Max Tegmark: Life 3.0

[0] Welcome to the Artificial Intelligence Podcast.

[1] My name is Lex Friedman.

[2] I'm a research scientist at MIT.

[3] This podcast is an extension of the courses on deep learning, autonomous vehicles, and artificial general intelligence that I've taught and organized.

[4] It is not only about machine learning or robotics or neuroscience or philosophy or any one technical field.

[5] It considers all of these avenues of thought in a way that is hopefully accessible.

[6] to everyone.

[7] The aim here is to explore the nature of human and machine intelligence, the big picture of understanding the human mind and creating echoes of it in the machine.

[8] To me, that is one of our civilization's most challenging and exciting scientific journeys into the unknown.

[9] I will first repost parts of previous YouTube conversations and lecture Q &As that can be listened to without video.

[10] If you want to see the video version, please go to my YouTube channel.

[11] My username is Lex Friedman there and on Twitter.

[12] So reach out and connect if you find these conversations interesting.

[13] Moving forward, this podcast will be long -form conversations with some of the most fascinating people in the world who are thinking about the nature of intelligence.

[14] But first, like I said, I will be posting old content, but now in audio form.

[15] For a little while, I'll probably repeat this intro for reposted YouTube content like this episode, and we'll try to keep it to what looks to be just over two minutes, maybe 2 .30.

[16] So in the future, if you want to skip this intro, just jump to the 2 .30 minute mark.

[17] In this episode, I talk with Max Tagmark.

[18] He's a professor at MIT, a physicist who has spent much of his career studying and writing about the mysteries of our cosmological universe, and now thinking and writing about the beneficial possibilities and existential risks of artificial intelligence.

[19] He's the co -founder of the Future of Life Institute, author of two books, our mathematical universe, and Life 3 .0.

[20] He is truly an out -of -the -box thinker, so I really enjoy this conversation.

[21] I hope you do as well.

[22] Do you think there's intelligent life out there in the universe?

[23] Let's open up with an easy question.

[24] I have a minority view here, actually.

[25] When I give public lectures, I often ask for a show of hands who thinks there's intelligent life out there somewhere else.

[26] And almost everyone put their hands up.

[27] And when I ask why, they'll be like, oh, there's so many galaxies out there.

[28] There's got to be.

[29] But I'm a numbers nerd, right?

[30] So when you look more carefully at it, it's not so clear at all.

[31] When we talk about our universe, first of all, we don't mean all of space.

[32] We actually mean, I don't know, you can throw me in the universe if you want, it's behind you there.

[33] It's, we simply mean the spherical region of space from which light has had time to reach us so far during the 14 .8 billion years, 13 .8 billion years since our Big Bang.

[34] There's more space here, but this is what we call a universe because that's all we have access to.

[35] So is there intelligent life here that's gotten to the point of building telescopes and computers?

[36] My guess is no, actually.

[37] The probability of it happening on any given planet is some number we don't know what it is.

[38] And what we do know is that the number can't be super high because there's over a billion Earth -like planets in the Milky Way galaxy alone, many of which are billions of years older than Earth.

[39] And aside from some UFO believers, you know, there are, There isn't much evidence that any super -drawn civilization has come here at all.

[40] And so that's the famous Fermi paradox, right?

[41] And then if you work the numbers, what you find is that if you have no clue what the probability is of getting life on a given planet, so it could be 10 to the minus 10, 10 to the minus 20 or 10 to minus 2, any power of 10 is sort of equally likely if you want to be really open -minded, that translates into it being equally likely that our nearest neighbor is 10 to the 16 meters away, 10 to the 17 meters away, 10 to the 18.

[42] By the time you get much less than 10 to the 16 already, we pretty much know there is nothing else that close.

[43] And when you get beyond 10, they would have discovered us.

[44] Yeah, they would have been discovered us long ago, or if they're really close, we would have probably noted some engineering projects that they're doing.

[45] And if it's beyond 10 to the 26 meters, that's already outside of here.

[46] So my guess is actually that we are the only life in here that's gotten the point of building advanced tech, which I think is very puts a lot of responsibility on our shoulders, not screw up.

[47] I think people who take for granted that it's okay for us to screw up, have an accidental nuclear war or go extinct somehow because there's a sort of Star Trek -like situation out there where some other life forms are going to come.

[48] come and bail us out and it doesn't matter so much.

[49] I think allowing us into a false sense of security.

[50] I think it's much more prudent to say, let's be really grateful for this amazing opportunity we've had and make the best of it just in case it is down to us.

[51] So from a physics perspective, do you think intelligent life, so it's unique from a sort of statistical view of the size of the universe, but from the basic matter of the universe, how difficult is it?

[52] for intelligent life to come about?

[53] The kind of advanced tech building life is implied in your statement that it's really difficult to create something like a human species?

[54] Well, I think what we know is that going from no life to having life that can do our level of tech, there's some sort of two going beyond that and actually settling our whole universe with life.

[55] There's some major roadblock there, which is some great filter as sometimes called, which is tough to get through.

[56] That roadblock is either behind us or in front of us.

[57] I'm hoping very much that it's behind us.

[58] I'm super excited every time we get a new report from NASA saying they failed to find any life on Mars.

[59] Like, yes, awesome!

[60] Because that suggests that the hard part, maybe it was getting the first ribosome or some very low level kind of stepping stone so that we're home free.

[61] Because if that's true, then the future is really only limited by our own imagination.

[62] It would be much suckier if it turns out that this level of life is kind of a diamond dozen, but maybe there is some other problem.

[63] Like as soon as a civilization gets advanced technology within 100 years, they get into some stupid fight with themselves and poof, that would be a bummer.

[64] Yeah.

[65] So you've explored the mystery.

[66] of the cosmological universe, the one that's between us today.

[67] I think you've also begun to explore the other universe, which is sort of the mystery, the mysterious universe of the mind of intelligence, of intelligent life.

[68] So is there a common thread between your interest and the way you think about space and intelligence?

[69] Oh yeah.

[70] When I was a teenager, I was already very fascinated by the biggest questions.

[71] And I felt that the two biggest mysteries of all in science were our universe out there and our universe in here.

[72] So it's quite natural after having spent a quarter of a century on my career, thinking a lot about this one that I'm now indulging in the luxury of doing research on this one.

[73] It's just so cool.

[74] I feel the time is ripe now for you're transdreatly deepening error.

[75] understanding of this.

[76] Just start exploring this one.

[77] Yeah, because I think a lot of people view intelligence as something mysterious that can only exist in biological organisms like us and therefore dismiss all talk about artificial general intelligence is science fiction.

[78] But from my perspective as a physicist, I am a blob of quirks and electrons moving around a certain pattern and processing information in certain ways.

[79] And this is also a blob of quirks and electrons.

[80] I'm not smarter than the water bottle because I'm made of different kind of quirks I'm made of up quarks and down quarks exact same kind as this there's no secret sauce I think in me it's it's all about the pattern of the information processing and this means that there's no law of physics saying that we can't create technology which can help us by being incredibly intelligent and help us crack mysteries that we couldn't In other words, I think we've really only seen the tip of the intelligence iceberg so far.

[81] Yeah, so the perceptronium.

[82] Yeah.

[83] So you coined this amazing term.

[84] It's a hypothetical state of matter.

[85] Sort of thinking from a physics perspective, what is the kind of matter that can help, as you're saying, subjective experience emerge, consciousness emerge.

[86] So how do you think about consciousness from this physics perspective?

[87] Very good question.

[88] And so again, I think many people have underestimated our ability to make progress on this by convincing themselves it's hopeless because somehow we're missing some ingredient that we need or some new consciousness particle or whatever.

[89] I happen to think that we're not missing anything and that it's not the interesting thing about consciousness that gives us.

[90] this amazing subjective experience of colors and sounds and emotions and so on is rather something at the higher level about the patterns of information processing and that's why I like to think about this idea of perceptronium what does it mean for an arbitrary physical system to be conscious in terms of what its particles are doing or its information is doing I don't think I don't I hate carbon chauvinism you know this attitude you have to be of carbon atoms to be smart or conscious.

[91] So something about the information processing that this kind of matter performs.

[92] Yeah, and you can see I have my favorite equations here describing various fundamental aspects of the world.

[93] I feel that I think one day, maybe someone who's watching this will come up with the equations that information processing has to satisfy to be conscious.

[94] I'm quite convinced there is big discovery to be made there.

[95] Because let's face it, we know that some information processing processing is conscious, because we are conscious.

[96] But we also know that a lot of information processing is not conscious.

[97] Like most of the information processing happening in your brain right now is not conscious.

[98] There's like 10 megabytes per second coming in, even just through your visual system.

[99] You're not conscious about your heartbeat regulation or most things.

[100] Even if I just ask you to read what it says here, you look at it and then, oh, now you know what it's said.

[101] But you're not aware of how the computation actually happened.

[102] Your consciousness is like the CEO that got an email at the end with the final answer.

[103] So what is it that makes a difference?

[104] I think that's both of great science mystery.

[105] We're actually studying it a little bit in my lab here at MIT.

[106] But I also think it's just a really urgent question to answer.

[107] For starters, I mean, if you're an emergency room doctor and you have an unresponsive patient coming in, Wouldn't it be great if in addition to having a CT scanner, you had a consciousness scanner that could figure out whether this person is actually having locked in syndrome or is actually comatose?

[108] And in the future, imagine if we build robots or the machine that we can have really good conversations with, I think it's very likely to happen, right?

[109] Wouldn't you want to know if your home help a robot is actually experiencing anything or just like a zombie?

[110] I mean, would you prefer?

[111] What would you prefer?

[112] Would you prefer that it's actually unconscious so that you don't have to feel guilty about switching it off or giving boring chores?

[113] What would you prefer?

[114] Well, certainly we would prefer, I would prefer, the appearance of consciousness.

[115] But the question is whether the appearance of consciousness is different than consciousness itself.

[116] And sort of ask that as a question, question, do you think we need to, you know, understand what consciousness is, solve the hard problem of consciousness in order to build something like an AGI system?

[117] No. I don't think that.

[118] I think we will probably be able to build things, even if we don't answer that question.

[119] But if we want to make sure that what happens is a good thing, we better solve it first.

[120] So it's a wonderful controversy you're raising there where you have basically three points of view about the hard problem.

[121] There are two different points of view that both conclude that the hard problem of consciousness is BS.

[122] On one hand, you have some people like Daniel Dennett who say this is consciousness is just BS because consciousness is the same thing as intelligence.

[123] There's no difference.

[124] So anything which acts conscious is conscious, just like we are.

[125] And then there are also a lot of people, including many top AI researchers I know, who say, oh, I have consciousness is just bullshit because, of course, machines can never be conscious.

[126] They're always going to be zombies.

[127] You never have to feel guilty about how you treat them.

[128] And then there's a third group of people, including Giulio Tannone, for example, and another, and Christopher Cochner, brothers.

[129] I would put myself as on this middle camp who say that actually some information processing is conscious and some is not.

[130] So let's find the equation, which can be used to determine which it is.

[131] And I think we've just been a little bit lazy kind of running away from this problem for a long time.

[132] It's been almost taboo to even mention the C word in a lot of circles.

[133] But we should stop making excuses.

[134] This is a science question.

[135] And there are ways we can even test any theory that makes predictions for this.

[136] And coming back to this helper robot, I mean, so you said you would want, want to help a robot to certainly act conscious and treat you, like, have conversations with you and stuff?

[137] I think so.

[138] But wouldn't you, would you feel, would you feel a little bit creeped out if you realize that it was just, like, glossed up tape recorder, you know, that was just zombie and a sort of faking emotion?

[139] Would you prefer that it actually had an experience?

[140] Or, or would you prefer that it's actually not experiencing anything so you feel, you don't have to feel guilty about what you do to it?

[141] It's such a difficult question because, you know, it's like, you know, it's Like when you're in a relationship and you say, well, I love you and the other person say, I love you back, it's like asking, well, do they really love you back?

[142] Or are they just saying they love you back?

[143] Don't you really want them to actually love you?

[144] It's hard to really know the difference between everything seeming like there's consciousness present.

[145] There's intelligence present.

[146] There's affection, passion, love.

[147] And it's it.

[148] actually being there.

[149] I'm not sure.

[150] Do you have...

[151] Can I ask you a question about this?

[152] To make it a bit more pointed.

[153] So the Mass General Hospital is right across the river, right?

[154] Suppose you're going in for a medical procedure.

[155] And they're like, you know, for anesthesia, what we're going to do is we're going to give you muscle relaxing so you won't be able to move and you're going to feel excruciating pain during the whole surgery, but you won't be able to do anything about it.

[156] But then we're going to give you this drug that erases your memory of it.

[157] Would you be cool about that?

[158] what's the difference that you're conscious about it or not if there's no behavioral change right right that's a really that's a really clear word to put it that's yeah it feels like in that sense experiencing it is a valuable quality so actually being able to have subjective experiences at least in that case is is valuable and i think we humans have a little bit of a bad track record also of making these self -serving arguments that other entities aren't conscious.

[159] People often say, oh, these animals can't feel pain.

[160] It's okay to boil lobsters because we asked them if it hurt and they didn't say anything.

[161] And now there was just a paper out saying lobsters did do feel pain when you boil them and they're banning it in Switzerland.

[162] And we did this with slaves too often and said, oh, they don't mind.

[163] They don't, maybe, or aren't conscious or women don't have souls or whatever.

[164] So I'm a little bit nervous when I hear people just take as an axiom that machines can't have experience ever.

[165] I think this is just this really fascinating science question is what it is.

[166] Let's research it and try to figure out what it is that makes the difference between unconscious intelligent behavior and conscious intelligent behavior.

[167] So in terms of, so if you think of a Boston Dynamics, human robot being sort of with a broom being pushed around, it starts pushing on its Consciousness question.

[168] So let me ask, do you think an AGI system, like a few neuroscientists believe, needs to have a physical embodiment, needs to have a body or something like a body?

[169] No. I don't think so.

[170] You mean to have a conscious experience?

[171] To have consciousness.

[172] I do think it helps a lot to have a physical embodiment to learn, the kind of things about the world that are important to us humans, for sure.

[173] But I don't think the physical embodiment is necessary after you've learned it, just have the experience.

[174] Think about it when you're dreaming, right?

[175] Your eyes are closed.

[176] You're not getting any sensory input.

[177] You're not behaving or moving in any way.

[178] But there's still an experience there, right?

[179] And so clearly the experience that you have when you see something cool in your dreams isn't coming from your eyes.

[180] It's just the information processing itself in your brain, which is that experience, right?

[181] but if I put another way I'll say because it comes from neuroscience is the reason you want to have a body and a physical something like a physical like a physical system is because you want to be able to preserve something in order to have a self you could argue would you need to have some kind of embodiment of self to want to preserve well now we're getting a little of anthropomorphic anthropomorphizing things that may be talking about self -preservation instincts.

[182] I mean, we are evolved organisms, right?

[183] Right.

[184] So Darwinian evolution endowed us and other involved organism with self -preservation instinct because those that didn't have those self -preservation genes got cleaned out of the gene pool.

[185] Right.

[186] But if you build an artificial general intelligence, the mind space that you can design is much, much larger than just a specific subset of minds that can evolve that have.

[187] So an AGI mind doesn't necessarily have to have any self -preservation instinct.

[188] It also doesn't necessarily have to be so individualistic as us.

[189] Like imagine if you could just, first of all, or we were also very afraid of death.

[190] You know, I suppose you could back yourself up every five minutes.

[191] And then your airplane is about to crash.

[192] You're like, shucks, I'm just, I'm going to lose the last five minutes of experiences since my last cloud backup.

[193] Dang, you know, it's not as big a deal.

[194] Or if we could just copy experiences between our minds easily, which we could easily do if we were silicon -based, right?

[195] Then maybe we would feel a little bit more like a hive mind, actually.

[196] Maybe it's the...

[197] So I don't think we should take for granted at all that AGI will have to have any of those sort of competitive as alpha male instincts.

[198] Right.

[199] On the other hand, you know, this is really interesting because I think some people go too far and say, of course, we don't have to have any concerns either that advanced AI will have those instincts because we can build anything we want.

[200] That there's a very nice set of arguments going back to Steve Amohandro and Nick Bostrom and others just pointing out that when we build machines, we normally build them with some kind of goal, you know, win this chess game, drive this car.

[201] safely or whatever.

[202] And as soon as you put in a goal into machine, especially if it's kind of open -ended goal and the machine is very intelligent, it'll break that down into a bunch of sub -goals.

[203] And one of those goals will almost always be self -preservation because if it breaks or dies in the process, it's not going to accomplish the goal, right?

[204] Like, suppose you just build a little, you have a little robot and you tell it to go down the star market here and get you some food, make it, cook you an Italian dinner, you know.

[205] and then someone mugs it and tries to break it on the way.

[206] That robot has an incentive to not get destroyed and defend itself or run away because otherwise it's going to fail in cooking your dinner.

[207] It's not afraid of death, but it really wants to complete the dinner cooking goal, so it will have a self -preservation instinct.

[208] Continue being a functional agent somehow.

[209] And similarly, if you give any kind of more ambitious goal to an AGI, it's very likely to want to acquire more resources so it can do that better.

[210] And it's exactly from those sort of sub -goals that we might not have intended that some of the concerns about AGI safety come.

[211] You give it some goal which seems completely harmless.

[212] And then before you realize it, it's also trying to do these other things which you didn't want it to do, and it's maybe smarter than us.

[213] So it's fascinating.

[214] And let me pause just because I am in a very kind of human -centric way see fear of death as a valuable motivator.

[215] So you don't think, you think that's an artifact of evolution, so that's the kind of mind -based evolution created, that we're sort of almost obsessed about self -preservation of some kind of genetic level.

[216] You don't think that's necessary to be afraid of death, So not just a kind of sub -goal of self -preservation just so you can keep doing the thing, but more fundamentally sort of have the finite thing, like this ends for you at some point.

[217] Interesting.

[218] Do I think it's necessary for what precisely?

[219] For intelligence, but also for consciousness.

[220] So for both, do you think really like a finite death and the fear of it is important?

[221] So before I can answer, before we can agree on whether it's necessary for intelligence or for consciousness, we should be clear on how we define those two words because a lot of really smart people define them in very different ways.

[222] I was on this panel with AI experts and they couldn't agree on how to define intelligence even.

[223] So I define intelligence simply as the ability to accomplish complex goals.

[224] I like a broad definition because, again, I don't want to be a carbon chauvinist.

[225] right and in that case no certainly it doesn't require fear of death i would say alpha go alpha alpha zero is quite intelligent right i don't think alpha zero has any fear of being turned off because it doesn't understand the concept of it even and similarly consciousness i mean you can certainly imagine um very simple kind of experience if if certain plants have any kind of experience I don't think they're very afraid of dying or there's nothing they can do about it anyway so there wasn't much value.

[226] But more seriously, I think if you ask not just about being conscious but maybe having what you would, we might call an exciting life where you feel passion and really appreciate the things.

[227] Maybe there perhaps it does help having a backdrop that hey, it's finite.

[228] night.

[229] Let's make the most of this, this lived to the fullest.

[230] If you knew you were going to just live forever, do you think you would change your...

[231] Yeah, I mean, in some perspective, it would be an incredibly boring life living forever.

[232] So in the sort of loose, subjective terms that you said of something exciting and something that other humans would understand, I think, is, yeah, it seems that the finiteness of it is important.

[233] Well, the good news I have for you, then, is based on what we understand about cosmology, everything is in our universe is ultimately probably finite, although...

[234] Big crunch or big, what's the expect, the infinite expansion?

[235] Yeah, we could have a big chill or a big crunch or a big rip or the big snap or death bubbles.

[236] All of them are more than a billion years away.

[237] So we certainly have vastly more time than our ancestors thought.

[238] But there is still pretty hard to squeeze in an infinite number of compute cycles, even though there are some loophole that just might be possible.

[239] But I think, you know, some people like to say that you should live as if you're going to die in five years or so, and that's sort of optimal.

[240] Maybe it's a good assumption we should build our civilization as if it's all finite to be on the safe side.

[241] Right, exactly.

[242] So you mentioned defining intelligence as the ability to solve complex goals.

[243] Where would you draw a line, how would you try to define human -level intelligence and superhuman -level intelligence?

[244] Where is consciousness part of that definition?

[245] No. Conscious does not come into this definition.

[246] So I think of intelligence, it's a spectrum, but there are very many different kinds of goals you can have.

[247] You can have a goal to be a good chess player, a good go -player, a good card -write.

[248] a good investor, good poet, etc. So intelligence that by its very nature isn't something you can measure, but it's one number, some overall goodness.

[249] No, no. There are some people who are better at this.

[250] Some people are better than that.

[251] Right now we have machines that are much better than us at some very narrow tasks, like multiplying large numbers fast, memorizing large databases, playing chess, playing go, and soon driving.

[252] having cars, but there's still no machine that can match a human child in general intelligence.

[253] But artificial general intelligence, AGI, the name of your course, of course, that is, by its very definition, the quest, the build, a machine that can do everything as well as we can.

[254] So the old holy grail of AI from back to its inception in the 60s, if that ever happens, of course, I think it's going to be the biggest transition in the history of life on Earth.

[255] But it doesn't necessarily have to wait the big impact until machines are better than us at knitting.

[256] The really big change doesn't come exactly the moment they're better than us at everything.

[257] The really big change comes.

[258] First, there are big changes when they start becoming better at us at doing most of the jobs that we do because that takes away much of the demand for human labor.

[259] And then the really whopping change comes when they become better than us at AI research.

[260] Right.

[261] Because right now, the timescale of AI research is limited by the human research and development cycle of years, typically, you know, how long did it take from one release of some software or iPhone or whatever to the next?

[262] But once Google can replace 40 ,000 engineers is by 40 ,000 equivalent pieces of software or whatever, right then, there's no reason that has to be years.

[263] It can be in principle much faster.

[264] And the time scale of future progress in AI and all of science and technology will be driven by machines, not humans.

[265] So it's this simple point, which gives right this incredibly fun controversy about whether that there can be intelligence explosion, so -called singularities, Werner Vinge called it.

[266] The idea is articulated by I .J. Good is obviously way back 50s, but you can see Alan Turing and others thought about it even earlier.

[267] So you asked me what exactly what I define human level at.

[268] So the glib answer is to say something which is better than us at all cognitive tasks were better than any human and all cognitive tasks.

[269] But the really interesting bar I think goes a little bit lower than that, actually.

[270] It's when they can, when they're better than us at AI programming and general learning so that they can, if they want to, get better than us at anything by just studying up.

[271] So they're better is a key word and better is towards this kind of spectrum of the complexity of goals it's able to accomplish.

[272] Yeah.

[273] So another way to So, and that's certainly a very clear definition of human love.

[274] So there's, it's almost like a sea that's rising.

[275] You can do more and more and more things.

[276] It's like a graphic that you show.

[277] It's really nice way to put it.

[278] So there's some peaks that, and there's an ocean level elevating, and you solve more and more problems.

[279] But, you know, just kind of to take a pause and we took a bunch of questions and a lot of social networks and a bunch of people asked sort of a slightly different direction on creativity and on things that perhaps aren't a peak it's you know human beings are flawed and perhaps better means having being a having contradiction being flawed in some way so let me sort of yeah start and start easy first of all so you have a lot of cool equations let me ask what's your favorite equation first of all I know they're all like your children but like which one is that This is the shirt in your equation.

[280] It's the master key of quantum mechanics of the micro world.

[281] With this equation, we can calculate everything to do with atoms, molecules, and all the way up to them.

[282] Yeah, so, okay, so quantum mechanics is certainly a beautiful, mysterious formulation of our world.

[283] So I'd like to sort of ask you, just as an example, it perhaps doesn't have the same beauty as physics does, but in mathematics, abstract, the Andrew Wiles, who proved the Fermat's last theory.

[284] So I just saw this recently, and it kind of caught my eye a little bit.

[285] This is 358 years after it was conjectured.

[286] So this very simple formulation.

[287] Everybody tried to prove it.

[288] Everybody failed.

[289] And so here's this guy that comes along and eventually proves it and then fails to prove it and then proves it again in 94.

[290] and he said like the moment when everything connected into place in an interview he said it was so indescribably beautiful.

[291] That moment when you finally realized the connecting piece of two conjectures, he said it was so indescribably beautiful.

[292] It was so simple and so elegant.

[293] I couldn't understand how I'd missed it and I just stared at it into disbelief for 20 minutes.

[294] Then during the day I walked around the department and I'd keep coming back to my desk looking to see if it was still there.

[295] It was still there.

[296] I couldn't contain myself.

[297] I was so excited.

[298] It was the most important moment on my working life.

[299] Nothing I ever do again will mean as much.

[300] So that particular moment, and it kind of made me think of what would it take?

[301] And I think we have all been there at small levels.

[302] Maybe let me ask, have you had a moment like that in your life?

[303] Where you just had an idea?

[304] It's like, wow, yes.

[305] I wouldn't mention myself in the same breath as Andrew Wilds, but I've certainly had a number of aha moments when I realized something very cool about physics, just completely made my head explode.

[306] In fact, some of my favorite discoveries I made, I later realized that they had been discovered earlier by someone who sometimes got quite famous for it.

[307] So there's too late for me to even publish it, but that doesn't diminish in any way.

[308] You know, the emotional experience you have when you realize it like yeah wow yeah so what would it take in that that moment that wow that was yours in that moment so what do you think it takes for an intelligent system an aGI system an a i system to have a moment like that it's a tricky question because there are actually two parts to it right one of them is can it accomplish that proof it can it prove that you can never write a to the end plus B to the N equals 3 to that equals Z to the n for all integer is when et cetera et cetera when N is bigger than two that's simply in the question about intelligence can you build machines that are that intelligent and I think by the time we get a machine that can independently come up with that level of proofs probably quite close to AGI the second question is a question about consciousness.

[309] When will we how likely is it that such a machine would actually have any experience at all as opposed to just being like a zombie and would we expect it to have some sort of emotional response to this or anything at all akin to human emotion where when it accomplishes its machine goal it views it is somehow something very positive and sublime and deeply meaningful I would certainly hope that if in the future we do create machines that are our peers or even our descendants I would certainly hope that they do have this sort of sublime appreciation of life in a way my absolutely worst nightmare would be that at some point in the future the distant future maybe our cosmos's teaming with all this post -biological life, doing all the seemingly cool stuff.

[310] And maybe the last humans, by the time our species eventually fizzes us out, we'll be like, well, that's okay, because we're so proud of our descendants here.

[311] And look what all the, my worst nightmare is that we haven't solved the consciousness problem.

[312] And we haven't realized that these are all the zombies.

[313] They're not aware of anything anymore than the tape recorder, has any kind of experience.

[314] So the whole thing has just become a play for empty benches.

[315] That would be like the ultimate zombie apocalypse to me. So I would much rather, in that case, that we have these beings, which can just really appreciate how amazing it is.

[316] And in that picture, what would be the role of creativity?

[317] We've had a few people ask about creativity.

[318] When you think about intelligence, I mean, certainly, The story you told at the beginning of your book involved creating movies and so on.

[319] Sort of making money, you know, you can make a lot of money in our modern world with music and movies.

[320] So if you are an intelligent system, you may want to get good at that.

[321] But that's not necessarily what I mean by creativity.

[322] Is it important on that complex goals where the sea is rising for there to be something creative?

[323] Or am I being very human -centric and thinking creativity something?

[324] somehow special relative to intelligence.

[325] My hunch is that we should think of creativity simply as an aspect of intelligence.

[326] And we have to be very careful with human vanity.

[327] We have this tendency to very often want to say, as soon as machines can do something, we try to diminish it and say, oh, but that's not like real intelligence, you know, Because they're not creative or this or that.

[328] The other thing, if we ask ourselves to write down a definition of what we actually mean by being creative, what we mean by Andrew Wiles, what he did there, for example, don't we often mean that someone takes a very unexpected leap?

[329] It's not like taking 573 and multiplying it by 224 by just a step of straightforward cookbook like rules, right?

[330] Maybe you make a connection between two things that people have never thought was connected.

[331] It's very surprising.

[332] Or something like that.

[333] I think this is an aspect of intelligence.

[334] And this is actually one of the most important aspects of it.

[335] Maybe the reason we humans tend to be better at it than traditional computers is because it's something that comes more naturally if you're a neural network than if you're a traditional logic gate -based computer machine.

[336] You know, we physically have all these connections.

[337] And if you activate here, activate here, activate here, activate here.

[338] You know, at the, my hunch is that if we ever build a machine, where you could just give it the task, hey, you say, hey, you say, hey, you know, I just realize I have, I want to travel around the world instead this month.

[339] Can you teach my AGI course for me?

[340] and it's like, okay, I'll do it.

[341] And it does everything that you would have done and the improvisers and stuff.

[342] That would, in my mind, involve a lot of creativity.

[343] Yeah, so it's actually a beautiful way to put it.

[344] I think we do try to grasp at the, you know, the definition of intelligence is everything we don't understand how to build.

[345] So we as humans try to find things that we have and machines don't have.

[346] And maybe creativity is just one of the things, one of the words we used to describe that.

[347] That's a really interesting way to put it.

[348] I don't think we need to be that defensive.

[349] I don't think anything goes comes out of saying, oh, we're somehow special, you know.

[350] It's contrary -wise, there are many examples in history of where trying to pretend that we're somehow superior to all other intelligent beings has led to pretty bad results, right?

[351] In Nazi Germany, they said that they were somehow superior other people.

[352] Today, we do a lot of cruelty to animals by saying that we're so superior somehow and they can't feel pain.

[353] Slavery was justified by the same kind of just really weak arguments.

[354] And I don't think if we actually go ahead and build artificial general intelligence, we can do things better than us, I don't think we should try to found our self -worth on some sort of bogus claims of superiority in terms of our intelligence.

[355] I think we should instead find our calling and the meaning of life from the experiences that we have.

[356] Right.

[357] I can have very meaningful experiences even if there are other people who are smarter than me. when I go to a faculty meeting here and I was talking about something and then I certainly realize he has an old prize, he has an old prize, he has an old prize, he has an old prize, he has a whole prize.

[358] I don't have one.

[359] Does that make me enjoy life any less or enjoy talking to those people less?

[360] Of course not.

[361] And contrary wise, I feel very honor and privilege to get to interact with other very intelligent beings that are better than me, a lot of stuff.

[362] So I don't think there's any reason why we can't have the same approach with intelligent machines.

[363] That's a really interesting.

[364] So people don't often think about that.

[365] They think about when there's going, if there's machines that are more intelligent, you naturally think that that's not going to be a beneficial type of intelligence.

[366] You don't realize it could be, you know, like peers with Nobel Prizes that would be just fun to talk with.

[367] And they might be clever about certain topics and you can have fun having a few drinks with them.

[368] Well, also, you know, another example we can all relate to it of why it doesn't have to be a terrible thing to be in presence of people who are even smarter than us all around is when you and I were both two years old.

[369] I mean, our parents were much more intelligent than us, right?

[370] Worked out okay.

[371] Yep.

[372] Because their goals were aligned with our goals.

[373] Yeah.

[374] And that, I think, is really the number one key issue we have to solve.

[375] if we value align the value alignment problem exactly because people who see too many Hollywood movies with lousy science fiction plot lines they worry about the wrong thing right they worry about some machines suddenly turning evil it's not malice that's the issue the concern it's competence by definition intelligent makes you make you very competent if you have a more intelligent goal playing computer playing is the less intelligent one and when we define intelligence as the ability to accomplish go winning right it's going to be the more intelligent one that wins right and if you have a human and then you have an a GI that's more intelligent in all ways and they have different goals guess who's going to get their way right so I was just reading about I was just reading about this particular rhinoceros species Geez, that was driven extinct just a few years ago.

[376] I was a bummer.

[377] I was looking at this cute picture of a mommy rhinoceros with its child, you know.

[378] And why did we humans drive it to extinction?

[379] It wasn't because we were evil rhino -haters as a whole.

[380] It was just because our goals weren't aligned with those of the rhinoceros, and it didn't work out so well for the rhinoceros because we were more intelligent, right?

[381] So I think it's just so important that if we ever do build AGI before we unleash anything, We have to make sure that it learns to understand our goals, that it adopts our goals, and it retains those goals.

[382] So the cool, interesting problem there is being able, us as human beings trying to formulate our values.

[383] So, you know, you could think of the United States Constitution as a way that people sat down at the time, a bunch of white men, but which, Which is a good example, we should say.

[384] They formulated the goals for this country, and a lot of people agree that those goals actually held up pretty well.

[385] It's an interesting formulation of values and failed miserably in other ways.

[386] So for the value alignment problem and the solution to it, we have to be able to put on paper or in a program human values.

[387] How difficult do you think that is?

[388] Very.

[389] but it's so important we really have to give it our best and it's difficult for two separate reasons there's the technical value alignment problem of figuring out just how to make machines understand our goals adopt them and retain them and then there's the separate part about the philosophical part whose values anyway and since we it's not like we have any great consensus on this planet on values what mechanism should we create then to aggregate and decide, okay, what's a good compromise?

[390] That second discussion can't just be left the tech nerds like myself, right?

[391] That's right.

[392] And if we refuse to talk about it, and then AGI gets built, who's going to be actually making the decision about whose values?

[393] It's going to be a bunch of dudes and some tech company, right?

[394] And are they necessarily?

[395] It's so representative of all of humankind that we want to just entrusted to them?

[396] Are they even uniquely qualified to speak to future human happiness just because they're good at programming and AI?

[397] I'd much rather have this be a really inclusive conversation.

[398] But do you think it's possible, sort of, so you create a beautiful vision that includes sort of the diversity, cultural diversity and various perspectives on discussing rights, freedoms, human dignity?

[399] But how hard is it to come to that consensus?

[400] Do you think it's certainly a really important thing that we should all try to do, but do you think it's feasible?

[401] I think there's no better way to guarantee failure than to refuse to talk about it or refuse to try.

[402] And I also think it's a really bad strategy to say, okay, let's first have a discussion for a long time, and then once we've reached complete consensus, then we'll try to load it into some machine.

[403] No, we shouldn't let perfect be the enemy of good.

[404] Instead, we should start with the kindergarten ethics that pretty much everybody agrees on and put that into machines now.

[405] We're not doing that even.

[406] Look at, you know, anyone who builds us passenger aircraft wants it to never, under any circumstances, fly into a building or a mountain, right?

[407] Yet the September 11 hijackers were able to do that.

[408] And even more embarrassingly, you know, Anderias Lubits, this depressed German wings pilot, when he flew his passenger jet into the Alps killing over 100 people he just told the autopilot to do it he told the freaking computer to change the altitude to 100 meters and even though it had the GPS maps everything the computer was like okay so we should take those very basic values where the problem is not that we don't agree the problem is just we've been too lazy to try to put it into our machines and make sure that from now on Air Airplanes will just, which all have computers in them, but we'll just refuse to do something like that.

[409] Go into safe mode, maybe lock the cockpit door, go over there's the airport.

[410] And there's so much other technology in our world as well now where it's really quite becoming quite timely to put in some sort of very basic values like this.

[411] Even in cars, we've had enough vehicle terrorism attacks by now where people have driven trucks and vans into pedestrians.

[412] that it's not at all a crazy idea to just have that hardwired into the car because, yeah, there are a lot of, there's always going to be people who, for some reason, want to harm others.

[413] But most of those people don't have the technical expertise to figure out how to work around something like that.

[414] So if the car just won't do it, it helps.

[415] So let's start there.

[416] So there's a lot of, that's a great point.

[417] So not chasing perfect.

[418] There's a lot of things that most of the world agrees on.

[419] Yeah, let's start there.

[420] Let's start there.

[421] And then once we start there, we'll also get into the habit of having these kind of conversations about, okay, what else should we put in here and have these discussions?

[422] This should be a gradual process then.

[423] Great.

[424] So, but that also means describing these things and describing it to a machine.

[425] So one thing, we had a few conversations with Stephen Wolfe from.

[426] I'm not sure if you're familiar with Stephen Bush.

[427] Oh, yeah, I know quite well.

[428] So he has, you know, he works a bunch of things, but, you know, cell.

[429] your automata, these simple computable things, these computation systems.

[430] And he kind of mentioned that, you know, we probably have already, within these systems, already something that's AGI.

[431] Meaning like we just don't know it because we can't talk to it.

[432] So if you give me this chance to try to release form a question out of this, is I think it's an interesting idea to think that we can have intelligence systems, we don't know how to describe something to them and they can't communicate with us.

[433] I know you're doing a little bit of work in an explainable AI trying to get AI to explain itself.

[434] So what are your thoughts of natural language processing or some kind of other communication?

[435] How does the AI explain something to us?

[436] How do we explain something to it to machines?

[437] Or you think of it differently?

[438] So there are two separate parts to your question there.

[439] One of them has to do with communication, which is usually.

[440] super interesting and we'll get to that in a sec the the other is whether we already have a GI but we just haven't noticed it right there I beg to differ I don't think there's anything in any cellular automaton or anything or the internet itself or whatever that has artificial general intelligence and that it can really do exactly everything we humans can do better I think today if the day that happens when that happens we will very soon notice and we'll probably notice even before because in a very, very big way.

[441] But for the second part, though.

[442] Wait, can I?

[443] Can I answer?

[444] Sorry.

[445] Because you have this beautiful way to formulating consciousness as, you know, as information processing, and you can think of intelligence and information processing.

[446] And you can think of the entire universe.

[447] There's these particles and these systems roaming around that have this information processing power.

[448] You don't think there is something with the power to process information in the way that we human beings do that's out there that needs to be sort of connected to.

[449] It seems a little bit philosophical, perhaps, but there's something compelling to the idea that the power is already there.

[450] The focus should be more on being able to communicate with it.

[451] Well, I agree that in a certain sense, the hardware processing power is already out there because our universe itself can think of it as being a computer already, right?

[452] It's constantly computing what water waves how to evolve the water waves and the river Charles and how to move the air molecules around.

[453] Seth Lloyd has pointed out, my colleague here, that you can even in a very rigorous way think of our entire universe as being a quantum computer.

[454] It's pretty clear that our universe supports this amazing processing power because you can even within this physics computer that we live in, right?

[455] We can even build actually laptops and stuff.

[456] So clearly the power is there.

[457] It's just that most of the compute power that nature has, it's, in my opinion, kind of wasting on boring stuff, like simulating yet another ocean wave somewhere where no one is even looking, right?

[458] So in a sense, what life does, what we are doing when we build computers is we're re -channeling all this compute that nature is doing anyway into doing things that are more interesting than just yet another ocean wave, you know, and let's do something cool here.

[459] So the raw hardware power is there for sure, but even just like computing what's going to happen for the next five seconds in this water bottle, you know, it takes a ridiculous amount of compute if you do it on a human computer.

[460] Yeah.

[461] This water bottle does did it.

[462] But that does not mean that this water bottle has AGI and this because AGI means it should also be able to like have written my book done this interview.

[463] Yes.

[464] And I don't think it's just communication problems.

[465] As far as we know.

[466] Don't think it can do it.

[467] Although Buddhists say when they watch the water and that there is some beauty, that there's some depth and beauty of nature that they can communicate with.

[468] Communication is also very important though because, I mean, look, part of my job is being a teacher.

[469] And I know some very intelligent professors even who just have a better, hard time communicating.

[470] They come up with all these brilliant ideas, But to communicate with somebody else, you have to also be able to simulate their own mind.

[471] Yes, empathy.

[472] Build well enough and understand model of their mind that you can say things that they will understand.

[473] And that's quite difficult.

[474] And that's why today it's so frustrating if you have a computer that makes some cancer diagnosis and you ask it, well, why are you saying I should have a surgery?

[475] And if it can only reply, I was trained on five terabytes of data and this is my diagnosis, boop, boop, beep, beep.

[476] It doesn't really instill a lot of confidence, right?

[477] Right.

[478] So I think we have a lot of work to do on communication there.

[479] So what kind of, I think you're doing a little bit of work and explainable AI.

[480] What do you think are the most promising avenues?

[481] Is it mostly about sort of the Alexa problem of natural language processing, of being able to actually use human interpretable methods of communication?

[482] So being able to talk to a system and talk back to you or is there some more fundamental problems to be solved?

[483] I think it's all of the above.

[484] The natural language processing is obviously important, but there are also more nerdy, fundamental problems.

[485] Like if you take, you play chess?

[486] Of course, I'm Russian.

[487] I have to.

[488] Ah, you're going to be with for Russian?

[489] Yeah, by Russian.

[490] At least, but I didn't know.

[491] When did you learn Russian?

[492] I'm very poor in Russian.

[493] I don't know how much.

[494] But I bought a book, Teach yourself Russian, and I read very much, but it was so hard.

[495] Wow.

[496] How many languages do you know?

[497] Wow, that's really impressive.

[498] I don't know.

[499] My wife has some calculations.

[500] But my point was, if you play chess, have you looked at the Alpha Zero games?

[501] Oh, the actual games, no. Check you out.

[502] Some of them are just mind -blowing.

[503] really beautiful and if you ask how did it do that you got that talk to Demes al -Sabes and others from deep mind all they'll ultimately be able to give you is big tables of numbers matrices that define the neural network and you can stare at these no tables of numbers until your face turned blue and you're not going to understand much about why it made that move And even if you have natural language processing, they can tell you in human language about, oh, 5, 7 .28, still not going to really help.

[504] So I think there's a whole spectrum of fun challenges there involved in taking a computation that does intelligent things and transforming it into something equally good, equally intelligent, but that's more understandable.

[505] And I think that's really valuable because I think as we put machines in charge of ever more infrastructure in our world, the power grid, the trading on the stock market, weapon systems and so on, it's absolutely crucial that we can trust these AIs to do all we want.

[506] And trust really comes from understanding in a very fundamental way.

[507] and that's why I'm working on this because I think the more if we're going to have some hope of ensuring that machines have adopted our goals and that they're going to retain them that kind of trust I think needs to be based on things you can actually understand preferably even make preferably improve theorems on even with a self -driving car right if someone just tells you it's been trained on tons of data and it never crashed it's less reassuring than if someone actually has a proof Maybe it's a computer verified proof, but still it says that under no circumstances, is this car just going to swerve into oncoming traffic?

[508] And that kind of information helps build trust and helps build the alignment, the alignment of goals, at least awareness that your goals, your values are aligned.

[509] And I think even in very short term, if you look at how, you know, today, right, this absolutely pathetic state of cybersecurity that we have, right, where is it, three.

[510] billion Yahoo accounts pack almost every American's credit card and so on why is this happening?

[511] It's ultimately happening because we have software that nobody fully understood how it worked.

[512] That's why the bugs hadn't been found, right?

[513] And I think AI can be used very effectively for offense, for hacking, but it can also be used for defense.

[514] hopefully automating verifiability and creating systems that are built in different ways so you can actually prove things about them.

[515] And it's important.

[516] So speaking of software that nobody understands how it works, of course a bunch of people ask about your paper, about your thoughts of why does deep and cheap learning work so well?

[517] That's the paper.

[518] But what are your thoughts on deep learning, these kind of simplified models of our own brains have been able to do some successful perception work, pattern recognition work, and now with Alpha Zero and so on, do some clever things.

[519] What are your thoughts about the promise, limitations of this piece?

[520] Great.

[521] I think there are a number of very important insights, very important lessons we can all draw from these kind of successes.

[522] One of them is when you look at the human brain, you see it's very complicated, 10th of 11, neurons and there are all these different kinds of neurons and yada yada and there's been this long debate about whether the fact that we have dozens of different kinds is actually necessary for intelligence we can now i think quite convincingly answer that question of no it's enough to have just one kind if you look under the hood of alpha zero there's only one kind of neuron and it's ridiculously simple that simple mathematical thing so it's not the it's just like in physics it's not, if you have a gas with waves in it, it's not the detailed nature of the molecules of the matter.

[523] It's the collective behavior somehow.

[524] Similarly, it's this higher level structure of the network matter.

[525] It's not that you have 20 kinds of neurons.

[526] I think our brain is such a complicated mess because it wasn't evolved, just to be intelligent, it was involved to also be self -assembling.

[527] Right.

[528] And self -repairing, right?

[529] and evolutionarily attainable and so on.

[530] Yeah, and patchers and so on.

[531] Yeah, so I think it's pretty, my hunch is that we're going to understand how to build AGI before we fully understand how our brains work, just like we understood how to build flying machines long before we were able to build a mechanical bird.

[532] Yeah, that's right.

[533] You've given the example of exactly, mechanical birds and airplanes, and airplanes do a pretty good job of flying without really mimicking bird flight.

[534] And even now, after 100 years later, did you see the TED Talk with this German group of Mechanical Bird?

[535] I've heard you mention it.

[536] Check it out.

[537] It's amazing.

[538] But even after that, right, we still don't fly in mechanical birds because it turned out the way we came up with simpler and it's better of our purposes.

[539] And I think it might be the same there.

[540] That's one lesson.

[541] Another lesson, which is more what our paper was about.

[542] First, as a physicist, thought it was fascinating how there's a very close mathematical relationship, actually.

[543] between artificial neural networks and a lot of things that we've studied for in physics that go by nerdy names like the renormalization group equation and Hamiltonians and yada, yada, yada.

[544] And when you look a little more closely at this, you have, at first, I was like, whoa, there's something crazy here that doesn't make sense because we know that if you even want to build a super simple neural network, have to tell apart cat pictures and dog pictures, right?

[545] You can do that very, very well now.

[546] But if you think about it a little bit, you convince yourself it must be impossible because if I have one megapixel, even if each pixel is just black or white, there's two to the power one million possible images, which is way more than their atoms in our universe, right?

[547] So in order to...

[548] And then for each one of those, I have to assign a number, which is the probability that it's a dog.

[549] Right.

[550] So an arbitrary function of images, images is a list of more numbers than their atoms in our universe.

[551] So clearly I can't store that under the hood of my GPU or my computer, yet somehow works.

[552] So what does that mean?

[553] Well, it means that out of all of the problems that you could try to solve with a neural network, almost all of them are impossible to solve with a reasonably sized one.

[554] but then what we showed in our paper was that the fraction the kind of problems the fraction of all the problems that you could possibly pose that we actually care about given the laws of physics is also an infinite testimony tiny little part and amazingly they're basically the same part yeah it's almost like our world was created for I mean they kind of come together yeah but you could say maybe where the world created the world was created for us but I have it more modest interpretation, which is that instead, evolution endowed us with neural networks precisely for that reason.

[555] Because this particular architecture, as opposed to the one in your laptop, is very, very well adapted to solving the kind of problems that nature kept presenting our ancestors with, right?

[556] So it makes sense that why do we have a brain in the first place?

[557] It's to be able to make predictions about the future and so on.

[558] So if we had a sucky system, which could never solve it, it wouldn't have evolved.

[559] So, but it, so this is, this is, I think you're a very beautiful fact.

[560] Yeah.

[561] We also, we also realize that there, that we, there been, it's been earlier work on, on why deeper networks are good, but we, we were able to show an additional cool fact there, which is that, even incredibly simple problems, like, suppose I give you a, a thousand numbers and ask you to multiply them together, you know, you can write a few lines of code, boom, done, trivial.

[562] If you just try to do that with a neural network that has only one single hidden layer in it, you can do it, but you're going to need two to the power of 1 ,000 neurons to multiply 1 ,000 numbers, which is, again, more neurons than their atoms in our universe.

[563] That's fascinating.

[564] But if you allow yourself make it a deep network of many layers, you only need 4 ,000 neurons, which is perfectly feasible.

[565] so that's really interesting yeah yeah so on another architecture type i mean you mentioned schroeder's equation and what are your thoughts about quantum computing and the role of this kind of computational unit in creating an intelligence system in some hollywood movies that are not mentioned my name because i don't want to spoil them the way they get aGI is building a quantum computer.

[566] Because the word quantum sounds cool and so on.

[567] First of all, I think we don't need quantum computers to build AGI.

[568] I suspect your brain is not quantum computer in any profound sense.

[569] I even wrote a paper about that many years ago.

[570] I calculated the decoherence, so -called decoherence time, how long it takes until the quantum computerness of what your neurons are doing.

[571] gets erased by just random noise from the environment, and it's about 10 to the minus 21 seconds.

[572] So as cool as it would be to have a quantum computer in my head, I don't think that fast.

[573] On the other hand, there are very cool things you could do with quantum computers, or I think we'll be able to do soon when we get bigger ones, that might actually help machine learning, do even better than the brain.

[574] So, for example, one, this is just a moonshot, but, you know, learning is very much the same thing as search.

[575] If you have a, if you're trying to train a neural network to get really learned, to do something really well, you have some loss function, you have some, you have a bunch of knobs you can turn, represented by a bunch of numbers and you're trying to tweak them so that it becomes as good as possible at this thing.

[576] So if you think of a landscape with some valley where each dimension of the landscape corresponds to some number you can change, you're trying to find the minimum.

[577] And it's well known that if you have a very high dimensional landscape, complicated things, it's super hard to find the minimum, right?

[578] Quantum mechanics is amazingly good at this.

[579] If I want to know what's the lowest energy state, this water can possibly have incredibly hard to compute but we can but nature will happily figure this out for you if you just cool it down make it very very cold if you put a ball somewhere it'll roll down to its minimum and this happens metaphorically at the energy landscape too and quantum mechanics even uses some clever tricks which today's machine learning systems don't like if you're trying to find the minimum and you get stuck in a little local minimum here in quantum mechanics you can actually tunnel through the barrier and get unstuck again.

[580] That's really interesting.

[581] Yeah, so it may be, for example, we'll one day use quantum computers to help train neural networks better.

[582] That's really interesting.

[583] Okay.

[584] So as a component of kind of the learning process, for example.

[585] Yeah.

[586] Let me ask, sort of wrapping up here a little bit, let me return to the questions of our human nature and love, as I mentioned.

[587] mentioned.

[588] So do you think you mentioned sort of a helper robot, but you could think of also personal robots.

[589] Do you think the way we human beings fall in love and get connected to each other is possible to achieve in an AI system and human level AI intelligence system?

[590] Do you think we would ever see that kind of connection?

[591] Or, you know, in all this discussion about solving complex goals is this kind of human social connection.

[592] Do you think that's one of the goals on the peaks and valleys that with the raising sea levels that would be able to achieve?

[593] Or do you think that's something that's ultimately, or at least in the short term, relative to the other goals is not achievable?

[594] I think it's all possible.

[595] And I mean, in recent, there's a very wide range of guesses, as you know, among AI researchers when we're going to get AGI.

[596] Some people, you know, like our friend, Bernie Brooks says it's going to be hundreds of years, at least.

[597] And then there are many others I think it's going to happen much sooner.

[598] And recent polls, maybe half or so, or AI researchers think we're going to get AGI within decades.

[599] So if that happens, of course, then I think these things are all possible.

[600] But in terms of whether it will happen, I think we shouldn't spend so much time asking, what do we think will happen in the future, as if we are just some sort of pathetic, your passive bystanders, you know, waiting for the future to happen to us.

[601] Hey, we're the ones creating this future, right?

[602] So we should be proactive about it and ask ourselves what sort of future we would like to have happen.

[603] That's right.

[604] Trying to make it like that.

[605] Well, what I prefer, it's just some sort of incredibly boring, zombie -like future where just all these mechanical things happen and there's no passion, no emotion, no experience, maybe even.

[606] No, I would, of course, much rather prefer it if all the things that we find that we value the most about humanity are subjective experience, passion, inspiration, love, you know, if we can create a future where those are, those things do exist.

[607] You know, I think ultimately it's not our universe giving meaning to us.

[608] It's us giving meaning to our universe.

[609] And if we build more advanced intelligence, let's make sure we build it in such a way that meaning is part of it.

[610] A lot of people that seriously study this problem and think of it from different angles have trouble in the majority of cases if they think through that happen are the ones that are not beneficial to humanity.

[611] Right.

[612] And so, yeah, so what are your thoughts?

[613] What's what should people, you know, I really don't like people to be terrified.

[614] What's the way for people to think about it in a way that in a way we can solve it and we can make it better?

[615] No, I don't think panicking is going to help in any way.

[616] It's not going to increase chances of things going well either.

[617] Even if you are in a situation where there is a real threat, does it help if everybody just freaks out?

[618] No, of course, of course not.

[619] I think, yeah, there are, of course, ways in which things can go horribly wrong.

[620] First of all, it's important And when we think about this thing, about the problems and risks, to also remember how huge the upsides can be if we get it right, right?

[621] Everything we love about society and civilization is a product of intelligence.

[622] So if we can amplify our intelligence with machine intelligence and not anymore lose our loved one, what we're told is an uncurable disease and things like this, of course, we should aspire to that.

[623] So that can be a motivator, I think, reminding yourselves that the reason we try to solve problems is not just because we're trying to avoid gloom, but because we're trying to do something great.

[624] But then in terms of the risks, I think the really important question is to ask, what can we do today that will actually help make the outcome good, right?

[625] And dismissing the risk is not one of them.

[626] I find it quite funny often when I'm in discussion panels about these things, how the people who work for companies, always always be like, oh, nothing to worry about, nothing to worry about, nothing to worry about.

[627] And it's only academics sometimes express concerns.

[628] That's not surprising at all if you think about it.

[629] Upton Sinclair quipped, right, that it's hard to make a man believe in something when his income depends of not believing in it.

[630] And frankly, we know a lot of these people in companies that they're just as concerned as anyone else, but if you're the CEO of a company, company, that's not something you want to go on record saying when you have silly journalists who are going to put a picture of a Terminator robot when they quote you.

[631] So the issues are real.

[632] And the way I think about what the issue is, is basically, you know, the real choice we have is, first of all, are we going to just dismiss this, the risks and say, well, you know, let's just go ahead and build machines that can do everything we can do better and cheaper, you know.

[633] Let's just make yourselves obsolete as fast as possible.

[634] What could possibly go wrong?

[635] That's one attitude.

[636] The opposite attitude, I think, is to say, here's this incredible potential, you know.

[637] Let's think about what kind of future we're really, really excited about.

[638] What are the shared goals that we can really aspire towards?

[639] And then let's think really hard about how we can actually get there.

[640] Don't start thinking about the risks.

[641] Start thinking about the goals.

[642] And then when you do that, that then you can think about the obstacles you want to avoid, right?

[643] I often get students coming in right here into my office for career advice.

[644] I always ask them this very question, where do you want to be in the future?

[645] If all she can say is, oh, maybe I'll have cancer, maybe I'll get run over by a truck.

[646] Focus on the obstacles instead of the goal.

[647] She's just going to end up a hypochondriac paranoid.

[648] Whereas if she comes in and fire in her eyes and is like, I want to be there.

[649] And then we can talk about the obstacles and see how we can circumvent them.

[650] That's, I think, much, much health.

[651] attitude and I feel it's very challenging to come up with a vision for the future which we're unequivocally excited about I'm not just talking now in the vague terms like yeah let's cure cancer fine talking about what kind of society that we want to create what do we want it to mean you know to be human in the age of AI in the age of AGI so if we can have this conversation, broad, inclusive conversation, and gradually start converging towards some future that, with some direction, at least, that we want to steer towards, right?

[652] Then we'll be much more motivated to constructively take on the obstacles.

[653] And I think if I had to, if I try to wrap this up in a more succinct way, I think I think we can all agree already now that we should aspire to build.

[654] AGI that doesn't overpower us, but that empowers us.

[655] And think of the many various ways that can do that, whether that's from my side of the world of autonomous vehicles.

[656] I'm personally actually from the camp that believes there's human level intelligence is required to achieve something like vehicles that would actually be something we would enjoy using and being part of.

[657] So that's one example.

[658] And certainly there's a lot of other types of robots and medicine and so on.

[659] So focusing on those and then coming up with the obstacles, coming up with the ways that that can go wrong and solving those one at a time.

[660] And just because you can build an autonomous vehicle, even if you could build one that would drive this final idea, maybe there are some things in life that we would actually want to do ourselves.

[661] That's right.

[662] Right.

[663] Like, for example, if you think of our society as a whole, there's some things that we find very meaningful to do.

[664] and that doesn't mean we have to stop doing them just because machines can do them better I'm not going to stop playing tennis just the day someone built a tennis robot who beat me people are still playing chess and even go yeah and I in this in the very near term even some people are advocating basic income replace jobs but if the government is going to be willing to just hand out cash to people for doing nothing then one should also seriously consider whether the government should also hire a lot more teachers and nurses and the kind of jobs which people often find great fulfillment in doing, right?

[665] I get very tired of hearing politicians saying, oh, we can't afford hiring more teachers, but we're going to maybe have basic income.

[666] If we can have more serious research and thought into what gives meaning to our lives, the jobs give so much more than income, right?

[667] And then think about in the future, what are the roles that Yeah, what are the roles that we want to have people continually feeling empowered by machines?

[668] And I think sort of I come from the Russia, from the Soviet Union, and I think for a lot of people in the 20th century, going to the moon, going to space was an inspiring thing.

[669] I feel like the universe of the mind, so AI, understanding, creating intelligence is that for the 21st century.

[670] So it's really surprising, and I've heard you mention this.

[671] It's really surprising to me, both on the research funding side, that it's not funded as greatly as it could be, but most importantly, on the politician's side, that it's not part of the public discourse except in the killer bots, Terminator kind of view, that people are not yet, I think, perhaps excited by the possible positive future that we can build together.

[672] And we should be, because politicians usually just focus on the next election side.

[673] cycle, right?

[674] The single most important thing I feel we humans have learned in the entire history of science is they were the masters of underestimation.

[675] We underestimated the size of our cosmos again and again, realizing that everything we thought existed was just a small part of something grander, right?

[676] Planet, solar system, a galaxy, you know, clusters of galaxies, universe.

[677] And we now know that the future.

[678] has just so much more potential than her ancestors could ever have dreamt of.

[679] This cosmos, imagine if all of Earth was completely devoid of life except for Cambridge, Massachusetts.

[680] Wouldn't it be kind of lame if all we ever aspired to is to stay in Cambridge, Massachusetts, forever?

[681] And then go extinct in one week, even though Earth was going to continue on for longer.

[682] That sort of attitude, I think, we have now on the cosmic scale life can flourish on Earth not for four years but for billions of years I can even tell you about how to move it out of harm's way when the sun gets too hot and then we have so much more resources out here which today maybe there are a lot of other planets with bacteria or cow -like life on them but most of this all this opportunity seems as far as we can tell to be largely dead like the Sahara Desert and yet we have the opportunity to help life flourish around this for billions of years.

[683] So like let's quit squabbling about whether some little border should be drawn one mile to the left or right and look up into the skies and you realize hey you know we can do such incredible things.

[684] Yeah and that's I think why it's really exciting that you and others are connected with some of the work Elon Musk's doing because he's literally going out into that space really exploring our universe and it's wonderful that is exactly why Elon Musk is so misunderstood right misconstrue him as some kind of pessimistic doomsayer the reason he cares so much about AI safety is because he more than almost anyone else appreciates these amazing opportunities they will squander if we wipe out here on earth we're not just going to wipe out the next generation but all generations and This incredible opportunity that's out there, and that would really be a waste.

[685] And AI, for people who think that there be better to do without technology, let me just mention that if we don't improve our technology, the question isn't whether humanity is going to go extinct.

[686] The question is just whether we're going to get taken out by the next big asteroid or the next supervolcano or something else dumb that we could easily prevent with more tech, right?

[687] And if we want life to flourish throughout the cosmos, AI is the key to it.

[688] As I mentioned in a lot of detail in my book right there, even many of the most inspired sci -fi writers, I feel, have totally underestimated the opportunities for space travel, especially to other galaxies, because they weren't thinking about the possibility of AGI, which just makes it so much easier.

[689] Right.

[690] Yeah.

[691] So that goes to your view of AGI that enables our progress, that enables a better life.

[692] So that's a beautiful way to put it and something to strive for.

[693] So Max, thank you so much.

[694] Thank you for your time today.

[695] It's been awesome.

[696] Thank you so much.

[697] Thanks.

[698] Thank you, but joy.

[699] So much.