Future of Science and Technology Q&A (April 26, 2024)

Hi everyone. Welcome to another episode of about future of science and technology. And I see a bunch of questions here. Let's see, what should we try first here? M. Rudo asks why use cloud computers if we can compute on the moon, if we could get rid of the latency? Well, you know, I think one of the jokes among some people has been can we have the cloud in the cloud, so to speak, can we have either satellites or high altitude drones that get their power from solar power and get their cooling because they're in a cold place like in the upper atmosphere or in space and just compute there? And the answer is probably yes. I think it might be easier. I mean as you get the, you have to be in kind of low Earth orbit or in high altitude drones or you have too long latency and also you end up with radiation damage and things if you're outside of, if you're too far out from the magnetic field of the Earth and so on. I think it seems quite feasible that one day the cloud could be in the clouds or above the clouds actually, because we're talking probably 60,000ft or something, which is above pretty much every cloud. I think the, and there's, you know, now there are all kinds of tests and probably going into production soon. Laser based communication between, optical communication between satellites. One can imagine the same thing between high altitude drones. It seems fairly realistic to do this kind of thing. I don't know. So, you know, what does that mean for the, you know, is that economical compared to just racks of things in data centers? Might very well be. Depends on the cost of power, depends on, you know, and the, and the corresponding costs associated with cooling and so on. I think it's one of these things where as soon as people are producing gazillions of high altitude drones or satellites or whatever that are compute capable, the cost of the actual hardware to do that is going to go way down and seems like a thing we may very well see. I mean, I think one of the funny things that certainly with a high altitude drone solution you can imagine. Well, first of all there's kind of an advantage if you're at closer to the equator latitudes because then the amount of solar radiation falling per unit area is higher. It's kind of like it's a bad deal for the folks in, well, the very northern parts countries. There aren't so many countries in the very southern latitudes, but just because the southern ocean takes up that, that range of latitudes. But I think the kind of amusing vision is you've got all of this compute and communications going on in, for example, high altitude drones. And kind of every, you know, every year or so they come in for maintenance or something like that. And it's very much like the kind of migratory bird patterns, but it's high altitude drones, not birds. Which brings me to another question here from Onik. Would we consider artificially engineered protein based organisms or robots or nanobots? Well, there's sort of different levels of this. I mean there's things that are cells, regular biological cells with the usual cell membranes and things like this. There's things that are kind of artificial cells with non lipid bilayer biological type membranes. Then there are things that are more like individual molecules or clusters of molecules that don't have the sort of whole machinery of a cell. I think one has to realize that drugs are basically individual molecules that bind to things in us or whatever else to have certain effects. And typically they are just molecules that have a certain shape and certain charge distribution and so on. And they just, you know, they fit this particular slot in some protein or some other kind of molecule or whatever and then have some biological medical effect. The first level of can you make a kind of machine for that kind of thing is sort of an algorithmic drug where instead of the drug just being, oh yes, I have this shape and I fit into this slot in this other molecule. It's something where it kind of has the equivalent of sensors on the molecule where in a sense there are processes going on right inside the molecule which are like, well, you know, if it hits this piece and that piece, then this will happen. So that's kind of one level. Another level would be sort of a cluster of molecules that all sort of act together. I don't think that's a thing that is done. I mean, there are certainly drugs where particularly for things like cancer, where there are a whole cocktail of different drugs. But that's not usually something where one thing happens and then the other molecule comes along and it does that thing. But that is really common in biology, at least at the, well, let's see. Certainly at the cellular level that's really common to have sort of a cascade of actions. So I mean, one thing is you have an individual molecule that has sort of, has lots of different things going on on that molecule. The other is you have sort of a sequence of different molecules. I mean, any typical synthetic chemistry reaction consists of you do this, then you do that and one molecule will sort of protect some region on another molecule. So that when the other, when some new kind of Molecule comes in, it won't bind to that thing, but then you remove the protection in a later step and then it does bind to that and you get different things to happen. So I think one level is kind of almost the chemical level. If you send in the molecules and the molecules will either for an individual molecule or between multiple molecules, essentially execute what one might think of as kind of logic for doing things. Another level is where you have more sort of cellular scale things and sort of one approach is much used in. Well, in, I guess the. That's actually not yet much used. But, you know, stem cell therapies and so on are based on the idea of you've actually got cells that can go in and do things. There are also things like monoclonal antibodies, where you've got things that are just molecules that have very specific binding to very specific sequences, molecules, things in biological systems. But I think the potential vision is, do you get kind of one question would be, can you get kind of a general purpose sort of biological actor, so to speak, which is kind of like your own specially programmable cell where you've got a thing that you can stick in the bloodstream or something, and it will go through and, you know, you program it somehow and it will go and it will bind to this, it will bind to that. It will do certain actions you wanted to do. Now, we already have something a bit like that in the immune system. We already have the idea in the immune system of whether it's B cells or T cells or whatever, these are antibodies or T cells. These are, well, in the case of antibodies, just molecules. In the case of T cells, it's a whole cell with a particular receptor hanging out of the side, so to speak, that are set up to very specifically bind to particular amino acid sequences that occur on proteins and which have a certain shape that's determined by that sequence. So I think. There are different kinds of things one can imagine having happened in sort of the nanobots. One is just essentially binding to things as where, you know, you have this little cell or whatever. And it's just I'm going to attach myself to such and such and, you know, have this action. Now the thing is, I'm going to respond to a signal. I'm going to notice the level of such and such in the blood. And if it's above this, I'm going to start secreting some, some particular chemical. That's something that we have, you know, plenty of organs in our bodies work by doing that kind of thing. One can imagine making kind of the nanobot version of that, where you've got just the individual cell that does those things. I suppose another level is where you do things like the immune system does, where you can actually engulf a cell, you can kill a cell, Things like this. That's more sort of mechanical action, so to speak. Another thing that one can do is things like transporting different kinds of atoms around and sort of delivering this atom to this particular place and so on. But I mean, I guess that the. There's sort of a question of how do you get there? You know, in other words, how do you get to the place where you have these kind of nanobot like things? One approach is to say that you do something very much like drug discovery, where you're just saying, can I find a molecule that has the right properties to do this particular kind of thing? Another is, do I, you know, use a CAD system and design this whole sort of micro spaceship, so to speak. That is my artificial cell that's supposed to have all these different features and has a door here and a set of pincers here and so on. Is that what one ends up doing? There's sort of been a long time challenge, I suppose, in thinking about sort of nanotech kinds of things which has to do with, do you take machinery that we kind of mechanisms that we know work that we're familiar with on a large scale, you try and find a microscopic version of that, or do you say, well, we know what objectives we're trying to satisfy. Let's make something that may look sort of structurally very different from things we've already recognized. One of the things that's a surprise in molecular biology in the last few decades is the extent to which there is mechanism that's quite identifiable that one notices whether it's muscles that have these little walkers that walk along them, whether it's flagella that sort of let bacteria swim, that have these kind of rotary motors in them, things like this. These are mechanisms that in some cases would have been hard for us to kind of come up with as, oh, let's just invent a molecule that can do this little walking thing. We probably wouldn't have managed that, but yet somehow that's been found by biological evolution and is now used for these purposes. And I guess that's sort of a question of how do you do the design for this nanobot that you're going to use? Are you taking mechanisms that already exist in nature and putting them together? Are you inventing things from scratch? Are you figuring out sort of how these different molecules will interact and which molecules you should use. And I think my own sort of guess is that the engineering of these things is going to be more difficult than just the kind of finding by our sort of simulation and effect of biological evolution or sort of search through biological space. In other words, that if you can define the objective well enough that it will probably be easier to just find something that achieves that objective, even though we wouldn't have designed it and we might not immediately understand how it works, rather than saying, let's try and design from scratch and make this complicated machine that we understand. Now, having said that, kind of the history of drug discovery is a little bit like that, and it isn't necessarily the happiest one in the sense that the typical sort of workflow of drug discovery is find a target, find a target, bind this thing that you're trying to bind to this effect you're trying to have, et cetera, et cetera, et cetera. Then try to identify which molecule will sort of achieve the objective you want there. And there are different approaches that have been taken. I mean, most big pharma companies have these giant libraries of compounds and you basically can just try a bunch of molecules and see what works, given that you know what target you're aiming for. Another, another thing that's been widely done is, well, it's had waves of enthusiasm. It's sort of what used to be called rational drug design. That is, I know the kind of thing that I'm supposed to plug my drug molecule into. Let me design from scratch a molecule that has the right shape and so on. And that's kind of segued in modern times to sort of AI drug discovery, where it's kind of learn from all the examples we've had before and try and figure out what will work now. And there are endless companies that are trying to do this with, with so far without dramatic successes. I mean, there are, there are drugs that are being, you know, in various kinds of trials that have been developed that way. It's always a little bit unclear in the end how much AI and how much scientist there is in what comes out. I mean, another approach that's pretty commonly used is scan some part of the sort of tree of life. You know, you go after the fungi, the bacteria, the bacteriophages, the whatever, these different kinds of things from the tree of life, and you say, let's find one of these things where biological evolution already had a solution to this, already had, like the bacteriophages that led to CRISPR and gene editing and so on that already had sort of a solution for, for doing something. And then let's piggyback on that solution and use it to make our sort of thing that achieves the objective we're trying to achieve. But I think the ultimate story is sort of the general purpose artificial cell that you just say, I've got all these artificial cells and I can program them to become, to start, you know, doing whatever you want them to do. You know, secreting this or that chemical in this or that in response to this or that, surveilling for this or that kind of cell that exposes the fact that it's been affected by a virus or all kinds of other different things. I mean, that's probably the thing that one would like to see happen. It's kind of like what happened with general purpose computers in the idea of, you know, what launched sort of the computerization of the world was the fact that you could have a general purpose computer. You didn't have to build a separate machine for every function that you wanted to do. You could just have that one piece of hardware and you could sort of modularly add software that does those particular things. Right now we are in a situation with biomedicine and so on, where we have a drug, it does a particular thing. We have, you know, what we. The question is, can we be in a situation where we can have sort of the arbitrarily reprogrammable kind of platform for biological action, so to speak. And I think the, you know, that, that one, one approach to that is sort of upstream. Can you make molecules that have some particular form? Another is you've got this sort of artificial cell that can do all these different functions. And that is sort of like the general purpose computer of biomedicine, so to speak. And I think it's conceivable that such a thing could be made. We're not super close to making it yet. Well, W7 asks, is it possible to design sharpen the brain by some kind of biological engineering? You know, we have a hundred billion neurons, give or take. What would it be like if we had a trillion neurons? We don't really know. It's a notable feature if you look at sort of the progression of organisms with bigger and bigger brains, so to speak, that sort of. Some things seem to happen as you get bigger brains, like maybe the invention of language, which is a very important piece of being able to sort of communicate thoughts formally and abstractly. It's sort of the driving force behind kind of the achievements of our species that the development of language as a sort of modular way to communicate things might have happened at some brain size. It might be that by the time you have 100 billion neurons, you're at the point where you can invent something like that. And so then the question is, okay, let's say we got to a trillion neurons or 10 trillion neurons. What's the next level of what we might invent? Interesting question. Hard for us to foresee. I'm not sure that if you asked your average cat or dog whether, you know, human, like, compositional language, you know, it's like, I don't think it would conceptualize that. Hard to know really, but it's kind of. It's hard to know from the level we're at, what's possible at a higher level. I think the, you know, the question always. For brains. Well, they're different kinds of things. There are, you know, at one level, it's, you know, people say, you know, drink caffeine and you'll be sharper. I. I don't personally drink caffeine at all, but. But that's just my. My idiosyncrasy or something. I. It's. It's kind of like, you know, take. Take your. Your thing that. That sharpens your brain by. And it's very clearly the case that if you look at the actual behavior of nerve cells and junctions between nerve cells and you make them swim in alcohol or something like this, it changes those responses. They become more sluggish, they become this, they become that. You clearly can have those effects. Is the brain set up so that it is optimized, given what it is? Or can you just feed in this thing and make it better? I guess my personal. Perhaps squeamish prejudice would be that you're better off with what you have, so to speak. But hard to know for sure when it comes to, could we imagine something? I mean, it's hard to interface with the brain, so to speak. You have some. The main interfaces are. Well, there are purely chemical ones, and then there are just what we see, what we hear and so on. Obviously, one can imagine. And there are all sorts of efforts along these lines to have direct connections between digital electronics and nerve cells. That's, you know, ultimately nerve cells operate in an electrical way, and so one can expect to have those interfaces happen. I think I've talked about this a bunch of times here. The kind of. The neural user interface problem of what is it when you've actually got sort of that interface to nerve Cells you've got, the nerve cells sort of are naturally doing certain kinds of things. Can you kind of learn what those nerve cells, when those nerve cells do certain things, what does that mean to the brain that has those nerve cells in it? You know, when there's this particular configuration of nerve neuron firings, does that correspond to the concept of, you know, now I want to figure out this or that thing, or now I'm thinking about this or that thing. That's kind of the challenge is to, is to kind of figure out what it means when you see such and such a thing happening. And there are a bunch of experiments, companies and so on that are making some progress along these lines, I think. And similarly, on the actuation side, when you think this, does it mean, you know, can you, can you get to the point and people have been doing experiments on this since the 1950s of, you know, when you think this, does it mean, you know, flip the, the military plane upside down or whatever it is? The, you know, in, in what. How do you decode kind of the. Both the ideas that happen in brains and the sort of the, the. The desire for action that's defined by a brain? And I think, Yeah, I think that that's, you know, I think we can expect that there will be sort of neural user interfaces to brains and what their effect will be on kind of, kind of the. At first it will be kind of like, you know, even when you drive a car, when you use a computer, whatever, you eventually get to the point where it becomes an extension of yourself. And we feel that as a kind of. We feel this is an extension of ourselves that will presumably become more direct if we have direct neural connections rather than the, the thing. That's the thing being connected to you through your fingers and your eyes and, and things like this. I think that, You know, that that's something we can expect will be coming now. You know, what will it feel like when you sort of more and more of your brain and your memories and things like this are taken over by digital electronics? It's an interesting question. I think if it happens gradually, one will just feel like, well, it's me still and it's me, and I guess I'm running partly inside the digital electronics as opposed to inside the neural tissue. And you probably won't notice the difference. If it's like, well, we've got this pure digital electronics and that seems to be doing you like, things well, we won't have sort of an internal view of that. My, My guess is if it's sort of a, a gradual process. We'll mostly just feel like it's just us in there and, and then, you know, the characteristics of the us that we have. For example, being able to potentially make a, you know, a, A, a mind backup, so to speak, where you're just, by the time it's all digital electronics, you're going to be able to make a copy of that mind. And that will have all sorts of interesting and strange effects. I mean, I certainly wonder, even at the level of kind of modern AI bots of ourselves, what the interaction with a bot that's an accurate model of yourself will be like. I'd like to think in, in my case that I have a good time with that, but I'm not sure Amon is asking, do I. How do I see neuralink and similar technologies improving lives or having drawbacks? I mean, it's clear that in cases where sort of things have gone medically wrong with the brain or connections to the brain, it's clearly, it's clearly a win, so to speak, to have some prosthetic mechanism for, for having those connections and so on in terms of, of, you know, what could possibly go wrong. I mean, there are many things that could go wrong. I mean, it is the case right now that kind of the, I mean, the, the insides of brains are kind of our, in a sense, you know, private domain, so to speak, in which we get to think, whatever we get to think and it's, you know, as one makes more visibility into that, it becomes both for the person themselves and for their sort of role in the world and society and so on. There are all sorts of things that could go sort of horribly wrong there. I mean, I think that we tend to, at least many people, places, cultures, tend to value kind of free will, tend to value sort of the existence of individuals who can do different kinds of things. By the time sort of brains are all kind of connected with digital electronics and sort of everybody's on the same web, so to speak, right down to the level of what's happening inside their brains. In a sense, you are, you are, you know, you're becoming much more hive mind than you are becoming sort of a collection of billions of individuals. And in the, what could possibly go wrong, much could go wrong in that situation. I mean, I think that the thing that's good about lots of individuals, it's a thing that's sort of been discovered by biology, is having lots of individuals means lots of different things can happen. And if some things go wrong, it doesn't mean everything goes extinct because even though some individuals walked off the cliff, others did not. But if you have something where everything's locked together and you know, all, all the critters are yolked together and they're either going to all walk off the cliff or not, then you can have a situation. It's much easier to have a situation where they all walk off the cliff. And I think that's a, you know, that's something where it's complicated because in a sense then the issues of, I don't know, computer security and all these kinds of things of know, do you end up with, you know, people talk about mind viruses, as in ideas that kind of, you know, that that kind of very much replicate for themselves, so to speak. But that can become quite literal by the time there's lots of digital electronics involved and we no longer have the kind of the firewall of. Well, it's inside a brain. And the only way you can get to the inside of that brain is, is by talking to that brain or by showing that brain something. You can't just go in directly inside and do different kinds of things. So much could go wrong there. And I don't know, I think there are many kind of coming issues of sort of the future of the human condition. I mean, throughout the history of our species, the human condition has been sort of unchanged. I mean, there are details that have changed. We lead in many places kind of cushier lives and we have different motivations and all this kind of thing. But fundamentally the living and dying kind of arc of the human condition has remained pretty much the same. It will presumably change. Presumably there'll be various forms of immortality, whether biological, digital, whatever else. There'll be different ways in which there's sort of communication. I mean, the, you know, in a sense when people say the world has sort of been changed by the presence of sort of instant, broad communication, at some level, maybe we haven't seen anything yet because by the time that communication is at a much more deeply sort of inside the mind kind of way rather than just from the outside of the mind, it might be very different. So know, I think we are, we are sort of heading for lots of things that can be very different. And I think the, the kind of sort of we're having AIs that can do lots of kinds of things that we thought were just things only humans could do. We are, you know, there'll be. There's the question of sort of what, how we think about kind of humanhood in a world where sort of humans have lots of digital electronics sort of integrated into them. AIs are just start from the digital electronics side and so on. It gets even more complicated when you think about sort of organoids, where there are biological nerve cells that are directly that were sort of grown just to be clumps of nerve cells and connected to digital electronics. It's kind of a complicated configuration of things. And I think that the kind of how we feel about and how we treat personhood is going to be sort of a complex issue in the future of what happens. I mean, I think one of the things that in a sense it's like, well, we don't get to choose what time in history we live at, but we are sort of given that we're at that time, we sort of develop to hopefully prosper in that time and to be sort of enculturated into that time. And I think one thing that perhaps is interesting to think about is at some time in the future when it's kind of disembodied souls operating in some pure digital electronics framework and where the things that are going on feel like to us it's like, oh, there's a bunch of disembodied souls playing video games for the rest of eternity type thing that has the interesting feature that to us today that probably mostly seems like, gosh, that's really pointless and a bad future for humanity. But in a sense that's just saying that we are sort of enculturated to our time and the things that we're doing now that we think are worth doing and so on, that's, you know, that's where we are and we're happy to be there. And hopefully and you know, this sort of time in the future that looks pointless to us today if you're embedded in that, in that setup and sort of enculturated that setup, it may seem like that's the thing, that's the most important thing ever. So it's kind of like you can say, well, it would be nice to kind of experience the distant future. But the distant future may be so kind of may have kind of aspects of it that are like, why do we care about this? We're sort of stuck in this situation that we absolutely don't care about. And the only thing that we really care about is sort of what things are like in the environment into which we've been enculturated, so to speak. Anyway, perhaps too elaborately, too much elaborate kind of future philosophy or something. Let's see. Luis comments. The speed of neuron influx limits brain size. Not quite sure what you mean by that. I think it is a challenge if you have a bigger brain, to do the thing that we humans very much rely on, which is to have a single thread of experience, to believe that we are experiencing the world in a definite way, rather than to say there are lots of incoherent things that are happening and they're sort of all happening separately and so on. We very much believe in the fact that a definite sequence of things happened to us. And as brains get bigger, it may be more and more difficult to maintain that. That notion of kind of unique attention to one thing. I don't know. I mean, it may be that at some level there's an optimal size of brain, given the hardware of neurons and so on. That sort of does as much as possible while just being one brain. And that if you go to bigger, more neurons, that effectively, in order to maintain this kind of three thread of experience, you'd end up with, you know, five brains inside the skull type thing where they're all in a sense having different thoughts and they're just sort of interfacing to each other in some high level way rather than having sort of a gazillion neural connections between them. They're almost. They're kind of separated in the way that individual human brains now are separated, that there isn't sort of direct neural transfer of information. Rather the information is packaged up into things like language and in that way sort of communicated to another brain and unpacked there. Let's see. Well, let's turn to a very different topic. It's a question from David here. In the future, how likely is it the physicists will be able to detect individual gravitons? You know, the history of sort of measuring things is a long history of places where just more and more sensitive measurements could be made. One's often at first saying, oh, one will never be able to figure this out. One will never be able to detect individual molecules. One will never be able to sort of pull out this tiny signal from this noise. The prediction of one will never be able to do this. Pulling out of this from that has essentially always been wrong. Even when there are theorems that say, given this situation, there is this much noise, we'll never be able to get a signal out of it. Turns out noise isn't always what you think it is. And turns out many of those successes have come from analyzing the noise. You know, where is the apparently random noise coming from? What Regularities are there, in fact, in that noise that we can make use of to pull a signal out of it. Okay, so having said that, the. I mean, at this point we are far, far away from detecting individual gravitons in a gravitational wave. We're dealing with kind of. Well, I'm sure I could work it out how many gravitons one has in a typical gravitational wave. But it's immense numbers. I mean, One's just got. It will be like saying, you know, detecting the individual graviton, it's like saying we've got the swell on the ocean and it involves all these water molecules, but let's pull one water molecule out of that and notice that it arrived in a particular place. It seems infeasible to us today. Now, how might we detect individual gravitons? I think the. Well, how do we detect individual photons? The typical thing we do for photons is. And by the way, the detection of individual photons. Well, okay, several things to say. First of all, photons have different energies. So, you know, if we go across the electromagnetic spectrum, visible light photons of energies of a few electron volts, if we go to something like radio waves, the photons have vastly lower energy. If we go down to X rays, gamma rays and so on, we can be getting photons where every individual photon packs, you know, millions of times billions of times the energy that a photon in invisible light would have. And it's to the point where, you know, you could have a photon with so much energy that if it, that it would actually impart. So, you know, you could have a photon with a jewel of energy. I'm not sure. I don't think we've. There are protons with a jewel of energy that we've detected in cosmic rays, but not, not photons, probably. But in any case, the. So an individual photon can have very high energy. When it has high energy, when it collides with matter, that photon will produce a whole cascade of particles. So in cosmic rays, for example, if we had a 1 Joule photon, 10 to the 19 electron volt photon then, and it, you know, hit the upper atmosphere, it would produce a giant shower of particles that probably be a mile wide by the time it, it got to the surface of the Earth, it wouldn't be, it would be, I don't know, millions of particles, maybe billions of particles that would be kind of the result of this cascade of taking the energy of that original photon and kind of breaking it down into all these particles. And many of those particles would be readily detectable because they would be charged particles that would produce signals in various kinds of detectors. Now, okay, what about. So that's for high energy photons. Photons have the feature that they sort of immediately can impart their energy to electrons and other charged particles. Gravitons are sort of the particles of gravitation. And to this point, we've certainly never detected an individual graviton. I mean, photons were originally identified back in. Well, their effects were kind of inferred around 1900 there. They're sort of the main effect that really sort of clinched the idea of photons was the photoelectric effect where high energy photon comes in, it kicks an electron out of a metal and you get to detect the electric current associated with that electron. So the question is, is there a similar thing that can happen with gravitons? And if you had a high energy graviton of that kind of energy, then I think the answer is yes, you would. Every massive particle would be kind of kicked by that very high energy graviton. As it is right now, the gravitons that exist. So, so one question would be, where would the highest energy graviton come from? How do high energy photons come from? Come to be? They come to be because they're very high energy charged particles that get their energy by being accelerated by sort of repeatedly passing through a magnetic field, for example, and going faster and faster as a result of that. And then those charged particles, when they are decelerated or when they hit something, they produce photons. That's how the so called bremsstrahlung breaking radiation, where when you take a charged particle and you accelerate it or decelerate it, it will emit photons. That's just a feature of how quantum electrodynamics works. And in fact, classical electromagnetism works as well. But in the case of gravitons, any massive object, if accelerated, will produce gravitons, just not very many of them. And there's a sort of a question of whether if you have a sort of. By the time you have something which has sort of enough energy to be producing gravitons, when it's accelerated and so on, you have, basically you're ripping space time apart. You can have a particle that has very high energy. And yes, its effect on the electromagnetic field is to do something very complicated and to have a very sort of discontinuous things that's happening in the electromagnetic field. But that's different from actually ripping apart the structure of spacetime So I think having something sort of high enough energy, high enough deceleration and so on to emit high energy gravitons, I suspect that's not possible if you have kind of any kind of continuum, space, time, if you haven't torn apart the structure of spacetime. So then the question would be, is there a way to detect, you know, how would you detect individual, individual gravitons? The way you detect individual photons is, is the story even for visible light photons is always. Well, you get that photon to have an effect that has, produces sort of a cascade where eventually electrons are involved, electric currents are involved, those kinds of things which you can readily measure. And how do you make that cascade? The most common way to make that cascade is, is by having a system that's somehow, a little bit, that's somehow unstable where, oh, you know, no photon has come in. Okay. It just hangs out and doesn't do anything. It's right on the edge of kind of a lot of things changing. And when the photon comes in, it kicks it over the edge and the system then changes and produces a bunch of electrons or whatever else that's true in, well, for example, a photomultiplier tube is what you have is a bunch of voltages between plates and so on, where they sort of really want to be pulling electrons out of things. And as soon as something comes in that kind of kicks it over, then there's this big cascade of electrons that gets produced and accelerated and so on. Could you do something similar for gravitons? You know, in, you know, I, I, I did some work once on a, on a movie called Arrival, which featured interstellar spacecraft and so on. And at some point the, the guys there wanted a whiteboard that was kind of a, a possible interpretation by physicists of today of the interstellar spacecraft that just arrived. That was the particular shape they wanted it to be because they thought it was cool for the movie and it was cool. And so I kind of spent a few hours writing this, inventing this kind of fictional physics that one might conclude if one saw this spacecraft. And actually the thing that I came up with was the thing that I was basically referring to as a graviton laser and the sort of amusing that I wrote out some things on this whiteboard and you can find it in a blog that I wrote, including some annotation of what the different pieces meant. But some physicists have actually taken that and thought a bit more seriously about the kinds of things that I raised in that sort of fictional physics. For that whiteboard. But the idea there was, could you make a graviton laser? And what is a laser? A laser is something where photons have this characteristic that they like to get together in the same state. Electrons have the exclusion principle, and two electrons have the feature that you'll never be able to get them in the same state in the same place with the same momentum, same energy and so on. They won't go there. They just refuse to do that. That's a large part of why matter is stable, for example, why we don't just fall through things. It's because those electrons are excluding themselves from being in the same place. So electrons, they have spin a half. That turns out to make them fermions, which means that they have this characteristic of kind of excluding each other. Photons have spin one, they have the characteristic, they have integer spin as opposed to half integer spin. That turns out as a result of kind of features of quantum field theory to mean that they are things that act as bosons and they like to be in the same state. They photons. If you try and you have something and you try and sort of shoot more electrons into a thing, and you say, I want to shoot an electron to produce something which has more than one electron in the same state. It's like, that can't happen. Zero probability. Do the same thing with photons. Then the probability that the photon will sort of catch and go into that state increases with the number of photons that are already there. So photons like to get together in the same state. And that's what's happening in a laser. In a laser, you're basically getting a bunch of photons to sort of coherently be in the same state. You do that by. Well, it. I mean, you can also think of a laser as being something where you've. Well, basically you're. You're getting atoms to be in an excited state where all the atoms kind of want to kind of jump down to the less excited state and emit energy in the form of photons. And they all want to do it together because photons all like to be in the same state. And that's kind of why lasers make this coherent lump of electromagnetic radiation that does what lasers do. So the question is, can you do the same thing with gravitons? Well, the way a laser works is you are pumping these atoms into this state where they're all ready to emit photons and go and produce a laser effect. You've got to sort of pump the atoms into this excited state. Where they will emit photons. And when that happens, you have to have the thing kind of in some sort of cavity where these photons are being kind of being reflected around so that the photons can kind of hang around together, so to speak. Well, the problem with gravity is that with photons, we know how to kind of have metal walls and so on that will reflect. We have mirrors, we know how to reflect photons, we have no idea how to reflect gravitons. Gravitons, so far as we know, aren't they. There isn't the same kind of notion of reflecting them. And part of the reason for that is that whereas with electromagnetism we have both positive and negative charges, and we have the feature unlike charges attract like charges repel. With gravity, all we have is mass and positive mass. And so far as we know, everything that undergoes that is subject to gravity has sort of the same sign of force. All things acting according to gravity attract. People in science fiction have had anti gravity forever in physics, in cosmology, there's the idea of dark energy, which is basically matter with negative mass that repels other matter as opposed to attracting other matters. But in terms of what we can have in a lab, we certainly don't know how to get that in the lab, whether it exists in cosmology or not. And so this question of can we make a graviton mirror, can we make a cavity that will sort of get the gravitons to be in a kind of laser like state, we don't know how to do that at all. And I think there are ways that you could potentially use existence of a graviton laser to have some way of detecting individual gravitons, maybe. I will say that the question of whether we can detect sort of the discreteness of gravity, the fact that gravity has discrete particles associated with it, that's sort of. Well, it's a different question from whether we can detect the discreteness of spacetime. My guess is that it may be easier to detect the discreteness of space time, which doesn't necessarily imply even the existence of gravitons, though it tells one something about what the fundamental theory has to look like and that from that one can deduce things about gravitons, Although we don't completely know how to do that yet. But I think it may be easier to see the discreteness of space time than to see the discreteness of gravity. And it's kind of a story. It's only been 100 and something years since we knew that matter was really discrete and made of atoms, and similarly that light could be thought of as made of discrete photons. The question of how one detects discreteness of space time is an interesting one. In our big physics project over the last few years, one of the sort of launching ideas is that spacetime really is discrete, which is something that hasn't been a popular idea for the last hundred years or so because people sort of couldn't figure out the mathematics of how to make that work, which is something that we have been able to figure out. And so there's a question, how do you tell if spacetime is discrete? And the guess is that there are phenomena associated with black holes and particularly sort of extreme situations with black holes where the black hole is kind of spinning as fast as it can be spinning and so on, in which kind of there will be one sort of pulling things out to the point where sort of the individual connections in discrete space time be more visible than they would otherwise be. But we don't yet know really how to make this work. There are other kind of ancillary effects, like for example, the presence of non dimension fluctuations in space. That space isn't precisely three dimensional. That's something that's to be expected if spacetime is discrete. It will be pretty weird to find 3.01 dimensional space time if in the standard kind of continuum theory of how spacetime works. So that would be sort of circumstantial evidence for the discreteness of space. But again, detecting that is, we don't completely know how to do that yet. I think I have to go soon to my day job here. But maybe one or two other questions there because this is fun, at least for me. This is one I'm afraid I don't really know the answer to. From Juni asking about latest innovations in battery technology, alternatives to lithium and cobalt in energy storage. I, I don't know the answer to that. I, I follow that only a little bit. I think in general there are, you know, we are subject to sort of the things that chemistry throws at us. And it is always an interesting question whether there is. I mean, sometimes it's like we just have to find a better chemical. We just have to find something. And sometimes we're really up against the properties of chemical elements and we're sort of boxed in and there really isn't, there isn't a better place to go in that direct path. I think in terms of batteries, there are a lot of detailed issues about what happens with electrodes in batteries. And Whether they start forming dendrites that sort of start. They start getting kind of this furry stuff on them that eventually shorts out pieces of the battery, things like this. Those are things which one can readily expect are sort of engineering problems that, oh, you have the right shape of electrodes and this and that, you have the right extra material around the battery that gets solved, as opposed to things where you're running right up against the electrochemical potential of this or that element. And there's just no way to get out of that box, so to speak, or no way to get out of that box by sort of a direct frontal assault, that is that in order for that problem to be solved by one isn't going to solve it by just saying, well, let's find another element, because there isn't another one to find. It'll be some completely different technology that allows one to solve it. But I don't know the details of that. Let's see, maybe one more. Well, okay. Weasel asks my last thing for today. Curious about my thoughts on hologram technology. That's right around the corner. So, you know what is a hologram? Hologram is kind of reconstructing the wave fronts of light. Normally we can just think of light arriving at our eyes as just rays of light. Just, oh, there's a photon here, there's a piece of light here, it arrives at our eyes. But in fact, light can be thought of as having this wave character where not only are sort of pieces of light arriving, but they arrive at a certain phase. They're like waves that are. That are oscillating electromagnetic waves that actually do oscillate, in the case of light, a trillion times a second or so. And there's a question of, are you at the place where the wave is going up or the place where the wave is going down? And a hologram takes kind of the light that's coming from something. It reconstructs the wave fronts of light as opposed to just saying there's this intensity of light coming from here and there. And the other place, it reconstructs the phase relationship. That is we. Where the wave was. Is that a wave that's coming up or going down at that particular angular position? So holograms. And then the. And so that's kind of the thing you're making kind of this thing where not only is there intensity, but there's also phase information. And that means that the light that you see is just like the actual light that would come from that object as opposed to what you might see. On a screen on display where it's just like it has the right intensity, but it doesn't have the right phase relationships. In radio, the concept of a phased array is a, is kind of a well known concept. Instead of if you just have one antenna, it's kind of sending out radio waves and it's going out in some antenna pattern that's kind of going out in different directions. But you can have a phased array where you're generating radio energy, where instead of just the intensity of the radio waves being important, the phase relationships are also important. And that allows you to just have this fixed antenna. And just by changing these phase relationships, you can get the electromagnetic energy to concentrate in one direction or another direction and so on. And for example, in 5G telephony, there's a whole bunch of phased array technology that has to do with saying let's not just transmit through from this one radio source kind of going in all directions and have kind of an inverse square law roughly fall off in the intensity. Let's instead arrange the phases so that things only so that all the energy is concentrated in one particular direction. So that's kind of the idea of holograms. Now the big challenge with holograms is yes, you can make holograms with lasers. You can etch plates with them. You can make things where you've got this static hologram that's made from kind of the, that's made sort of photographically. The question is, can you make a hologram dynamically? Could you make a hologram? Could you make a display that is a hologram that is producing kind of phased phase information where every pixel on the display not only has an intensity, but also has a particular phase. And that technology doesn't quite exist yet, although it's getting closer. And as one looks at things like mems, little kind of mechanical devices that use kind of the effect that certain materials, when you will change their, the will generate forces inside the material when you put an electric current, when you put electric potential, when you put voltage across them. And that allows you to move things around. That's what a bunch of display technology, digital light mirrors, I guess are. Digital light displays work this way. They have an array of mirrors effectively that are being moved using this effect, but the voltage moves the mirror. And so that causes the light to either get reflected in the direction towards the screen or not. And that's something that you can, you know, there are many projectors for many years have worked this way that sort of have a mirror that is microscopically configurable but it's not microscopically configurable. On the length scale, you need to change the phase of light. That length scale has to be comparable to the wavelength of light. So we're talking, you know, 500 nanometers. So that's. See, that's. That's half a. Yeah, half of a. Half of a millionth of a meter or something. The. So it's. The question is, can you make sort of a device that will kind of affect a material device that will sort of affect light on that smaller distance scale and fast enough a trillion times a second to be able to affect the phase of light? And we're not quite there yet. When we are there, the kind of the Princess Leia style projector, hologram into space type thing will absolutely be a reality. Actually, there are companies now that have, gosh, I was going to see a demo of one of these, but realize I haven't seen it. And these are demos that, you know, many demos. People say, I can show you this cool technology and it's like, sure, let's get on a video conference call and you can show it to me. But when the technology involves kind of a hologram, by definition that's not going to work. You actually have to see it with your own eyes, so to speak, and have those wave fronts of light intersect with your eyes. But I think the, the big coming attraction is being able to make dynamic holograms where you could make a holographic movie, for example, where you have a projector that is in real time changing the. Has something which in real time is changing the phase of light and allowing you to kind of project this thing as a movie in real time. And I think that. And probably that technology is both. The production technology and the measurement technology are things that are sort of coming attractions of what can you then capture sort of a hologram by. Instead of a camera, which is just measuring the intensity of light coming in different directions, your measuring all this phase information as well, so that you're capturing sort of the full wavefront of light. I mean, when people make 3D or movies, they just have a certain number of cameras, maybe just two or more, that are dealing with the fact that we perceive 3D because we have two eyes and we're able to see the parallax of different things. But a hologram is really making the complete kind of wavefront of how the light is being produced. All right, I should wrap up there. Need to go to my day job. Actually. I think I'm doing a live stream for my day job in just a moment here. Well, thanks for for joining me and thanks for bunch of those transfer questions and look forward to doing this another time. Bye for now.