Lex Fridman argues with physicist about a theory of everything

let me ask you just on this tension we've been dancing between physics and Mathematics um what to you is an interesting line you can draw between the two uh you have done some very complicated mathematics in your life to explore the laws of nature what's the difference between physics and Mathematics to you well um um I love math I think my first love is is physics and the math that I've done I've I've drawn to because it was needed and serve as a physics in service of physics but then of course in the In the Heat of it it has its own appeal and uh In the Heat of it I like it sure it has its so it's its own appeal and I certainly enjoyed it and ultimately I would like to think I wouldn't say I believe but I would like like to think that there's no difference between physics and Mathematics that or mathematics is realized in the physical world and all physics has a for a mathematical basis that they're really the same thing I mean why would there be math that had no physical manifestation it seems a little odd right you have two kinds of math some that are relevant to the real world where they don't have to be contradictory but you can have a can't you not have mathematical objects that are not at all connected to the physical world so I mean this is to the question of is math discovered or invented I said to you math is just discovered and and and there's a deep linkage between the two yeah yeah yeah uh do you find it all compelling these ideas uh like something like Max tegmark where our universe is actually fundamentally a mathematical object that math is our universe is mathematical fundamentally mathematical in nature my expertise is a a physicist doesn't add anything to that um it's not really you know physics is you know I was once very interested in philosophy and you know physics physics I like questions that can be answered that it's not obvious what the answer is and that you can find a an answer to the question and everybody will agree what the answer is and that there's a an algorithm for for getting there um not that these other questions aren't interesting um and they don't somehow have a way of preventing presenting themselves but to me the interesting thing is to is is motion in what we know is learning more and understanding things that we didn't understanding before things that seemed totally confusing having them seem obvious that's wonderful so I think that's those questions are there I mean I would even go further you know the whole Multiverse I don't I don't think there's too much we concrete we're ever going to be able to say about it this this is fascinating because you spend so much time in string theory which is devoid from a connection to the physical world for a long time like it did not devoid but it it it travels in a mathematical world that seems to be beautiful and consistent and seems to indicate that it could be a a good model of the laws of nature but there's it's still traveling independently because it's very difficult to experimentally verify but there's a promise Laden in it in the same way Multiverse or uh you could have a lot of kind of very far out there questions where your gut and Instinct and intuition says that maybe in 1500 200 years you'll be able to actually have strong experimental validation right I think that with string theory um I don't think it's likely that we could measure it but we could get lucky in other words just to take an example about 10 or 20 years ago it was thought that they had seen a string in the sky and it that it was seen by um you know doubled stars that were gravitationally lensed around the gravitational field produced by some long string there was a line of double instant now the signal went away okay but people were hoping that they'd seen a string and could be a fundamental string that it somehow gotten stretched and that would be some evidence for string theory there was also bicep two which it was a the experiment was wrong but it could have happened it could have happened that we got lucky and this experiment was able to make direct measurements certainly would have been measurements of quantum gravity if not string theory so it's a logical it's a very logical possibility that we could get experimental evidence from strength that is a very different thing than saying do this experiment here's a billion dollars and after you do it we'll know whether or not strings are real but I think it's a crucial difference it's measurable in principle and we don't see how to get from here to there if we see how to get from here to there in my eyes it's boring right so when I was a graduate student they knew how to measure the Higgs boson took 40 years but they did it I not just say that stuff is boring I don't want to say that stuff yeah but I I you know you know would Magellan set out he didn't know we could get around the world there was no map you know so I don't know how we're gonna um thank you connect in a concrete way all these ideas of string theory to the real world and you know when I started out in graduate school I said what is the what is the most interesting problem that there might be the deepest most interesting problem that there might be progress on in 60 years and I think it could be you know uh that you know in another 30 years that maybe we'll learn that uh we have understood how black holes store information you know that doesn't seem wild that that we're able to abstract what we learned from string theory and show that it's operative and and and and and you know I mean the Bose Einstein condensate they did you know they if you win Bose and Einstein uh predicted it when was that the 30s maybe early 30s it took they were there were 20 orders of magnitude that were needed in order to Improvement in order to measure it um and they did 50 years later so and then you couldn't have guessed how that had happened how they could have gotten that and it could happen that we I don't think we're gonna light see the heterotic string Spectrum at an accelerator but but it could be that things come around and and uh in an interesting way and somehow it comes together and the fact that we can't see to the end isn't a reason not to do it you know we just you know what did they do when they were trying to find the specific right they just they took every route they just tried everything and that's what we're doing and we're taking and I'm taking the one that my nose tells me is the best you know and other people are taking other ones and that's good because we need every every person taking every route and you know if somebody on another route uh find something that looks really promising you know I'm gonna make a Portage over the mountain and get on their stream you know so the fact that you don't see the experiment Now isn't to me a reason to give up on what I view as the most fundamental Paradox in 20th century 20th in present physics 20 21st century physics absolutely you can see that it's possible you just don't know the way but that's what I mean why some of the philosophical questions could be formulated in a way that's explorable scientifically so uh some of the stuff we've talked about but you know for example this topic that's become more okay to talk about which is the topic of consciousness uh you know to me as an artificial intelligence person that's a very practically interesting topic but there's also philosophers uh Sean Carroll loves to argue against them but there's the philosophers they're pen psychists I'm not against philosophers it's just not as fun I don't it's not fun right but they they uh they start a little flame of a fire going that some of those Flames I think eventually become physics So eventually it becomes something that we can really like having them around is really important because you'll discover something by modeling and exploring black holes it's really weird and having these ideas around like the ideas of pan psychists that Consciousness could be a fundamental force of nature just even having that crazy idea swimming around in the background could really spark something where that you were missing something completely and it's just that's where the philosophy done right I think um is very useful that's where even the you know these thought experiments which is very fun and sort of the the tech sci-fi world that we live in a simulation uh that you know taking a perspective of the universe as a as a computer as a as a computational system that process information which is a pretty intuitive notion but you can just even reframing it that way for yourself could really open up some different way of thinking could be and then you have I don't know if you're familiar with Stephen wolfram's work of like cellular atom and complexity yeah I did a podcast with Stephen Hey Stephen that's awesome I I to me forget physics to get all that uh cellular automata make no sense they're so beautiful they're so like they're from simple rules you can create complexity I I just don't think you know he wrote a book A New Kind of Science um basically hinting at which a lot of people offended as like we don't have a good way to talk about these objects we don't we can't figure out what is happening here these simple these trivial rules can create incredible complexity uh he's totally right about that yeah and I and physicists I guess don't have don't know what to do with that don't know what to do with cellular automata because you can describe the simple rules that will govern the system or how complexity it can emerge like incredible complexity yeah of course wolfram's version of that is that physicists will never be able to describe it right yeah exactly he tries to prove that it's impossible what do you make of that what do you what do you uh what do you make about the tension of being a physicist and potentially not being able to it's like uh Freud or somebody that maybe uh Sigmund Freud maybe you'll never be able to actually describe the human psyche is that a possibility for you that you will never be able to get to the core fundamental description of the laws of nature yeah so I had this conversation with Weinberg yeah how'd it go so Weinberg has this book called dreams of a final Theory yeah and I had this conversation with him I said why do you think um there's ever going to be a final Theory why should there ever be a final Theory I mean what does that mean the physics departments shut down we've solved everything um and you know what is it doesn't it seem that every time we answer some old questions we'll we'll just find new ones and that it will just keep going on forever and ever and he said well that's what they used to say about the Nile they were never going to find the end and one day they found it yeah so I don't I don't know um String Theory doesn't String Theory doesn't look like a candidate to me for a final Theory it as it stands now it doesn't get to the bottom of the world yeah it seems to me that even if we kind of solved it and we've did experiments there still would be more questions like why are there four dimensions instead of six it doesn't seem to have any anything that in it that would explain that you could you can always hope you know that there's something that we don't know about string theory that we'll explain it but but it but it still doesn't look like it's going to answer every question and um why is there one time not two you know why is this but you know it doesn't seem like it's I don't even know what it would mean to answer every question but to answer every question obviously so when you refer to The Theory of Everything you'll be able to have a if it exists it would be a theory that allows you to predict precisely the the behavior of objects in the universe and their their movement right what what about them their movement yeah like like precisely no matter the option right that's true so so that would be a really interesting State of Affairs if we could predict everything but not necessarily understand everything so for example let's just forget about gravity I mean we're not too far from that situation if we forget about gravity the standard model in principle given a big enough computer predicts almost everything but if you look at the standard model it's kind of a laundry list with neutrino masses and all that stuff there there are hundreds of free parameters where do they come from is there an organizing principle is there some further unification sure so so being able to predict uh everything is not the only goal that physicists have so on the way to trying to predict you're trying to understand that's actually probably the goal is to understand yeah but but but right we're more interested in understanding than actually that actually doing the predictions but the predictions are more focusing on how to make predictions is a good way to improve your understanding because you know you've understood it if you could do the predictions yeah one of the interesting things that might come to a head with is our artificial intelligence there's an increasing use of AI in in physics we might live in a world where AI would be able to predict perfectly what's happening and so that will as physicists you'll have to come to to the fact that you're actually not that interested in prediction I mean it's very useful but you're interested in really understanding the Deep laws of nature versus a perfect predictor yeah like uh you want to play jazz even within AI yeah AI people are trying to understand what it is that the AI Bots have learned in order to produce whatever they produce for sure but you still don't understand deeply especially because they're getting you know uh especially language models if you're paying attention uh the systems that are able to generate text they're able to have conversations Chad GPT is the recent manifestation of that they're they just seem to know everything they're trained on the internet they seem to be very very good at uh something that looks like reasoning they're able to generate you can ask them questions they can answer questions it just feels like this thing is intelligent right uh and I could just see that being possible with physics you ask any kind of physical question and it'll be able to very precise about a particular star system or a particular black hole you'll say well these are the numbers it's it'll perfectly predict uh and then sure you can understand uh how the neural network is the architecture structured actually for most of them now they're very simple you can understand what data is trained on huge amount of data you're giving it a huge amount of data from a very nice telescope or something and then but it it seems to predict everything perfectly you know how a banana Falls when you throw it like everything is perfectly predicted you still don't have a deep understanding of what governs the whole thing um and maybe you can ask it a question it'll be some kind of Hitchhiker's Guide to the Galaxy type answer uh that you know it's a funny world uh we live in of course it's also possible that there's no such deep simple governing laws of nature behind the whole thing I mean you there's something in US humans it's possible that wants it there to be yeah it doesn't have to be right right I do what's where do you again you're betting the for you already bet the farm but if you were to have a second Farm do you think there is a theory of everything that we might get at so um simple laws that come in the whole thing I don't I don't I I honestly I I don't know but I'm pretty confident that if there is we won't get to it in my lifetime I don't think we're near it but it doesn't feel like they're like the fact that we have the laws we do they're relatively simple already that's kind of incredible it's just there seems to be there seems to be simple laws that govern things right by Theory of Everything you mean Theory a theory of of everything an algorithm to predict everything but a simple algorithm a relatively simple algorithm to predict everything so for me it would be a sad day if we arrived at that without answering some deeper questions sure of course it definitely is that but the question yes but one of the questions before we arrive there we can ask does such a destination even exist so because the asking the question and the possible answers and the process of trying to answer that question is in itself super interesting because is is it even possible to get there well there's an equals mc squared type of there's a function okay you can have many parameters but a finite number parameter function that can predict a lot of things about our universe well okay but just to sort of throw one thing in in order to answer every question we would need a theory of the origin of the universe right and that is a huge uh task right so and the fact that the Universe seems to have a beginning defies everything we know and love right because we you know one of the one of the basic principles the physics is determinism that the past follows from the the present follows from the past the future follows from the president so on but if you have the origin of the universe if you have a big bang that means before that there was nothing and you can't have a theory in which something Falls from nothing so somehow sounds like you don't like singularities well I thought for somebody that works with black holes you'll get used to them by now no no I I like this because it's it's so hard to understand I like it because it's hard to understand but but it's really challenging us it's not a I don't think we're close to solving that problem so even uh and string theory string theory has basically had nothing there's been almost nothing interesting said about that in the last many decades so String Theory hasn't really looked at the big bang it hasn't really tried to get to the origin not not successfully not not in there there aren't compelling papers that lots of people have read that that people have taken it up and tried to go at it but but there aren't there aren't there aren't compelling String Theory doesn't seem to have a trick that that helps us with that puzzle do you think we'll be able to sneak up to the the origin of the universe like reverse engineer it from experimental from theoretical perspective like okay if if we can what would be you've already gotten yourself in trouble really because you use the word reverse engineer sure so if you're going to reverse engineer that means you know you you forward engineering means that you take the present and term in the future reverse engineering means that you take the present and determine the past yeah but estimate the best but yes sure but but but if the past was nothing how are you ever gonna reverse engineer to nothing well that's hard to do run up against the nothing right until they have mathematical models that break down nicely to where you can actually start to infer things foreign but it is people try to do things like that yeah but have not succeeded it's not it's not something that we we you know we're getting a pluses in sure let's pretend we live in a world where in a hundred years we have an answer to that yeah what would that answer look like who what department is that from What fields left led us there what uh uh not what feels what set of ideas and theoretical physics um is it experimental is it theoretical like what can you imagine possibly could have possibly lead us there is it through gravitational waves and some kind of observations there is the investigation of black holes is it simulation of universes is it uh maybe you start creating black holes somehow I don't I don't know uh maybe some kind of high energy physics type of experiments well I have some late night ideas about that that aren't really ready for Prime okay sure but you have some ideas yeah yeah but but the but um and many people do it could be that some of the advances in Quantum information Theory are important in in that they kind of go beyond taking Quantum systems and just replicating themselves but combining them with others

Channel: Lex Clips

Views: 140,251

Keywords: ai, ai clips, ai podcast, ai podcast clips, andrew strominger, artificial intelligence, artificial intelligence podcast, computer science, consciousness, deep learning, einstein, elon musk, engineering, friedman, joe rogan, lex ai, lex clips, lex fridman, lex fridman podcast, lex friedman, lex mit, lex podcast, machine learning, math, math podcast, mathematics, mit ai, philosophy, physics, physics podcast, science, tech, tech podcast, technology, turing

Id: kK8oKwBb9eM

Length: 24min 54sec (1494 seconds)

Published: Thu Feb 16 2023