has um it has ethical questions uh that would need to be navigated very carefully but i think it's a really interesting uh you know point to leave the audience with which is would you you know would you rather have the best ai and we think that one integrated with a homeostatic engine is that or would you want to have one that you can control and i think that's a really kind of a nuanced ethical question but it's also um you know highly commercially relevant to a lot of the work that's been done in ai right now so stuart do you see this tying into your own that big paper you did a few years ago on pleasure seeking as an essential evolutionary driver yeah i've never been uh totally happy with the idea of evolution oh stewart can you turn up your uh volume turn it what oh there i can hear you i've never been totally happy with the idea of uh evolution and uh gene survival that that everything we do is to promote our genes if you look any organism in a lab uh on up to us humans uh act said to optimize pleasure in some way certainly reward in in a in in rats or any experiment you want to do and then we learn to you know delay our gratification and it's not all hedonism or pleasure but basically we're here now to because we're pursuing our profession and we enjoy it and in some way we're getting we're getting some pleasure out of it or some gratification so all we do you know it doesn't need to be hedonistic it could be altruistic it could be spiritual it could be uh you know it's better to give than to receive so it's not a hedonistic approach but it's certainly i think it certainly started out that way and when i realized that roger's objective reduction with these organic molecules uh could have could have uh did should have preceded the origin of life you know and the origin of life is still somewhat of a mystery still a mystery and uh if it was uh you know simple pleasure or avoidance of displeasure could have been the spark sparks uh that originated life and just kept on going as the motivation for the uh development and assembly of organisms into more and more complex uh entities to optimize pleasure in various ways so um i like the uh the quantum pleasure principle idea and it's uh you know i i'm kind of kind of uh uh become used to being uh an outlier in in science uh in many ways so you know let's just add evolution to the mix but uh and i'm not sure how roger feels about that so i he certainly i mean i'm using his ideas of o.r in the in the uh you know early or mid-universe uh uh to to promote uh life and evolution but it's logical to me it makes a lot more sense than uh gene survival and you know genes don't feel or think as far as we know if they do then then maybe that's another case but uh uh i think it's all about optimizing uh pleasure and avoiding displeasure wonderful thanks so ani are you are you there you dropped off my screen here but we do have a number of audience questions for you i hope he hasn't gone to bed i know it's in the middle of the night for him sleep me so ani if you come back in i will pose these questions people are dying to know more about quantum cloaking um going back to um to peter and and hartmut a question came in about well essentially asking how this works in collective systems how for instance what whether markets could actually exhibit some of these feelings or feedback loops that you're describing homeostatic feedback groups you're describing is is this applied to biological organisms and ai systems or is it a more general feature of the world the kind of approach you're taking peter go yeah maybe um you know i i gave sort of this um explanation of why we think a conscious experience is a limited observable it's observable strictly in the duncan experiment of two clones and i have always thought okay between us and the robots as the um on this golf is too big we are too different even though we can think of this cheesy move of giving it self-reporting abilities via word embeddings but i always thought if there would be a society of um or multiple animats that could interact with each other then for them they become observable their conscious state becomes observable relative to each other and can we exploit this somehow um but i have to admit i never got this to work as of now so yeah it would be very interesting to think of this animat we proposed in a group setting so you have multiple ais interacting with each other what happens then um but as of today i don't have a good answer for that i'll just add that i think the the principle that we were trying to embody there was um the well-known um it takes one to no one uh which we thought you know was an interesting place to start and and we actually sort of felt that that was much easier to model between i mean the model we had in our minds was a you know a parent and child uh you know very often the expression on a child's face will tell a parent quite a lot about what that child the conscious experience of that child is having um so i think we were we were thinking in a very in a relatively limited uh you know one-to-one rather than many-to-many very good thanks so roger the top trending question on the q a session right now actually goes back to some of the emperor's uh new mind arguments about uh kind of a goodellion argument and the question is whether the same godelian restriction on logical mathematical systems also applies to science as a whole whether science as a whole is somehow restricted or its methods are restricted in this kind of self-referential way do you have any thoughts on that i don't know of any theorem to do with that i mean the thing about the girdle theorem is it's a theorem and it really does show you that uh we don't well i mean people don't think about it this way but but it's and people i thought i'm glad to hear people still worrying about it because it's it's something which you can't explain simply by an ai system as far as i can see i mean um it's the question of understanding i always like to phrase it that way that what it is that i mean consciousness does all sorts of things i mean clearly i mean people talk about pain and pleasure and uh and that sort of thing and but the main thing that i could say anything about i mean i can't quite see why you could say something enjoys itself and and uh worry about whether you can make a theory out of that but but what you can say is understanding of mathematical truth now you may say it's a very limited activity but if in that limited activity you could see that there is something non-computational going on then it seems to me that's a serious argument even though uh well i've been worried about the people forgotten the argument but the argument is is is pretty strong um especially in the context of evolution i mean you could see that okay the girdle theorem is pretty sophisticated or the good stein theorem which i often give in talks is uh i mean you can understand what the theorem is and you can even see sort of why it's true for the number four for example but how is it that the understanding that this thing is true came about when it's so far from from natural selection phenomena so it obviously was not specifically selected for that is to say an algorithm for doing very sophisticated mathematics was not naturally selected for what was naturally selected for was both general quality of understanding and that's the ability if you like to stand back and think about what you were thinking about and things like that and then girl serum is very much that sort of a thing you stand back and think about what you were thinking about and that's that's that's exactly you're trying to think about what somebody else would think about something and you can see in behavior i mean one of the examples i was like to quote is this the african hunting dogs and how they are they you could see them dividing themselves into two groups one group go and hide just where the the pass over the river has to be and the other one going to chase the antelopes there and then they pounce on them i mean they must have been communicating with themselves in some way to work out that strategy okay that's not solving the girdle problem or something but that's understanding of some kind and that i can well believe goes way way down in in in the animal kingdom not just humans and that quality okay it's been developed very greatly in human thinking and sure but it's something which has been there a long time and it has been naturally selected for and what consciousness is doing in this particular respect is giving us the quality of understanding that does other things too do you think that the difficulties we have in science at least in theoretical physics of understanding the measurement problem or other great questions of our age the question is beyond this i suppose that was rather lost the question in rambling in my way no yeah the question is is there some limit to our understanding in some sense of physics i don't quite see where this should be and i don't know any sort of girdle type theorem it says you can't have that okay you can't have a theory which tells us what i'm going to do next and then do something else i mean sure yeah you have that kind of limitation so it can't be a theory where you could um make a computer work out what i'm going to do in 10 minutes and tell me what the answer is before 10 minutes and that's how damn we're not going to do it so you can see you've got you you've got to evolve avoid that kind of paradox and clearly you've got to avoid that kind of paratox with the retroactive scheme which i'm putting forward so so that's absolutely clear so that's the thing which is but it's not the same as i think the question was worrying about whether a limit ultimately to the to the development of physics i don't see why they should be it may be you know too complicated to work things out or make difficult another some other sense to work it out but i don't see why principle one shouldn't get to the root of it all right we're not there yet away from it well those of us trained in physics will be relieved to hear that because that's what our life's work is yes so stuart there's been a couple questions that um actually has been a number of questions about some of the nitty gritty of orca arm microtubules but let me just pick one of them as a representative question which is memory so can you elaborate for a couple minutes on how memory works because this is i think you're proposing if i'm following a pretty radical shift in neuroscience of memory yes uh this the uh standard idea is synaptic plasticity synaptic strengths you have a neural network and the strengths of particular synapses in the network will channel information activity through the network in a particular way and synaptic strengths and the famous long-term potentiation experiments bliss and o'keeffe and those guys showed that a high frequency input causes a prolonged sensitization of that synapse and donald hebb and all that so that's that's pretty standard dogma the problem with that is that the membrane proteins at the synapses the both on the postsynaptic side the receptors and the site and the gtp cyclic g proteins and all that and on the release side are transgender only last hours to days and memories can last lifetime so these proteins have to be replenished and how are they replenished well by the microtubules with the motor proteins carrying those things along as i showed in one of those early slides so the microtubules are involved in in memory but most people would still say it's in the synaptic strength but that's too short now what we showed in our paper uh with the cam k2 travis craddock jack jacinski and i was that cam k2 which is not well understood this hexagonal enzyme and is involved in memory can imprint up to six bits of information per cam k2 and with every synaptic influx there's hundreds if not thousands of these cam k2s that spread all over the the neuron the dendrite uh tree dendritic tree so uh and the phosphorylation fits so um it i think it's a it's a valid mechanism and uh um we get a lot of hits on our paper but it hasn't made a dent in neuroscience uh because everybody's so entrenched uh just like everybody's so entrenched in neurons being the fundamental units of information processing the brain and we know that can't be right but when you think about a single cell like a paramecium or even an amoeba that can solve problems a paramecium swims around finds a mate can learn all these things it doesn't have any it's not part of a network it's one cell all those creatures use their microtubules so would our microtubules be sitting around acting as bony skeletal support and not taking advantage of their information processing abilities i don't think so and we need the we need the capacity for memory storage and each tubulin can be modified in so many different ways post-translational modifications phosphorylation uh binding of various things so that each tubulin can have let's just say 10 states so you've got 10 to the ninth raised to the 10th power of possible states per neuron so the capacity of memory storage uh at that level is enormous then the problem is well how does it relate globally and so you need entanglement you need something quantum so that memory is distributed and we know memory is distributed anyway through lashley and priebus and so forth so um i think it's a logical uh proposal uh i'm somewhat dismayed that it hasn't been picked up on because memory is still a mystery and uh the whole thing with alzheimer's and and memory loss uh people should be looking at the microtubules and uh in fact we've proposed using ultrasound ultrasound uh this study yesterday and uh sasha bastritsky mentioned a couple of different mechanisms he said well stewart has his own ideas well my idea is that the ultrasound acts on microtubules because they have megahertz uh vibrational frequencies and we actually have some preliminary data on that we've we've never uh published and uh but i i think uh that's a way of addressing memory issues and alzheimer's and and other uh neurological disorders is by approaching the microtubules not just working on receptors that act on membrane proteins wonderful thank you so um part but we've had a few questions that are in very much like tell me about x uh perspective so let me kind of pack those together can you give us just a quick rundown of where some of your experiments to the extent you can tell us where your experiments and progress stand what's the next milestone we can look forward to now that supremacy's been achieved just a quick status report yeah so this would my day job i should say is um not about building uh conscious or feeling any meds my day job is um concerned with building a large error corrected quantum computer and finding a scientifically or commercially relevant applications for it and for those who are interested we host um an annual event it's called the quantum summer symposium and the video of the talks just went online so there is actually a half hour talk if somebody wants to know what where does the quantum ilab at google stand today and what is our roadmap going forward there's a talk that people can look at um but i can maybe highlight where we want to go so we indeed spent significant time flashing out um a roadmap and the anchor points are we believe it will take us about 10 years we joke it will be before the end of the decade to allude to this kennedy quote that we have this error corrected machine ready and there's a sequence of technical milestones leading up to it our next technical milestone is to demonstrate that quantum error correction can work in principle it has never been shown that let's say you build a little grid of qubits let's say three by three qubits or five by five qubits and then you surround them with these measure qubits that do the round about parity measurements that i uh talked about so what you need to see is if i have let's say a five by five grid of data qubits relative to a 3x3 grid then the logical error rate needs to come down this has never been shown and this is our next milestone so we hope um to demonstrate that and then yeah there's more milestones than leading up all the way to the one million machine but i stopped here can i can i just add george to that again again from a layman or a more sort of commercial um perspective and i said during our talk you know what excites me about what uh hartman and his team are building there are these machines that um solve problems the way nature solves problems we would call it biomimicry i guess and um in the the commercial the the most likely commercial applications of that could potentially be in things like um quantum chemistry um and you know for me if we're starting to be able to build artificial trees that take carbon dioxide out of the air and turn it into oxygen that feels to me like a you know a good thing for humanity obviously in a you know a proper commercial application of relatively um you know early generation quantum machines wonderful thank you so um roger you have a couple of hundred audience members out here i don't know what the latest count 198 online and maybe 197 of them are still trying to get their minds including myself around the retroactive ideas that you have so one question came up that may help us zero in on this is whether it would help understand the delayed choice experiment or some of these other experiments quantum steering uh into the past that might how does your view bear on those those classic experiments well i'd forgotten them when i was i don't think really because i i don't think the multiple delayed choice is that what you're referring to there was a thing called it's a wheeler type of experiment yeah i never could quite see why that was a problem but maybe because the choice is made it's not made earlier i think it was a way of looking at quantum mechanics which isn't mine but i'd have to be reminded what it is again because i remember thinking that i didn't quite see why it was a problem um it's it's to do whether something is a position state or or is it is it an interference thing and you i think the idea is that between the state preparation so after state preparation but before measurement the decision is made that slits are opened or closed for example they're not oh yeah only the state preparation can't be responsible then for what the output is that's right i mean yeah i mean the state is something or other um and whether you choose to close this i mean suppose there's a photon coming along and it doesn't know whether you've chosen to open this listener well that photon has a state which is spread out it's not localized sure so only part of it goes through the slit so a lot of it gets lost but the part which goes through the slit is localized there it's not as though it's decided beforehand and it's going to go through the stitch or something i didn't quite under i think i never quite got the hang of the what the problem was um let me um then the the audience member's question in a slightly different direction there have been other proposals like ken wharton hugh price and others for retro causal interpretation of a bell experiment that the correlation might reflect a backward in time for john cramer ruth castner i don't know uh yeah yeah here did that kind of thing does your view relate to that i will you see when i was writing my notes i was perfectly well i was going to put haran offs and all that stuff into the model and thinking of it and then i realized it was different as far as i can see i mean they have two forward propagating and a backward propagating wave function and so that when you you have a measurement here and then a measurement here here and here means in time earlier and then later and this measurement reduces the wave function to to a reduced state here but then there's a corresponding wave function which is reduced by the laser measurement and they come back and it's quite useful to consider those two wave functions together because you can look at scalar products and so on and there are these vitamin experiments which are very complicated and new houses little things going on in the middle and so on it it helps you to understand some of those um situations i don't think it's the same thing that's what i'm saying because they tend not to be i don't know what a hair loss view is on on the reduction of the state i know i've talked to him about it but i can't uh um lev weidman i've talked to certainly about this well they tend to be many worlds type of view well for me what a key message though is independent of the specifics that your proposal which will strike a lot of people as just crazy oh yeah it has to be crazy but it's very much in keeping it's it's something a lot of people have thought about and it's not crazy it's really it's just objectively not a crazy idea and i'm very excited to see how you develop it when i say it's crazy i mean it's unlike our current thinking we see my cosmological model is crazy too but then we seem to see confirmations of her um this is crazier but that doesn't mean wrong it's just when i say it's crazy i mean you know quantum mechanics is crazy even the earth moving around the sun is probably pretty crazy isn't it it's increasing familiarity i suppose yeah and when you get used to the idea i mean even yeah the earth is rotating and not the sun moving around it it's pretty crazy because you can see so i need to get the final final word to rogers credit i would say it's experimentally testable and hence it's science yes you know and that's what excites me about okra in general that it can be tested it's it's not just floating off in the unempirical clouds but certainly the retroactive as the last experiment i mentioned rather quickly in some form is is testable i mean if that's certainly in theory testable whether you could make it a practical one i think it probably it's not so far from the bomb mr experiment that you just put a few more cavities in the experiment probably take another five years to do it or something but i can't quite see why you couldn't modify it to to test it in this way whether you could do make buzzing sound condensates that's really exciting i'd like to think more about those um because they're much more flexible you see you can do all sorts of things so stuart um i'm passing the baton to you i understand that you have an announcement to make for the group and then we can all break and go get some lunch or dinner or whatever it is in our time zone well thank you george and i'll hand it back to you and i i basically wanted to thank you and the other uh speakers i thought it was a great session uh this is my last time uh on screen for this conference so i want to thank all the speakers all the participants i want to particularly thank the guys from commotion studios who handle our av and tech stuff and i want to thank abby baejar montefiore who is wonder woman and she and commotion i gave them in retrospect unreasonable expectations uh and yet they uh pretty much they met them we've had a few glitches but uh generally it's run uh very well and thanks especially to abby she's always trustworthy she's always reliable she always comes through and the same can be said of commotion so thanks to all and i'll send it back to you to say goodbye and thanks good seeing all my friends out there good seeing you you

Channel: The Science of Consciousness (TSC) - Conferences

Views: 6,774

Keywords:

Id: 9YdNg2rEBW0

Length: 72min 55sec (4375 seconds)

Published: Tue Oct 20 2020