#2521 - Aravind Srinivas

25:03Look, it's possible. This is a possible endeavor. It can be done. Yeah. But imagine what kind of technology would we have to need to map it out to make sure that it was all precise, that it all aligned. I mean, it's precise within, like, millimeters from point to point. And everything is done out of one piece of stone. Like, what did they do? Was it chisels? Did you do that with chisels? That's crazy. How many times did you have to sharpen your fucking chisel? That's nuts. Or did you have something completely different? Because some of the more intricate ones, see if you can find these. Some of the crazy ones, inside these temples, there's sculptures that are three-dimensional.

25:45And they're carved, like, inside of the sculpture. So there's, like, an outer area. And then there's all these openings. and then inside it's highly detailed. Like, how'd you even reach in there? It just says they use chisels and hammers, and I don't think that's possible. And careful geometric planning. Geometric planning. I assume there's people trying to do that. They see it like this is how much work someone could do in like 12 hours with a hammer, and they get nowhere, let alone like perfect and looking good. Yeah, it's nuts, man. And there's a lot of evidence of stuff like that all over the world, which is really weird.

26:22You have the stuff in Peru, like Sacsayhuaman. When you look at these stones and it looks like they're melted into place. And they're 900 tons. Like, what did you do? Yeah. How did he even get it up there? How'd you? Where'd they get it? How'd you get it there? How'd you align it perfectly? Built in only 18 years. How do they know that? How do they know that? Because it's attributed to one king. Yeah. So King Krishna I, 756 to 773 CE? Maybe. How do you know, though? Yeah, they said the archaeologists said it would have, they calculated it would take them 100 years to do it.

27:00Aravind Srinivas:Yeah. Yeah, I mean, this is where, like, you know. I don't know. Different historian accounts are all, like, muddled up, you know. Uh-huh. Well, it's a real problem. History is a real problem. But, yeah, it goes back to, like, the thing you were saying, right? You know, what is one thing that's common across all these different ages is human curiosity. So, I mean, that's something that, you know, I would love to get your take on this. Like, I've been talking with this idea called the curiosity premium, which is the most effective people, the most successful people have always been the most curious people, the ones who have been good at asking the best questions.

27:34Aravind Srinivas:And they tend to do better in every aspect of their life. And you're a good example of that. So that's why I would love to get your take on this. And the reason I believe that is because long term, people who continuously ask questions tend to do better. They make more money. They have a higher quality of life. They're happy. They have more compounding relationships. People find them more interesting. And so they compound their relationships over time. And so naturally, they end up succeeding. but their spirit of inquiry their intrinsic curiosity doesn't actually stop once they succeed it only they just channelize it even more and so that's why it keeps compounding and i would argue that like it's the only quality it's the only like quality that makes us really human you know in this world where we can seek a lot of information and get information way faster than ever before, it feels like that's that one universal human quality that's existed since ancient time, since the oldest text.

28:37Aravind Srinivas:In fact, in the Rig Veda, you're explicitly encouraged to seek wisdom more than wealth. And it's not just an idea specific to Hinduism. That specific idea exists in the Bible, exists in the Quran, exists in the Torah. It's not that seeking wealth is admonished by religious texts. It's actually that it's more important to seek wisdom. And, you know, like you can, why I said you're a good example of that is like, sure, you have a very, very large podcast. But the way you're running it is like you're just curious about a lot of things and asking a lot of questions. And I think that's that one quality that's very important.

29:14Aravind Srinivas:So, and I feel like it's the oldest thing. It's the only thing that we've known since ancient time, being curious. Well, I think it's stimulating to people and genuine curiosity is stimulating to other people. When someone is genuinely curious about something, I become curious about it. I think it's contagious. And I think that it's also an authentic quality. And I think there's something about really wanting to know something and being interested in something. And if you're curious, generally you're going to ask more questions about something so you have a deeper understanding of it. So if you're trying to do whatever you're trying to do, a sport, a game, you'll probably get better at it because you're more curious.

29:55Because instead of just assuming things, you'll ask more questions. You'll reexamine things. It's one of the most important human qualities. And to me, it's one of the most attractive human qualities. It's always been. When I meet curious people, I'm always interested. I'm always like, tell me what you're curious about. I'll tell you what I'm curious about. Let's talk. You know, it's it's and this podcast started out genuinely because of well, it was a lot of just talking shit with friends. But it also led into like one of my very first guests, actual guests, was Graham Hancock. And it's just I was curious because I had read Fingerprints of the Gods and I'd seen him talk.

30:32I'd seen speeches and I'm like, I want to know, like, what do you know? What do you think's going on? And he's another guy incredibly curious and absolutely fascinated with his takes on ancient history. He has been talking about this subject a long time. And when he first wrote Fingerprints of the Gods, I think that came out in like – I want to say it was like 97 or 98 or something like that. And I remember reading it and so many of my friends, you know, educated friends like this is horseshit. Why are you paying attention to this? More and more and more as time goes on, it's been proven that he's correct.

31:11The timelines shifted back. And from the publication of that book, the discovery of Gobekli Tepe and the surrounding area, like it's like, OK, now we realize, well, there was some crazy shit going on at the very least 11 ,000 years ago. So we pushed civilization back 5 ,000 years. So like, and this is just what we found now. And we keep finding things, keep digging, keep looking. And then you see the stuff that they're finding underneath the pyramid with this radio tomography where they're looking under the pyramid that seems that there's structures under the pyramid. You've seen that stuff? I haven't seen that.

31:46I had the scientist that's involved in it. He's an Italian guy, Filippo Biondi, and he came on the podcast. Wonderful accent. Almost as good as yours. It was amazing. But he's describing the use of this stuff and that they've used it successfully on known areas in pyramids and other structures. And they can – in fact, there's a – in Italy, there is a particle collider that is underneath a mountain. And using this technology, which is satellite-based technology, they get an accurate description of this particle collider. I think it's 1 ,200 meters underground. Like how far is that thing underground?

32:29We'll find out. But it's like deep under stone. And they find that they can get an accurate – like they can actually give you the dimensions of this particle collider. They have like an image of it. And this same technology is showing that there's these columns underneath the pyramid in various places that are 20 meters wide and they have coils around them. They don't know what the hell they are. And the whole structure of this thing, it's not small. It goes almost a kilometer into the ground. There's like this enormous like bottom of it and it seems like it's something that's constructed. And so they're like, OK, well, the pyramid is crazy.

33:12It's crazy enough. But if there's something underneath it that's a man made or someone made it, that's a kilometer deep into the ground. Like, what the fuck are we even talking about? Like, who made this? What what did they have? 1.2 kilometers into the mountain. That's nuts. It's a half a fucking mile in it plus into the mountain. And this thing can see through all that and get this accurate depiction of this particle collider. And it's showing with multiple scans, not just one, multiple scans and different technology, the same exact images, the same exact structures underneath this fucking immense 2 ,300 ,000 stone structure that almost perfectly aligns to true north, south, east and west.

34:02Like what? What was going on? Don't tell me police. Don't tell me copper tools like what the fuck was going on something crazy And I have a feeling our simplistic explanation of it is just doing no one any justice It's doing no service to history. It's doing no service to our understanding They've got to be a little bit more open in the fact that they are perplexed and not just perplexed by stuff like this This is a 3d print of an actual vase that exists in Egypt that they found that is, they found it in tombs of the old kingdom. This thing was somehow or another, it's made with diorite. So it's incredibly hard stone and made to like a thousandth of a human hair in, in it's yeah.

34:49Like crazy dimensions, like the way the, the precision of it and wasn't turned on a lathe because it has handles. Yeah. So you look at the handles on the side. Well, you can't carve the, and those are perfect too. Like the alignment of everything and it's like you just look at it. Oh, it's a vase. No big deal But no, it's kind of fucking crazy. Like how did they cut that out? There's also these There's all these core marks in some of the stones that they find in Egypt and they've analyzed the amount of revolutions per minute that you would have to go through to be able to cut through something and leave these lines and Defies explanation like what is this?

35:29This is crazy. This is not sand and copper and just rubbing things. No, this is some insane technology that we don't understand. There's scoop marks out of the bottoms of some of these stones. It's like, what the fuck is this? How did you scoop rock? It looks like ice cream. They just went, roop. What are they doing? There's so many questions.

35:53Aravind Srinivas:What tools did they even have to do all these things? They had copper. I mean, there's some evidence that they had some iron. And then I think Tutankhamen had a dagger that was actually made from meteorite, which is interesting. You know, like when they could find meteorites and make things out of them was very valuable, obviously. But the just the sheer volume of work that they did there. It's like you look at the Temple in Man. You look at the three major pyramids. You look at all the different temples and all the construction. And the older you go, the deeper into the sand they go, the more complex these things are, which is even weirder.

36:32So it seems like civilization after civilization, there was probably a rise and fall with their technology as well.

36:39Aravind Srinivas:Absolutely. Uncovered windows can make your home feel up to 20 degrees hotter. Stay cool and save up to 50 % off custom window treatments during the 4th of July mega sale at blinds.com. From outdoor shades to room darkening blinds, finding the perfect fit is easy. Get free samples, expert design help, and professional measure and install services, or DIY with confidence and support every step of the way. Shop up to 50 % off site-wide, plus huge savings on doorbusters, right now during the 4th of July mega sale at blinds.com. I think it's just incredible that none of this knowledge was properly documented ever.

37:16Aravind Srinivas:And it's a whole line of work to just go understand how to even rebuild these things, leave alone how did they build it. Well, think about what we're doing, right? So all of our knowledge is essentially stored on hard drives and paper. Those are the two things that are going to deteriorate the quickest. Maybe we should take a dump of the internet and - Put it on a rock. Go preserve it somewhere so that even if our civilization is wiped out and all the data centers are gone or whatever, whoever comes next can go figure it out. Well, I mean then you've got to always assume that even if they found a hard drive that they would like How long would it take for them to back engineer what we did and figure out what these ones and zeros actually mean?

38:03Yeah, that what is which is one of the most bizarre and fantastic accomplishments yeah modern civilization is that like this is a terabyte yeah, which is nuts yeah I don't know what your first computer had. I don't remember.

38:19Aravind Srinivas:Definitely not even a gigabyte, probably. No. Yeah. Like a few hundred megabytes was your hard drive. Yeah. Yeah. I mean, I remember when they first came out with gigabytes, I was like, this is nuts. Yeah. You remember like when Gmail launched and gave everybody like free email storage, unlimited email storage, and the bottom sliding bar would just keep increasing in terms of the total allowed size? Yeah. And that was nuts to me. Yeah. Yeah. And I think, yeah, we take it for granted that we have, like, infinite RAM and infinite hard disk. And nobody has to worry about, like, you know, back in those days, you would worry about, like, taking too many photos on your phone.

38:56Aravind Srinivas:Right. Right. And then you'd have to go delete all the old ones or bad ones. Yeah, you'd run out of storage on your phone. Yeah. And then you'd have to buy, like, an external hard drive to keep storing things. Yeah, I remember those days. You'd keep transferring stuff from your phone to the hard disk. I remember the old Android phones, you'd get an SD card. Yeah. You could slip one of those in there and you could store images on that so you could save space. And all that stuff is so vulnerable. It's so vulnerable. And again, if a completely alien society had to come down and find our hard drives and they went a totally different path of technology, they'd have to back engineer, reverse engineer everything that we did, try to figure out what are we using, what operating system, how's the operating system work.

39:42Is it Unix? Is it Linux? Is it like, what is it? How do they do it? It would be a nightmare.

39:47Aravind Srinivas:They would need an advanced AI to figure it all out for them. Right. Yeah. Yeah. And so that's just if the hard drives survive, right? So if there's some massive flood, cataclysm, whatever, some horrific thing that damages all of our electronics, which is totally possible, you know, just some solar flare, some intense. Or just another lab leak. Right? Yeah. Just time. A lab leak in time. Yeah. Yeah. It's nuts. And we could go back to zero real quick and we would basically be like preppers and hunter-gatherers. It would be hard to reverse engineer everything again. It would be almost impossible, which is why I'm really fascinated by the post-flood timeline.

40:32Because if these people like Graham Hancock and a lot of these other folks that have speculated that there was probably a very advanced civilization that went in a completely different direction many thousands of years ago. If you look at like the emergence of like Sumer and Mesopotamia and that area, which a lot of people attribute to be the earliest known civilization, that's around 5 ,000 plus, 6 ,000 years ago, right?

40:59Aravind Srinivas:Yeah, roughly. So the flood's like 11 ,000 years ago plus. So you're looking at like 5 ,000 years of what? It's not even that long in the grand scheme of things. No, not to the earth, but for people. Yeah. Pretty fucking long. Exactly. Like think of how long it took us to get our shit together. Yeah. It took thousands and thousands and thousands of years of people probably being monsters, just being the worst of the worst. Yeah. And that's probably the only way they survived. There's probably a lot of cannibalism. Yeah. There's a lot of murder. There was a lot of like horrific shit going on for 5 ,000 years until people slowly but surely figured out agriculture again.

41:38Yeah. Started building walls. Everybody relaxed a little. Got some solid weapons to keep people away so you could work on math. And then next thing you know, civilization emerges again. And it goes right back onto the cycle. And then you start reading in the Rig Veda about stuff that happened thousands of years. You go, what the fuck is this? Like, what happened? And that's my belief. I think there was something going on on Earth many, many, many thousands of years before established beginnings of history. That was very bizarre. And probably technology that went in a completely different direction than what we're doing now with combustion engines and circuits and all the different things that we use.

42:22They probably figured out some other kind of technology. Exactly. Which is totally possible.

42:27Aravind Srinivas:And it's amazing. It's amazing to think of what if we could rediscover all that again. Yes. Well, I would love to be able to... I would love to just one window in time to go back to see what it would look like, I would 100 % pick ancient Egypt while they're building. The pyramids. Show me what the fuck was going on. Just put me in a big hamster wheel. There's a big plastic bubble where no one could see me. Just let me violate space and time and exist there for just a few minutes. Just let me look. I think that would be the most insane thing that you could see about humans in human history. Yeah.

43:07I just, I want to know what they knew, what they had, what they used. Just looking at this thing, Petra's same time period, at least attributed to, 7 ,000 roughly BC. Jesus. And they, you know.

43:20Aravind Srinivas:How would you do that? Right. How? The details of all those carvings is just insane. Insane. Yeah. And in 7 ,000 BC, what are the tools? What the hell were you using? how did you make a temple out of the side of a fucking mountain look at the size of it man the size of those columns it would be hard to do anything like this even today it would be incredibly difficult insanely time consuming oh uh yeah the kagliasa temple by the way i don't have it up right now but in 1650 or so someone sent a thousand people to try to destroy it and after three years of doing nothing they stopped they barely made a dent on a couple statues Yeah, a lot of times when invasions happen in India, like...

44:07They tried really hard to fuck it up and couldn't. Oh, wow. That's crazy.

44:12Aravind Srinivas:That's very robust.

44:17That's a great way to describe it. It's just, there's so much of that stuff that's so interesting. Because it's so undeniable. It's so undeniable in its scale. so undeniable in its complexity and the planning and the understanding that you had to have a deep knowledge of geometry, of measurement. You had to have accurate. Yes.

44:43Aravind Srinivas:Yeah. Everything. Sturdiness, like resist calamities, earthquakes, floods. And then even if you had that, what tools are you using? Yeah. Like how are you doing this? How are you coordinating all these people? Right. Getting them to do stuff. I mean, sure, conditions must have been way harsher. Like, I'm sure people didn't really have a choice but to do these things because back in those days, like, the only way you could take care of your food and clothing and shelter is, like, you commit yourself as a laborer to the state, to the kingdom. But you could also ask, like, what gave them the initiative or drive to go do these things?

45:20Yeah. Well, that description is perhaps of a later time. We don't even really know what civilization was like when these were constructed. Yeah, the real the real problem is the material science The real problem is like you there's a lot of things that you have to have to make those things It's not as simple as a sculpture like Michelangelo making a sculpture out of something That's like fairly easy to carve into stars as far as stone goes now This is the scale is impo. It's so undeniable that like something something some piece of our understanding is missing yeah yeah i mean it it

45:59Aravind Srinivas:oh like looking at all this like everyone should just be like a lot more humble right like like yeah we don't actually know that much like what we know is like so little like right whatever like the same thing as what socrates said what we know is very very little and the only thing we we should all strive to be is just be curious and I think there's a lot of tendency for people to like think like oh like we have all this advanced technology we're so amazing like look at us and it's like wait hold on like you don't even understand what happened thousands of years ago and there's so much out there to just go and explore and learn and like get better at understanding more what is this place this is yeah this is unreal this is called the Elora Caves timeless wonder carved in stone.

46:46I think it's all kind of the same area.

46:48Aravind Srinivas:Yeah, it's the same Allura cave and the Shiva temple that you saw. Look at that. My God. Look at this stuff. It's insane. And again, there's no steel back then. It's actually really symmetrical. It's not even like... Can you go back to the first one with the symmetrical top? Yeah. Look at the symmetry at the top. It's nuts. It looks like that mall in New York they made where the world transformed us. Yeah, but way more robust. I mean, how? What were they? This is the thing. It's like the material science aspect of it. It's like you don't have the ability to do. Look at that top one. Go to that top one again, the one that you just had, Jamie.

47:40Yeah, that one. look at that's crazy man i mean i am just blown away when i see stuff like that my mind just starts racing and i just think how did you do this who who was involved how was it planned how is it so symmetrical what were the tools like what were the tools man yeah if you don't have steel you don't have what are you using how'd you do i mean most of it is done with stone clearly I guess. I guess. I doubt it. I bet they had something else. I bet they had something else that over time eroded just like metal would today. I mean, if you left a shovel outside today and you came back to that same spot 500 years from now, there's nothing.

48:28That shovel's gone, right?

48:30Aravind Srinivas:Yeah. And you've got to assume that these many thousand-year-old temples that were carved out of a fucking mountain, whatever tools they used probably got absorbed by the earth. And the only thing that's remaining— Yeah. That's giving me a weird thought. Like when they make a big building downtown, though, they only bring the crane in for a temporary period of time. And there's only so many cranes on the planet currently, too. Right. True. You take it, you move it, you go take it to the next spot. Yep. Yeah. True. Yeah, I don't know. Yeah, especially something like this. Like if they had heavy equipment and machinery and whatever the fuck they were using, they probably moved it and then moved it out and then it probably rotted away.

49:08And now it's gone. If there was machinery. If there wasn't, like there must have been something else. Some other kind of like some technology that we haven't even imagined. Yeah. But it's like their commitment to art, too, was so fascinating. Because these aren't just structures. They're in passion projects. Intensely beautiful. Intensely ornate. Yeah. So it's not just that they wanted to build like a functional structure. Yeah. Good architecture. No, it's a fascinating artwork. And it's so intricate. There's so many different features and so many different images of different people and beings and animals and elephants.

49:54Aravind Srinivas:There's one more temple you could pull out. It's called the Tanjar Temple. Oh, I've seen that one too. Yeah. That was done more recently in the age of the Cholas, and it's pretty incredible. When did they do that one? I don't know the exact number, but more recent than the ones that you saw. All of them are nuts, man. And then there's stuff like that all over the world. Whoa. This was done as a project by the king to basically make a name for himself. Wow. That's incredible. Is that multiple pieces of stone or did he carve that whole thing out of stone too? Probably multiple pieces. So that's actually like construction.

50:38Yeah. Not like removal. The other ones are, it's essentially a giant sculpture. Wow, it's so pretty. Look how geometric it is too.

50:51Aravind Srinivas:That's what amazes me. Like they didn't actually have all these simulations and CAD tools and all these things. Right. And what year was this made, Jamie? Does it say? It's just so incredible how much of this stuff exists where it's really baffling. Like I just found out recently that the Aztecs didn't build those temples, that they found them. Really? Yeah. They found like the Tenochtitlan. They call it the place where the gods were born. The Aztecs found it and uncovered it. And then on the when is it to know shit lawn or Teotokan, whichever one it was on the consecration day when they were done with like whatever they were doing with it to celebrate.

51:38They killed somewhere between 20 ,000 and 80 ,000 people in four days. Damn. Not exactly the mindset of the type of people that would construct something like that. You know, so those are the people that found it and it might have been sitting there for a thousand years. And then they came along and said, oh, this is cool. Let's live here. OK, well, what was the society that lived there before them? And where are they? And what happened? And how they do this and why they do it? And why did they have it aligned with the constellations? Like, what were they doing?

52:10Aravind Srinivas:Yeah. Some of the calculations are pretty, pretty amazing. like how they timed it, how they positioned it, how they cared about planetary positions and stuff like that. Sure, like some of it could even be pseudoscience, but whatever. I think just the level of like calculations they were making back in those days without, you know, powerful computers is just outstanding. It's just nuts and it doesn't make sense. It's like, okay, they're making it without powerful computers. So what are they using? I mean, at one point, the word computer just meant a human. human beings would be doing the calculations.

52:48Aravind Srinivas:That was their only job, to literally multiply two numbers. Astronomers were actually the first mathematicians. The term mathematician and astronomer were used synonymously at one point. Really? Yeah. Why is that? Why studying the stars and math? Yeah, because studying the stars involved making a lot of geometry calculations. and that was kind of actually one of the first set of mathematicians in India. People like Aryabhata, Bhaskara, all these guys were actually astronomers too. They were not just mathematicians. And Aryabhata was earlier still like the idea of using zeros and then he had a lot of like contributions in geometry and he was doing all this like just because he was interested in astronomy.

53:39me isn't there evidence of pythagorean theorem in ancient yeah is it ancient sumerian is it it's some ain't some something that predates pythagoras interesting my my theory is that

53:58Aravind Srinivas:even though it was not formulated as a pythagorean theorem like i'm sure people had to understand concepts of science sines and cosines and like you know whatever is the right angle for the right inclined to get this right level of like geometry you need to you needed to have some implicit understanding of it to build these kind of structures there's no way you could do it without that yeah 100 and you have to have incredible measurement tools like not just the actual mathematics okay the oldest known evidence of pythagorean theorem dates from old babylonian clay tablets from about 1900 to 1600 bce roughly 1000 years before pythagoras isn't that wild like how how clay tablets often cited uh use what we now call the pythagorean theorem to complete to compute rather the diagonal of uh rectangles and squares including an excellent approximation look at this this is nuts man vedic ritual text explicitly states the rule equivalent i don't know how to say that what is that a squared b squared c squared but the diagonal of a rectangle that includes numerical examples predating or roughly contemporary with classical greek mathematics so completely different parts of the world yeah and they're coming up with the same stuff exactly because they're all curious that's it yeah they're all curious and eventually all curiosity leads to truth or some form of it i would argue that anything anything that's of impact on the world has only been done by curious people.

55:32Aravind Srinivas:In hindsight, we label those people as successful or smart or rich, but the common trait across all of them has been like curious. Well, that's certainly a powerful trait. And people that aren't curious are not fun. Yeah, they're not interesting. So because of that, they don't attract other smart or interesting people. And therefore, they won't be able to do something very meaningful level so it's it's it's it's kind of like um um it's less about and it applies to your personal relationships and personal life too it's not just about professional success like you have a more fulfilling life with your wife or your kids if you're a more curious person you ask them more questions you you take interest in them right so that's that's the one quality everybody wants in personal relationships is like taking interest in them and like actually understanding them better or like being curious about common things.

56:29Aravind Srinivas:And so it's not just that, you know, being curious leads to success. It's more that people around you want you to be successful if you're curious because you will have more compounding and fulfilling relationships. I would agree with that. Yeah. I'd say it's one of the more important qualities of human beings. I mean, it's led to everything that we have today. All curiosity has led to all of our architecture, math, everything, art, everything. The transistor. You know the story of the transistor? Yeah. So Bell Labs was basically employing as many, like history adjusted, as many telephone engineers back then as the number of software engineers today.

57:07Aravind Srinivas:But only three people cared enough to question whether you should use these really hot giant vacuum tubes for amplifying telephone signals. So vacuum tubes were very big, power hungry, and very hot. And so they were not fault tolerant, and it's very expensive. And so three people questioned the need for that and came up with the idea of the transistor to amplify current. And that was the Nobel Prize winning discovery. And not just that it was useful to amplify telephone signals, it basically led to the rise of modern computing and we wouldn't have an iPhone like this today if not for those three people.

57:44Do you know what the tinfoil hat conspiracy theory about transistors is? No. That they are back engineered from the Roswell crash along with fiber optics.

57:56Aravind Srinivas:Tell me more. So we read this on the podcast. Remember, Jamie, there's the two scientists that were attributed. There's this one scientist that said they weren't even remotely exceptional guys and that they gave them the credit for this so that they didn't have to reveal the true nature of where this technology came from. I see. Interesting. So, again, tinfoil hat securely on our heads. This is not something I believe. Okay. This is just something that's fun. there's a few inventions that came out of that time period roughly after 1947 that are weird and one of them is fiber optics and one of them is a transistor and these are supposedly attributed to back engineering programs so the roswell crash i don't know if you ever paid any attention to it it's a real weird one because the cover of the roswell daily record said that the government has a crashed disk that landed in the desert.

58:51A bunch of witnesses, a bunch of people saw it. It's also people that supposedly saw physical bodies of these creatures and supposedly, again, who knows what's true. But Truman went to the site, he visited it, and then the planes, two separate planes, were flown to Wright-Patterson Air Force Base, which was I think was just Wright Base back then. I don't think it was Wright-Patterson. But they flew them out, and the idea was this material was so important they didn't want to risk one plane crashing. So they flew it in two different planes. And that this stuff has always been known to be stored at Wright-Patterson Air Force Base.

59:36That's what everybody always talks about. And then a lot of it was moved to Bell Labs. Oh, wow. There was a company called the American Computer Company. and back in the day the american computer company was just like it was a consumer website where you could go and say oh i need a windows computer that does this that and the other thing and you could just put in whatever your specs were and they would build it for you but they all had a whole section of their website dedicated to bell labs and back engineered ufo technology and all they talked about and this one like whoever ran it was like a fucking kook i don't know is that still around yeah American computer company is it still around interesting so this is like the 1990s I say your theory I mean not that you believe in

1:00:21Aravind Srinivas:it but your theories that the transistor was not like invented it was known and it was given to the there's apparently a giant leap between the first ideas of the transistor and then what it what actually came about and how much money had to be spent to create it off of this leap this was this This is a assertion by these scientists that we're trying to examine this. The thing about Bell Labs is there's a military base right outside of Bell Labs. And they say, well, that military base is to guard New York City. But New York City is quite a flight away. But Bell Labs is right there. Yeah. And they're working on some deep, dark shit at Bell Labs for sure.

1:01:01Because I've had a bunch of people on that were talking about remote viewing exercises that they were doing out of Bell Labs. We've had a bunch of people that came on and talked about various programs that were going on that were like top secret programs that were happening that were being run through Bell Labs. There's some weirdness to that place, like real weirdness. Yeah. And it's fun. Yeah. The idea that like –

1:01:25Aravind Srinivas:Because it definitely feels very disconnected. Like, okay, like you were using all these vacuum tubes and then suddenly you're like, okay, like what if we just use semiconductors? okay, there's definitely a pretty far drift from what you're doing currently to what you're supposed to do. And also the idea of the first transistor and what ended up being used in chips, the junction transistor are quite different too. So there are like big leaps in terms of what the core idea was. It's not an incremental change. The way I thought about it was like, okay, that's like tens of years of work. and that's why they made a big change.

1:02:08Aravind Srinivas:And so if you actually looked into the individual milestones they had, maybe it would have looked pretty different. But your conspiracy theory is pretty interesting. It's always fun. Yeah. And also there's just too many stories of this. And David Grush has, you know, on oath said that there are back engineering programs and he was read into these and that they've been around for a long time. But this is the assertion of that movie, The Age of Disclosure, that the real problem is that they have misappropriated funds and lied to Congress. And so if they come out and tell you, okay, we do have this program, well, guess what?

1:02:42Everybody goes to jail because you guys are a bunch of liars and you've been stealing money and you've been doing whatever you want to do with this money. I don't know how much oversight is there on back engineering UFO programs. So probably a lot of people get in trouble. A lot of people go to jail. On top of that, these things are all being done by weapons manufacturers, right? Like, where are you going to bring them to? Well, you're going to bring them to Lockheed Martin, or you're going to bring them to, you know, Rocketdyne. It's going to be someone that does that kind of work. You're not going to do it on your own.

1:03:16It's not going to be like, we'll do it. No, you're going to have to bring it to people that already make spaceships or bring it to people that already make jets. And so they have a massive competitive advantage over any other company that's doing it. So then there's other companies that also had contracts with the United States government. They can sue. And so he lays out all the problems with disclosure. And their assertion is that the only thing what we need if we really want to find out the truth is we're going to need widespread amnesty for all these people that were involved. My problem with that is that's what I would say, too, if I had been stealing money for decades and decades.

1:03:53I'd be like, we need amnesty. And then I'll tell you where all this stuff is. I'm like, how do we know what this stuff is, whether or not these are just top secret military programs with advanced propulsion technology that's unavailable to the public? And they're going to say that is aliens and they back engineered this and they did that. Like they clearly don't want to tell people. They don't want people to know. I think a large part of it is probably because they could get in trouble. But I think also a large part of it is because it's fun to keep secrets from people.

1:04:23Aravind Srinivas:Yeah. Especially when you're the government. Why tell them? Fuck those people. yeah fuck them they don't even know ufos are real meanwhile you know we're going into a bunker in the middle of the mountain and we're remote viewing you know it's it's probably there's probably a lot of fun involved in having access to information that most people would kill for yeah i mean there's so much information that um we just we just don't have access to which brings me to this question with it seems like one of the things that's happening with uh both with ai and with technology in general is that you have more and more access to information and more and more answers to questions than ever before.

1:05:06At a certain point in time, there's going to be no bottleneck.

1:05:11Aravind Srinivas:Yeah. And we're going to know everything about everything. So how is anyone in government going to keep a secret? How is any corruption ever going to be possible? At a certain point in time, all of it will get uncovered. Like it's much more difficult to commit murder now with DNA evidence, right? Back in the 1800s, like I didn't see nothing. I wasn't there. And then you're free. Like now they do your fingerprints. Now they get your DNA. Now there's flock cameras. There's like more and more and more. It's harder to get away with things. So it seems like to me, like whatever they have, whatever anybody has, ultimately there's going to come a point in time where there's so much data and so much information and you could run all your questions like there's an ai fact checker for politicians now yeah so while a politician is giving a speech you can run an ai fact checker and in real time it will tell you whether or not these people are full of shit this it seems like the direction is there's not going to be anybody full of shit in the future because it's not going to be possible yeah i mean the government still would have access to things that we human beings wouldn't have access to like regular people and particularly defense related weapons related like for example when they did the Venezuelan thing I don't think people in Venezuela even understood like what even those weapons were we did they were described as something the literal words used were like alien like technology so even we didn't know that that the United States had access to that quality of defense technology until that incident happened.

1:06:53Aravind Srinivas:So there are obviously going to be secrets, right? Especially the highest stakes things. I would say like building frontier AI models is similar to that. Of course, as more and more models are getting open source, I think the knowledge is diffusing. But still, the true amount of details you need to actually train a really amazing frontier reasoning capability model is still not like widely diffused. So my hypothesis is that whatever is extremely high stakes will still not be widely diffused. At least there'll be enough structures in place to keep it secret. Forever? Not forever, but for a while.

1:07:39Aravind Srinivas:For a while. Yeah. That's the thing. Long term, sure. things do get out and people do understand. But it feels like long term is what I'm looking at. Look, when we're looking at history, we're talking in these, when we're looking at all these different temples and all these different things, we're talking about thousands and thousands of years. And thousands of years time span in between each individual one. With our world, we're talking about massive change in 200 years. This country is 250 years old. Think about how kooky that is. That is a blink of an eye in history. But do we understand everything that happened in the United States?

1:08:14Aravind Srinivas:No. Exactly. So there are still some details that are hidden from us. We don't fully understand everything, right? For now. Yeah. But my question is, as time goes on, 250 years from now, is it even possible to keep any secrets from anybody? And is that a good thing? It might be a good thing. It sounds horrible to people because they're like, oh, my God, what about privacy? Right. But also, what about lies? No more lies? Like everyone's going to know what you're thinking? Everyone's going to know everything people do all the time? Yeah. I mean, if you're a true surveillance state, obviously there are no secrets.

1:08:53Right.

1:08:55Aravind Srinivas:Except about the government itself. That's the problem. Yeah. Does it bottleneck with the government or does it get to a point where you can't even have government secrets? because as technology evolves and as human civilization evolves, secrets will be less and less, not just necessary, but secrets will be problematic because there'll be an impediment to knowledge, the impediment to understanding the true scope of what the world is, like the true nature of all of our various moving parts. Yeah. I mean, as long as the human quality, the intrinsic human quality of curiosity and truth-seekingness, which is, you know, universal, it's existed ever since we've known human beings, If that continues and that continues to be the case, then people will have enough incentives to figure out the truth.

1:09:42Aravind Srinivas:Yeah. And if something is actually hard to get to, it only motivates you more to actually go and find it. For sure. So my question is, where does this all go? And you obviously work in AI. And when you think about AI and when you think about just technology in general and you extrapolate, you just take it from here and you just plot it out. Like what is a possible scenario of 250 years from now? Like what does it even look like? What does the United States look like at 500 years old? It's very hard to know. I'll be very honest. I think it's very hard to know even five years from now how it's going to look like.

1:10:21That's crazy.

1:10:22Aravind Srinivas:Yeah. Five years ago was like. Five years ago, whoever is at the top most in AI, I don't even consider myself like that, but whoever is at the most frontier level of decision making in AI five years ago, I don't think they predicted the exact state we are in today. Nobody did. If they did, they would have already procured all the compute and manufactured all the chips, bought out all the fabs. They would have done all that, right? Just this is counterfactual. everyone's like bottlenecked by not having enough compute and like we have we don't have enough chips we don't have enough power these are all the problems that if you invite anybody in ai and ask what is the bottleneck in ai today and everybody would say power i think jensen was here and he said the same thing right yeah but okay like if you predicted this exact state five years before wouldn't you have secured enough power and started building more power plans yourself and start getting permits and like started like planning out capacity no nobody did that But everything is reactive to the demand that we're having today.

1:11:22And that's just five years.

1:11:23Aravind Srinivas:Yeah, that's just five years. So when you ask me to predict 250 years, I just have to honestly say I don't know. Do you ever sit back and think about it, though? I don't think about it. What it could be? I don't think about it. So there are a lot of fun. I use perplexity a lot for these kind of things, especially this new feature computer inside it. And this is just for hypothetical scenarios. Let's say there is an AGI. right i've seen you ask a lot of people about this and um and um a lot of conventional answers is like oh like we'll just become managers of the ais don't worry but um if if the price of cognition is the price of compute managing an ai is also pretty much doable by the ai itself because it's the bottleneck is not like unique cognition capability there so the value of the society will automatically shift to what is scarce and uh fundamentally what has been scarce is like asking like high quality questions about things okay like what if like we just completely spent all our time understanding the past like that's an interesting endeavor it was not cool before but it'll it'll become cool again um and like we usually use the view like archaeology or history is not something that's like worth having a career in because it doesn't pay well but what if it actually starts paying you a lot more now that like actual knowledge works being done by AIs and like it's all mundane and all the price of that is basically a zero right and archaeology would be one of the few things that it wouldn't have access to because it doesn't have the actual ground it can't get into the ground and do the scans and well let's say we have like robots to go do that but but you're still going to be the one probing right because you have incomplete information all the time even the idea of like okay let's go explore this particular area let's go understand better oh let's let's go try to reverse engineer let's go try to build this again how would it be if we wanted to do the same thing on the moon there are like so many interesting projects to work on for us as long as we are we stay curious and we stay interested in like a lot of things that we've done before and trying to understand like civilization that i'm not really concerned about like what things we get to do we might be doing a lot more cool things for what it's worth like i don't know if anybody will be like coming and telling you that oh it's so cool to like open an excel sheet every day and make financial models right compared to like there's got to be somebody out there that likes that i mean there's something about like the task you do and and and what you get paid for like what is the job title blah blah blah and some people associate their personal worth with like where they work at and how much they get paid And I think that thing is going to collapse in a world where the price of all that cognition is going to be the price of compute.

1:14:14What do you think happens to people if a large percentage of jobs get replaced by AI?

1:14:24Aravind Srinivas:I think they'll find new things. We've always gravitated towards things that are scarce because that's where the value lies. and so if you know have you one interesting analogy is the Gulf states where there's an abundance of resources and they export their resources to other states and that pays for the whole state you know how they offer everybody free electricity subsidized health, subsidized education and like no taxes when I first went to Dubai almost like 20 years ago. They told me people don't pay taxes here and nobody pays for electricity here and education is super cheap. And I was like, wait, how is that real?

1:15:18Aravind Srinivas:And the way that's real is that, I mean, of course, Texas also has no taxes and any well-run state can do this. But the way it's happening is that because the government provides you all these things, it becomes a rentier state. Like you offer political acquiescence to the state. And what ended up happening is citizens there expect the state to find them jobs, expect the state to take care of like job displacement for them. So they don't worry. So it made them a little more lazy. So that's not a good future to have where some people talk about AI subsidies and AI dividends that get paid to everybody.

1:15:55Aravind Srinivas:I think we need to do some form of that, but that in entirety, won't solve the problem. Right. Well, the thing about human nature is sort of undeniable. And if you give people the ability to be lazy, a large percentage of people will take that. That's right. A large percentage won't though. Yeah. There's going to be enough people that are inspired to do something and they say, okay, well now my basic needs are taken care of. Let me pursue my actual interests and find purpose in that. Yeah. Because that's a lot of people find purpose in whatever their occupation is. And if we can shift that to finding purpose in what your actual interests are, and then really pursuing something, whatever it is in that, then you'll still have meaning in your life.

1:16:38Aravind Srinivas:It keeps coming back to staying curious and finding value in your relationships, your family, caring for each other. If you ask a lot of retired people, actually retired people is a good demographic to understand what would happen, what are things people find meaning in after work's taken off them. And the majority of the answers are always family, caring, personal relationships, and community. These are the things retired people keep doing to keep themselves active and wake up every day and have something to look for. So all those things will become even more important at a time when work itself doesn't mean much.

1:17:23Aravind Srinivas:Doesn't mean humans won't be status-seeking. I think we'll still be. But status is not going to come from whether you're working at a particular famous bank or a tech company or whatever. It'll be driven by how interesting you are. Are you interesting to talk to? When I can talk to an AI, despite that, are you still interesting to talk to? Are there certain things I get out of talking to you that completely changed my perspective about a bunch of things? Or is it just fun to hang around you? Can we have a compounding relationship together? And I think, again, it goes back to being curious about things.

1:18:02Well, this is best case scenario, right? Worst case scenario is civilization upheaval, chaos, civil war.

1:18:09Aravind Srinivas:And it's possible even without an AI. Right. Yeah. Look, we've gotten real close to it a couple of times. Exactly. Yeah. And we did not need an AGI-like scenario for a civilization collapse in the past, as you've clearly seen. Right. A calamity can just take out all of us, wipe out everything. Sure, especially natural ones. Yeah, that's why I'm not a big fan of everybody claiming that the AI is going to kill us. So AGI is going to destroy humanity, and it's too dangerous, and we all need to stop doing these things. But at the same time, continuing to fill all these data centers and continuing to make money.

1:18:44Aravind Srinivas:You have to have one consistent position. My position is that whether AI or not, I think being curious is going to serve you really well. I think it's going to help you have a better life. And there are two paths to curiosity, one that can kill it and one that can supercharge it. In my opinion, the one that kills curiosity is algorithmic feeds. The brain rot that you're fed every day, which is continuous doom scrolling. that's bad. And the one that can supercharge it is AI. Okay, like now that you could just ask whatever you want, if everybody has like a pull it up Jamie for them, you know. Right.

1:19:28Aravind Srinivas:And that's amazing. So all you have to do is be curious about a lot of different things. And of course, talk to interesting people, engage in interesting activities together. If money is no longer an issue, you can fund passion projects yourself. You don't have to require government funding or venture funding. Like what if you just wanted to build a mini cave yourself? Okay, like you find a piece of land somewhere. There's a lot of land in America, way more land than we know what to do with it. And surely we can build a lot of interesting things there. Well, that's a good glass half full scenario.

1:20:06And one of the things that I keep coming to is this whole idea of people working and making money and having careers and having portfolios and bank accounts. This is all very recent in human history. Yeah. Very recent. Very recent.

1:20:23Aravind Srinivas:It's very recent. But we've become accustomed to this as a way of life. And Microsoft built this concept of a knowledge worker because they wanted to sell more office software. Really? Yeah. Like this whole idea of putting a PC on every desk and making you like glue to the PC was there. That was Bill Gates' mission. Put a PC on every desk. That fucking wizard. What an incredible accomplishment because, boy, did they nail it. Yeah. So it was not about making computing like beautiful or anything in the way like Steve Jobs and Vision did. Right. It was just about like sell more software. Sell more computers because that way you can sell more software.

1:21:01Aravind Srinivas:And if you sell more software, you become rich. And the company just was a machine that was just, you know, built essentially a large sales machine that's built to sell software.

1:21:15Aravind Srinivas:And now they sell cloud, but whatever. That's essentially the reason that, like, you know, we all got trained to use software. People went and did tutorials on how to use Excel, how to use Word, how to use all these email tools. And then now that became the upskilling you needed to go work at different companies and then write code and like whatever, right? So if that part is going to be done by an AI, it's not necessarily a bad thing because this is not actually the way you feel like real purpose and fulfillment in your own life. If you were never exposed to that, whatever you had as the intrinsic curiosity in you, that's probably what you should be doing.

1:21:57Yeah, there could be a completely new way to live life where we're not dependent upon labor for basic needs. But then it's going to be incumbent upon people. They're going to have to figure out a way to be either self-starting or we're going to have to expose people to things that are going to excite their curiosity and make that a mandate.

1:22:18Aravind Srinivas:Yeah, it has to start from schools. Yeah. And as long as we keep rewarding people for having answers instead of asking interesting questions, it's going to be a difficult change. Like in schools, you're always rewarded for being smart based on whether you have answers to like 20 different questions. Like who cares? Like all those 20 questions can be answered by AIs. Have you ever like flipped a script where you say, okay, like I'm going to, the smartest person in the room is the one who has the most interesting questions. Okay, like what kind of students can you cultivate based on that? Like imagine if the room had no pressure to always know the answer, but the freedom to ask a lot of questions.

1:23:02Right, because sometimes when someone asks a question, it'll make you pause and go, I never even thought of that. But that's it. Like that's the question.

1:23:11Aravind Srinivas:Yeah. And it takes a – I mean so many people have so many different perspectives, which is one of the more interesting things that I've experienced doing this podcast is I get to talk to so many different people and they vary so widely. There's so many different ways of looking at the world and so many different ways of engaging with the world and so many different things that people are fascinated with that they spent their entire life studying and pursuing. It's like you get this rich tapestry of the human experience that's just, I would have never been exposed to this many people. And in turn, I've been able to expose these people to all these other folks that are just listening and watching right now.

1:23:52And it's fucking incredible. And it's such a, for me, it's like the perfect job. I've never had a job that more aligns with my own personality as much as this. Because I've always been that kid, like, shut the fuck up with all the questions. I've always been that kid.

1:24:07Aravind Srinivas:That's the system, right? It's not your fault. Right. Like, it's actually, the reason you're successful now is the exact thing that people told you to shut up about in the past. Right? Yeah. You know, hey, you know, stop bothering my lecture, you know, asking all these unrelated questions. It's mainly a frustration of the teacher that they don't have the answers to you. Right? Sure. And now that bottleneck is gone. We did this experiment with one instructor at MIT who taught the Introduction to Biology class. where he came and told us that he's going to give perplexity to all the kids, all the students, and they would use it as part of the lectures.

1:24:49Aravind Srinivas:So instead of fighting AI, you just give AI to everybody and let them ask whatever questions they want, and they can actually use it in the exams too. Oh, wow. So how do you even design questions for an exam in such a world? is maybe you just encourage people to pose a question that AI can't answer right now, and that becomes your research project, and you turn everybody into a scientist. Fundamentally, there's this belief that scientists have to go through a rigorous PhD, and you have to get accredited by an amazing university to be that. Sure, but anyone who's curious can be a scientist. The only thing that's required to be a good scientist is intellectual humility to understand that you could be wrong about things, things that everyone takes for granted, you could still question them.

1:25:41Aravind Srinivas:And when you're presented with new evidence and new data, you're willing to change your mind, and you're willing to operate with ambiguity and uncertainty about the world. That's basically all the qualities you need to be a scientist. And you can run your experiments. You can gather data. You can gather evidence. And you can consult people. You can bring in experts and talk to them. And as long as you're uncovering more and more about the world, you are a scientist. You don't need a PhD to feel that you're allowed to be a scientist or not. And I think that's the most important quality we need to inculcate in our kids, the upcoming generation, so that they all feel more liberated.

1:26:22Aravind Srinivas:Okay, like finally, I don't have to memorize this textbook or these lecture materials. And like, I don't have to feel bad if I get like 12 out of 20. Okay, who cares? Like AI is always going to get 20 out of 20. That's not what you're meant to be like good at. Of course, master the foundations, the basics, great. But your job is to actually pose interesting questions. Yeah, and the intellectual, excuse me, intellectual humility is so important. Because one of the things that was really weird about the whole COVID pandemic was that we weren't supposed to question science. Yeah. Or when Fauci said, if you question Anthony Fauci, you are questioning science.

1:27:01Aravind Srinivas:That's because they try to assign credibility through their degrees. Yes. Through their affiliations. Yes. Appeal to authority. But not through the scientific method. Right. Anybody should be allowed to ask questions as long as they are open to new evidence. Yeah. And that's the most important quality of a scientist. Well, the scientific method alone, I mean, it's one of the most important things that we can use to try to figure out what's real and what's not real. And as soon as someone says, don't use it, don't question, well, wait a minute. And then there was an actual government push to silence questioning and legitimate researchers were kicked off of Twitter because they didn't back the narrative.

1:27:42Aravind Srinivas:Yeah. Like this is all anti-science. This is not, you're questioning science. Well, science demands questioning. Yeah. It's what it is. Yeah. When you don't understand something, the best thing you can do is ask all possible questions. Right. And so curbing that is almost like a way of saying, look, I'm going to tell you what happened. You need to believe in my worldview. And I'm not open to new perspectives. I wonder if anybody has used AI to try to map out possible scenarios for where technology leads human civilization and what could be done to mitigate the problems and push it in the proper direction, like have a bunch of different models of how this could play out.

1:28:27Aravind Srinivas:Yeah. I mean, I try to do that for fun, but I haven't done it in a serious enough way to have a proper answer to that. Right. But I think a lot of things that we are doing today will not be considered needed or valuable. And maybe a little bit of taking our own lessons from the past. I don't know if you, when you grew up as a student, did you have to be good at mental math, like multiplying arbitrary numbers? Was that considered a sign of smartness or remembering people's phone numbers or something? Well, you had to because there was, I mean, you had little address books. That's what we used to carry around.

1:29:05I had a little address book that I keep on my desk. It's a little tiny thing with everybody's number and name. That's the only way I knew people's numbers. And I remembered a bunch of them, like all my friends. I had all my friends. I don't have any of my friends' numbers remembered. Maybe my wife and my friend Eddie. I have two numbers in my head.

1:29:22Aravind Srinivas:But was there a time when people thought somebody was smart based on how good their memory power was? Oh, yeah, definitely. Would you say that now? Well, people are impressed if you know things now. I have a bunch of weird information, obviously, that I've gathered through so many years of doing this podcast and just so many years of being curious. Sometimes even my own daughter is like, how the fuck do you know that? I'm like, this is what I do. That's my thing. I pay attention to stuff. But yeah, I mean, memory itself is always very impressive. And someone has an excellent memory. and can pull out facts of the past, we automatically equate that to intelligence.

1:30:04Aravind Srinivas:Yeah. I think it's impressive, but it's not necessarily a sign of being intelligent, right? I think you have a very fast lookup table in your head. That's great. It's very valuable. But I still think being smart is all about posing the most interesting questions. Also, the decisions that you make and whether or not you self-correct when you make mistakes. Yeah. Yeah, all those things. Exactly. So when you have an amplifier to intelligence, like an AI, all the time, where lookups is essentially something you can delegate. Reasoning for decision making is something you can delegate. But posing the right questions to gather the right data, and then forming your own judgment based on what it reasons and comes up with, and finally having the courage to make the decision, that's still you.

1:30:52Aravind Srinivas:That agency, that intrinsic curiosity to ask the right question, and the scientific intellectual humility to, like, you know, gather new evidence, always questioning your beliefs, that is still you. And so I feel like that is essentially what would be considered smart in the ages to come. If somebody is, like, you know, like a proxy scientist or whatever, like, no more, doesn't have to go to, like, MIT or Harvard and get a PhD to be a scientist or to be considered a scientist. Because all scientific literature is open and, like, it's accessible to everybody. And you can even take a paper written by an expert and use an AI, understand it deeply, ask a lot of questions, and maybe even disprove what they claim to be true.

1:31:36Aravind Srinivas:That's the whole peer review process, right? The peer review process is all about questioning somebody's paper. And that's why, like, you know, whatever you said happened in COVID days is wrong. You should be allowed to ask questions about even eminent scientists' work. It's okay. If you're dumb and you had the wrong questions, sure, you're going to learn from that. It's worse than not being allowed to ask the question. Yeah, agreed. It's going to be interesting to see what the future of education looks like, like how valuable are degrees when essentially AI is going to be able to do the majority of whatever work you need done on a variety.

1:32:17Like, how good are they right now at just law? Like, you could ask questions. Pretty good. Pretty amazing, right?

1:32:27Aravind Srinivas:Yeah, yeah. How good are they at mathematics? Perfect. Like, how good are they at coding? Way better than people. Yeah. And at a certain point in time, it's going to be interesting to, like, what is education now? Is education just providing you with information? Because that information is readily available. Yeah. Or is education teaching you how to think? Yeah. Teaching you how to pursue your interests and be curious and have intellectual humility and understand what you know and what you don't know. I think that's where it should be. I still think institutions will preserve their brand value because there is a certain aspect of education that's outside of learning.

1:33:03Aravind Srinivas:It's just having access to other curious and intelligent people. Sure. Community. Yeah. And brands attract good communities, peer groups, blah, blah, blah. But the actual process of learning itself has to change. And what you're rewarded for has to change. So fundamentally, everything flows downstream of the incentive. So if the incentive is to score the highest on the exam based on answers, you're not really changing much. You need to change that process. You need to change the process of what do you reward a student? What is A plus or A? Right. That's where we need to start at. We talked about this the other day that the education system in this country was designed to make workers.

1:33:45Exactly. That's what they did when they first started doing it.

1:33:49Aravind Srinivas:Curriculum was designed around that. In India, it's still the case, by the way. Really? Even if you go into a computer science degree, I don't know if it's still the case. I shouldn't misspeak, but at least when I was there and for many years after, the first two years, you just spend learning hardcore electrical and mechanical engineering. You would learn welding using late machines. You would have to go and do workshops, carpentry, a lot of these things. It was fun. I would think there's a lot of value in that. So in hindsight, I actually think it was fun to learn soldering and how to make circuits on red boards and circuit boards.

1:34:27Aravind Srinivas:But if somebody was just interested in just writing code, let's say, back then, all this is kind of pointless to learn, but you had to go through it to be qualified as an engineer. And the reason the curriculum was designed that way is because that's what the labor force was required back then to build like oil factories and like all these things. So you had to learn mechanical engineering, you had to learn fluid mechanics, whatever. But I think that should also change because if the way like you do work changes, then what you're trained for in college should also change. And it's much harder to change these things.

1:35:06Aravind Srinivas:You know, people are much slower. They're scared to do changes. Disruption is always, like, looked down upon. And so I think let's at least start at the incentive structure right from the schools, right from the colleges. Like, let's not, like, reward people based on, like, how much they know. Well, it seems like in the future when things do radically change and it seems like they're inevitable, they're going to radically change. universities and schools are going to be rewarded for having developed thinkers that are able to adapt to this new world that's right yeah so they're going to have to figure out how to adjust their curriculum yeah because the the tools are so spectacular now that just this idea of just memorizing information is it's not that's not what you're going to need yeah to get by in the future It's not.

1:35:56Aravind Srinivas:And I guess like one proxy different schools use is like maybe if more entrepreneurs arise out of your school, you probably create a lot of independent thinkers because they are like willing to take a fresh perspective towards a problem. and build their own thing from scratch. And fundamentally, that's what America has always been about, is the American dream of coming here and having your own idea and still be taken seriously by a bunch of people. The whole idea of venture capital only exists here. Or like family and friends around, this whole idea of just having your friends help you to bootstrap a business and then turning it into a success.

1:36:36Aravind Srinivas:And success doesn't mean like multi-billion or 10 billion or whatever, right? Like as long as it pays you enough that you don't have to work for somebody else and you can live a fulfilling life and you can just go explore your passions, that's success. That's actually a better success than creating a company based on what other people want you to do and then hating your job for it. Yeah, and having a yacht and being miserable and working every day. That's why I said the smartest, the richest people are not always the ones who have the most fulfilling lives. The most curious people have the most fulfilling lives because they have better relationships.

1:37:11Aravind Srinivas:They're actually able to sit and look at something and be curious about it instead of being worried about what's going on. What did the American dream, what was it to you when you weren't in America? What is it like? How is it discussed? Well, to me, like I always thought, America is the only country where you can come here and have an idea and people listen to you and encourage you to go pursue it. The risk-seeking culture is just incredible here. Everywhere else you can are either explicitly or implicitly are forced to defer to authority. Okay, like go and ask the permission of this person. Go and ask the permission of that person.

1:37:54Aravind Srinivas:or get their approval or get their insight. Sure, you can consult everybody out there, but if you have a thought that challenges what they believe in, this country still encourages you to go pursue it. And so, yes, when I came here, obviously Google was the number one company that everybody wanted to work in. But it's also the same country where it allows you, as a new person, to start a new idea the challenge is one of the biggest companies in this own country. And it actually wants it. People actually want new ideas. And then you can consistently see that there are, like, always going to be more and more new ideas and new companies to be created here.

1:38:40Aravind Srinivas:And so that spirit of, like, questioning is encouraged a lot here. And it happens in academic research. I started off as an academic. even there a lot of ideas when I had it and I would share it with people you know people actually give you very honest feedback about things but they don't stop you from working on anything and that's fantastic because that's very fresh it's very liberating and that's not anywhere else I would say it's not it's not in India it's a simplification to say It's not anywhere else, but... It's not as encouraged. It's not as encouraged. The incentive structures are not quite there.

1:39:25Aravind Srinivas:And the ability to be taken seriously for some crazy ideas is why America is still at the top. But it's crazy to me that if the American dream is so compelling and so many people come here for it, why doesn't the rest of the world sort of adopt those values? It's hard. you know like a lot of it is cultural like america was was was born was made from from like you know a piece of land essentially right uh and uh a lot of ideas that we built here a lot of industries that we built here were all like created here from nothing and that required you to like go take bold risks I think Jeff Bezos said this in some podcast that where else would you like be able to go raise like a few million dollars for an idea that has like 5 to 10 % chance of working and then fail at it and still go and raise another few million dollars for your next idea nowhere else people are willing to like people who get rich here actually want to encourage and be part of somebody else's crazy journey because it's hard to pursue all crazy bets yourself.

1:40:42Aravind Srinivas:So it's an ecosystem. And once something becomes an ecosystem, there's network effects. So it's very hard to copy that elsewhere. And so your value is measured in your curiosity and your willingness to work on whatever it is that is your pursuit. Yes. And then eventually adjusting and learning and catching fire with one of them. Correct. And you have to work hard. Like, you know, like, I'm a big believer in intense hard work. I think nothing great can be accomplished by being soft. And so all this, like, recent push for, you know, having a lot of work-life balance, this and that. Sure, if you have work-life balance, if that's what you want, then I think there are certain jobs that will give you that.

1:41:26Aravind Srinivas:But when you're trying to do something from scratch, when you're trying to create something from nothing, it's not meant to be easy. Right. There are some sacrifices that have to be made, and you're signing up to be part of that experience, that surreal joy you get from doing something that's felt almost impossible to achieve. And you're not staying up late or waking up early because you're getting paid more. Maybe you might not get paid anything. Maybe this whole thing goes to nothing. But that experience you're getting of being part of something that feels very hard to achieve is what you're signing up for to be part of.

1:42:04Yeah, and if you're not, find something else. It's fine. There's nothing wrong with that. Exactly.

1:42:09Aravind Srinivas:And the country has enough jobs to provide for all kinds of needs, right? And everybody goes through different phases in their life. Sometimes they feel a little lazy or disillusioned. And so what I like about this country is that there's a lot of curious people here. There's so many different people, whether they use AIs or not AIs, They're all like finding meaning in like so many interesting projects. Well, obviously, I don't know any other country really because I was born here. But the people that do talk to me about what the American dream is like from another country, they're the most passionate and the most supportive of this idea, this experiment in self-government.

1:42:52And just the whole idea that the country operates on that anybody can chase their dream. You can. If you have a dream and you're willing to work hard, you could actually do it in this country. That's right. Yeah. That's, you know, it's most the people that are most passionate about that idea oftentimes are people that come from somewhere else where that wasn't available. And it's not just like, you know, people coming from one particular country or another.

1:43:19Aravind Srinivas:It's the attitude. It's the way the system works and rewards you to like be bold and take bets against established players. It's okay, right? It's okay to, like, be an upstart, a challenger. And people love that, like, underdog. And I think, you know, that's fantastic. And that culture is continuing. Yes, there are, like, multi-trillion dollar companies here, and they're all going to become even bigger. But people still want the young, hungry person to also be successful. Yeah. Well, they love disruptors. Yeah. And people love underdogs in this country. Yeah. Yeah. It's universal. It's not specific to technology.

1:43:58Aravind Srinivas:Right. Like, I'm sure everybody would love an underdog story that wants to go against, like, Coca-Cola or Pepsi or something, too. Sure. Yeah. Oh, in sports. It's our favorite thing. In sports, yeah. We don't like when the guy who's supposed to win wins. We love when the guy who's not supposed to win triumphs. Yeah. Yeah. The underdog story. Yeah. That's a very uniquely American story. To me, that's what this country is. I mean, sure, there's a lot of obstacles and challenges. Just like every other country, there are things here that are challenging, but it's one thing that has consistently stayed true.

1:44:33One of the big fears that people in America have about technology in particular is that without being aware that this was going to take place, everybody gave up their data. Everybody gave up their data and didn't recognize it was a commodity. That in turn made these corporations immensely wealthy and powerful. And then they have the ability to shape narratives. And that concerns people because using their ideological position as leverage to try to push that through technology that has immense control and influence over people. and that we didn't see technology and corporations as having that much control over how society views itself and how we interact with each other.

1:45:24And there's a real concern that these companies got so big. There's a guy named Robert Epstein who's done a lot of work on curated search engine results and how much that can affect. Have you seen any of his stuff? I think I've seen this, yeah. How much that can affect elections, how much that can affect people's perceptions on any societal issue that's coming up. And it's concerning. It really is because they do curate search results. It's not simply you just run it out there and you get this is the data. No, you get – if you look for specific political figures, depending upon where they fall in the right or left spectrum and depending upon which way the company forms – the corporation forms – falls rather, you'll get different results.

1:46:13And that sucks. You know, that's very concerning that people don't recognize, they don't have the ability to see how that is dangerous for all of society to have that kind of power and wield it in that way where you're not being honest about accurate objective information. You're pushing particular ideologies. Yeah.

1:46:36Aravind Srinivas:So I think it's kind of like this is almost an effect of the asymmetry that exists between the amount of AI power that centralized systems and centralized companies have and the amount of AI power as you as a sovereign individual has. so when you don't have the AIs to just go judge for yourself like what you should be reading and fed you're obviously like under the influence of what you know whatever big tech company is controlling the information for but when you have access to all those AIs you can actually just customize what you want to see by telling the AI like hey this is what I think you should actually question and tell me.

1:47:19Aravind Srinivas:Until now, you never had that power for yourself. You're finally getting it. And eventually, we'll be able to have our own LLMs, like our own models that we would be able to post in our own hardware. We don't have to rely on like one centralized model given to us by like any specific model company. And using that, you can shape it to your beliefs, your custom data. So when you're consuming a search result, you can actually task that AI that you control and you run so nobody can shut off access to it to tell you, hey, can you actually give me a contrarian perspective on this? Or can you tell me if these search results are actually biased?

1:48:04Aravind Srinivas:So I think we need to give individuals more sovereignty with more access to their own AIs that they own and run on a piece of hardware they own themselves. And this is going to be leading to the whole rise of local AIs. So as AI models, like today, they're very power inefficient. They're running on large data centers. But in a year or two from now, whatever capability that exists in the most power-hungry data centers, it'll be possible to run it in some box that you own. Really? Yeah. It's already happening. It's already happening that there are interesting hardware projects like the Apple Mac Mini, NVIDIA DGX, where you can actually host a reasonable-sized model and put it in a box and have it run, and you don't have to pay for all the tokens it produces you.

1:48:56Aravind Srinivas:You just have to plug it into your power core, and it works. I know Duncan, my friend Duncan Trussell, he does that. Yeah. And today the capability of that model that can run locally is not quite there. So you would still prefer to use something that runs in the data center. But eventually this is going to be a spectrum. There's going to be some percentage of tasks that you would start delegating to this local system. It'll be a hybrid model. And over time it could end up being the case that you could buy something that feels like a refrigerator for your home, which is your own AI box, and host a model that you control.

1:49:31Aravind Srinivas:So nobody can arbitrarily shut off access to it one day. And then you can have that be your weapon against what the big tech wants you to be fed or believe in. This is the only way we can fight this because they have far more computing power, far more data, far more algorithms than you. So the only way you can fight that is you have something you own yourself. and with the rise of open source models, open source LLMs, and progress in local hardware, and both Apple and NVIDIA and Intel, they're all doing amazing work here. You could potentially change the future and give people more power, and this may not be as expensive as people think.

1:50:12Well, that's a good solution because I've always wondered, like are these searches like using Google, is that going to be irrelevant one day? because you already can just ask your phone. Most of the time, if I want to have an answer for something, I just ask perplexity. I was like, what is it? Instead of having to sift through all these Google searches and try to figure out what it's showing me first and get to page three where it's what I really want to know, I can get the accurate information, then follow-up questions are instantaneous.

1:50:44Aravind Srinivas:Yeah, and even the models that are running the perplexity app today, they're all in the cloud. eventually you'll be able to do that on a box that you own you can still use the front end the UI of the app but you can control the compute that runs on a piece of hardware you may ask why do I care what if someday the data center gets taken off Iran was bombing data centers what if someday the government decides that model is no longer available you you want some control over like like what models you can run and like you can you may even want to shape it to like your context that you never want to be living on any data center and and and and uh i think that's where i believe the individual gets more sovereignty against big tech and um that's how like we fight uh surveillance or like centralization of power yeah and certainly pushing narratives um what do you think happens with social media because social media and as you were talking about before like algorithms like it's one of the biggest problems in terms of the way people view the world yeah i'm curious what you think like you know like my my opinion is that it's not good for the kids it's terrible for them yeah but i think they should have some exposure to it because i think it's good to know that it's a thing and i think children are fairly resilient and they learn but the anxiety levels of kids is much higher than ever before yeah suicidal ideation's higher self-harm yeah yeah i'm a little my belief is that um when you're just fed of feed and and and the algorithm of the social media company decides what you're going to see next it it curbs your curiosity and i i don't i don't think things that curb human curiosity should be encouraged.

1:52:41Aravind Srinivas:Yeah, I agree. And so if the app is designed in a way where it asks you what you're interested in and helps you to come up and find things that are very related to what you're interested in, that's awesome. But that's not how it works. It's literally like it starts with something. You start doing scrolling and then start showing you what you just scrolled. And then you end up in an echo chamber. And that's not necessarily good. Well, you can get trapped. I'm in a trap of schizophrenics lately on Instagram. It's just mostly schizophrenics. Like people that tell me they're the rightful president of the United States and tell the guy hasn't showered in days.

1:53:21And if you have a phone, you can create an account and you just start uploading nonsense. And then for whatever reason, I've watched a couple of them. So now they just keep showing them to me.

1:53:30Aravind Srinivas:And it's full of AI slop right now. Like a lot of AI slop. Like it's not even clear. and it's not labeled also clearly whether it's been made with AI or not. So often, so essentially it's leading to a complete loss in trust where when I see something, I don't even know if it's real anymore. Right. And it's going to get worse. Yeah, it's going to get worse. To the extent that you're going to like, your default would be that this is AI. And then like you're going to have to go through multiple layers to finally verify if it was real. and even like verified accounts post a lot of AI stuff so it's not about like whether the account is verified by Meta or whatever, right?

1:54:12Aravind Srinivas:So I think fundamentally I feel like, okay, the way I think about it is what are pieces of technology if did not exist would be a really bad thing for the world and what are pieces of technology if did not exist wouldn't even matter and and i feel like social media is more towards a second yeah like you know uh searching for information and answering questions and like getting you know ais to like do things for you uh help you learn new things faster all that stuff is some we need more of that but um because it supercharges our curiosity whereas like brain rot feeds with ai slop doesn't actually supercharge our curiosity.

1:54:57Aravind Srinivas:It actually curbs our curiosity. And so if we believe in the curiosity premium idea, we need to encourage things that supercharge our curiosity and discourage things that curb our curiosity. Do you anticipate a time where we recognize the dangers of algorithms and there is some discussion to either curb them or allow people to have control over them in a real meaningful way? Like you could dictate maybe through AI even that there's an AI interface to your algorithm that understands your particular emotional needs, your curiosity. Like only show me this. This is what I'm interested in. Carpentry and basketball games.

1:55:39Show me those things. I don't want to see who's getting divorced. I don't give a fuck about this.

1:55:46Aravind Srinivas:So here's the thing. You can still customize on most of the social apps. It'll be deeply buried somewhere in the settings somewhere, and you can go and say stuff. But the reason it's buried is because once you always have to say it or it's the starting entry point for your experience there, your engagement time would go down because once you consume the content that you really want, you would go back to your work, which is what you really need to be doing. But that doesn't help them sell more ads. And so the incentives are not aligned. And so Elon has this really good metric he talks about where it's like a total amount of unregretted minutes spent on the app should go up.

1:56:29It's hard to measure.

1:56:30Aravind Srinivas:It's hard to measure. It's more like in spirit the right metric. but this metric is also why it's hard to make money on ads if you care about this metric which is why x doesn't really make a lot of money on ads compared to you know instagram or youtube because uh you're kind of like optimizing for interestingness like doesn't mean x has everything right there's a lot of chaos there's a lot of memes there's a lot of like um weird shit going on there as well but uh in general social media is not necessarily like great for people i think is terrible for people. But it also provides you with a way better understanding of what's going on in the world than has ever existed before.

1:57:14Aravind Srinivas:X particularly. X particularly. Because it's a place for discourse. It's a text-based app more than a video-based app. Right. So naturally, people tend to engage in discussions and debates. And there's a lot of curious debates going on there and a lot of interesting viewpoints expressed by people. So I think in terms of the unregretted minutes, it's actually one of the better social media apps. But apps that are purely based on, like, video or images, and largely video these days, I think that's just trying to get your eyeballs in time. Yeah, those are the mind numbers. Yeah. They just numb your mind.

1:57:54Aravind Srinivas:I mean, it's depressing when you go to a metro and you just see people just scrolling through their feed. Nobody. Everybody doing it. You look at the entire car. Everyone's doing it. It's just insane. Yeah. It's weird. Yeah. I always say that if there was a drug that existed that made people stare at their hand for six hours a day, everybody would be like, get that out of here. Yeah. But that's essentially what we're doing. Yeah. Because, like, most of what people are looking at most of the time, they don't even remember. Yeah. They're just scrolling through this thing. It's brain rot. It's just brain rot.

1:58:21Aravind Srinivas:It curbs your curiosity. Yeah. I mean, Apple has these settings in different apps. Have you tried this? What? Where you can set the timer for every app? No. I just use discipline. I don't engage very much anymore. I dip my toe into X every day for a few seconds. I go, what's everybody mad at? What's going on? Who stole this? How much corruption's here? Who got killed there? Okay, bye. And then I just check out. I don't want to do it. And Instagram to me is just nonsense. I just look at that every now and then for nonsense and occasionally something interesting. Really, YouTube is my main go-to thing because YouTube is my most unregretted minutes.

1:59:06YouTube for me is always interesting. There's always like some cool thing on cosmology. There's some – I watch fights on YouTube. I watch professional pool matches. That's what I do for the most part. That's where I really like find my actual interests and fulfill my curiosity.

1:59:25Aravind Srinivas:Long-form content is what human mind should be trained to consume more of. whether it's books, whether it's like 30-minute videos explaining something. And you need to train your mind to actually complete it. That's actually the biggest problem of the younger generation. The more they're in the reels experience, short-form video, they're unable to actually complete long videos anymore. That's true, but also at the same time, the rise of podcasts is happening. Yeah, and it's great. It's great. So it's not universal. It's like there's a lot of people that don't find fulfillment and all the doomscrolling and all the nonsense.

2:00:03Aravind Srinivas:They really do want. Yeah. I'm particularly just focused on the younger generation. I'm sure like people like us can adapt to like, okay, let's say maybe we have a temporary addiction to social apps and we can. But a lot of the young people are the people like I meet kids like at the mall that are 11 that listen to my podcast. Really? Yeah. Wow. I know. It's nuts. They go, I love your podcast. I'm like, who lets you listen? Get out of here. no i'm always joking around about it that's really cool but no there's a lot of like particularly like young boys that come up to me all the time that are interested in it that's amazing i love it i love it because then they're going to get exposed to some interesting ideas and yeah it'll also encourage them to have those kind of conversations with each other right yeah whose whose podcast do you listen to i love tim dylan's he's probably my favorite because it's the most accurate and also satirical and hilarious view on everything that's going on in the world in terms of like war and world news and culture shit and he's my favorite he was just on here yesterday i fucking love that guy to death he's so funny he's so crazy it's like his mind works in such a unique way and it's developed because his podcast is different where he very rarely has guests so most of the time it's just him ranting and his producer laughing and he's the best ranter that's ever lived i don't think there's anybody that's even close he's the goat like there's like i don't think there's any argument every comedian agrees like as far as like just the ability to just sit in front of a microphone and rant like bill burr does it well he's good at it there's a few other guys that are good at it no one's as good at it as tim he's the most consistently entertaining and then for just mind not mindless but like to escape i listened to a lot of archery shows and hunting shows where they're talking about different tactics and hunting or different techniques and archery new equipment and new innovations archery is an interesting thing because every year bow manufacturers make a better bow and like tiny little engineering changes of these bows like it's a weapon that's been around for who knows how many thousands of years but what the and you're able to feel those improvements oh yeah yeah you feel the difference every year Hoyt puts out a new bow and every year I'm like motherfucker they did it again it's better so just tiny changes less vibrations in the hand more balance in the shot, you know, more forgiving in terms of accuracy.

2:02:44I love that stuff. So I get really fascinated by engineering, really fascinated by automotive engineering. I'm really interested in, like, that's another thing where, like, every year people figure out to make a car that can hold more Gs on a skid pad, that can get around the track quicker. Like, every year they're battling to see who can get around the Nurburgring quicker. And what are they doing? They're adding horsepower, increasing suspension travel and suspension tuning rather and making them more compliant or making them stiffer and making them more adjustable and then like tire compounds.

2:03:18And I'm just interested in anything that where someone's working on something and getting better at something or getting new information. I love history podcasts. I listen to a bunch of history podcasts. So that's most of the time. If I'm listening to something, I either want to be entertained or want to be educated. Educational. Yeah.

2:03:38Aravind Srinivas:Yeah. And that's entertaining. Yeah. What about you? What kind of stuff do you listen to? I mean, I listen to your stuff. I listen to Lex. There's this guy. I mean, you had him on Rick Rubin, of course. Sure. Yeah. Love that guy. Yeah. He's awesome. I listen to his stuff. and I mean I also watch like some interesting videos about you know concepts I don't understand there is this YouTube channel Veritasium you should check it out what is it called? Veritasium how do you spell that? V-E-R-I-T-A-S-E-U-M Veritasium what does it mean? I think is that someone's name? no Veritas just means like seeking truth kind of thing oh okay Is it this channel?

2:04:25Yeah, yeah, yeah.

2:04:26Aravind Srinivas:Yeah, yeah, yeah. Ooh, okay. 20.9 million subscribers. Yeah, that's pretty good. Obviously, a lot of people agree. So they make all these very interesting videos about, like, stuff that, you know, you would be curious about, but you never actually bother to ask that or learn more about and explain some of the most under-understood companies or, like, phenomena. And I just love watching it. This is kind of like my idea of doom scrolling. I like watching 20 videos at once. Yeah, I am going to subscribe to it right now. It's pretty cool.

2:05:05Veritasium. There it is. Got it. Subscribed. Bam.

2:05:10Aravind Srinivas:And it explains all these fun concepts that are, you know, you take it for granted. Like, okay, why is Google Maps really fast? Like, okay, it'll tell you what's going on, and how the data is used across so many different people at once in all these different departments. CIA's new tech doesn't make sense. Exactly. We were just talking about that yesterday. We were doubting it. You know the heart murmur thing? Do you know about that? No. So the pilots that were downed in Iran, they said that they have this technology that allows them, I think they could use it up to 70 miles, and they could detect a very unique heart rate.

2:05:49Like your heart rate is different than my heart rate. They could know it's you. You could be hiding in the mountains and they could find you from 70 miles away with this technology.

2:05:58Aravind Srinivas:Wow. A lot of people are like – Beams or waves or something? Well, it's called – what is it called? Quantum magnetometry. Is that what they call it? I think that's what it was. Remember we looked it up yesterday? They're using the word quantum and not explaining what they're doing, like how they're doing it. And you're like, okay, is that real? Or is this some invented horseshit to cover the fact that they have some very sophisticated satellite imagery where they can have a detailed map of literally the entire surface of the world? They know exactly where people are, but they don't want our enemies to know that they have this capability.

2:06:34So they're making up something.

2:06:36Aravind Srinivas:I see. That was my suggestion yesterday. That, like, maybe they're full of shit. Because the whole thing seems nuts. What is it called? You got it. It's quantum magnetometry? Sure. What does that mean? You tell me. I don't know. Exactly. Yeah. So this guy, he's saying it doesn't make sense. Yeah. And a lot of people say it doesn't make sense. Like it doesn't seem to vibe with anything that we know that we can do. Quantum magnetometry. Yeah. First time I'm hearing it. Pull up this description, the official description of what this stuff is capable of. So this is supposedly some very advanced CIA tech that allowed them to locate this downed pilot.

2:07:16Maybe. Or maybe there's something else going on. Or maybe there's some other methods that they use that they don't want the enemy to know about. Maybe some beacon these guys have on them.

2:07:27Aravind Srinivas:Yeah, I guess what's the incentive for CIA to actually describe how their technology works? Yeah, zero. Why would they tell you that? Why would they tell you they even have that? That's crazy. Yeah. And then Jamie had a good point. But the capability is insane. Detecting your heart rate 70 miles away is just insane. How? Yeah, how? And when they throw the word quantum in things, I always go, hmm. Hey, what happened with that White House announcement? Sorry, I keep... The quantum computing? Yeah, remember? There's Q News coming soon, and then at the bottom of the text, the Q sounds for quantum.

2:08:01Aravind Srinivas:Oh, I see. Is that what it is? I thought they just announced a bunch of investments in a bunch of quantum companies. Maybe that's it. IBM was getting some funding or whatever. So this quantum magnetometry, can you pull up a description of what it is? Sorry, I started looking up the... Sorry. I was asking you too many questions at the same time. Quantum sensor help rescuers. Yeah. So this is it. Ghost murmur. Yes, that's what it's called. Purported surveillance technology utilizes long-range quantum magnetometry. What is that? quantum magnetometers measure extremely faint magnetic fields, including the body's natural electromagnetic signals, by tracking changes in the energy states of atoms or subatomic particles.

2:08:47What? Technology reportedly uses microscopic defects in synthetic diamonds. When illuminated by a laser, these centers are hypersensitive to tiny magnetic fluctuations. The heart signal, while human heartbeats produce a magnetic field, is extremely weak, around 50 to 100 picoteslas, and typically degrades over very short distances. So the Ghost Murmur deployment, they reportedly used Ghost Murmur during a mission in southern Iran to pinpoint the location of a downed American airman hiding in dense mountainous terrain. By mounting these quantum sensors into a helicopter, the system purportedly registered the pilot's heartbeat from afar.

2:09:32okay does that sound like more shit i mean not it doesn't sound full of shit but like

2:09:38Aravind Srinivas:basically the part that sounds surprising to me is how they're able to deal with all this like distance and attenuation across the distance right and all this interference and they claim to use ai for that but nothing is really described on how they use it right so if they're not describing how they use it why are they even telling us they have it why so like there's a lot of skepticism Yeah. Laws of physics. Physicists point out that the heart's magnetic field is a million times weaker than the Earth's. Detecting it at a range of miles rather than centimeters defies currently published peer-reviewed physics.

2:10:14Alternative explanations. Suspect that while quantum sensors were likely on board, they were probably tracking the radio waves of a survival beacon. the metal in the pilot's equipment or using traditional thermal, infrared, and radar capabilities rather than detecting a raw heartbeat via magnetic fields. I do remember seeing a different part of a... When that story happened back in April, someone did report on one of the military websites that there was a survival beacon that they used to track them. And that the whole quantum armor stuff was nonsense. Yeah, I saw that too. But no one wants to report that because it's not fun.

2:10:49Right. No, the ghost murmur thing is awesome fun. And if that is real, like, boy. And you can imagine a world 100 years from now where that is real. So it's exciting.

2:11:00Aravind Srinivas:Oh, yeah. 100 years is a long time for this to be real. Yeah. 100 years, they probably got it down pat. Then that's the problem. You can't hide from the robot dogs from Black Mare. Yeah. You know? Yeah. Do you ever, while you're working in AI, do you ever wonder, like, is this the downfall of humanity? Is this a good thing to be worked on? Do you ever have, like, doom moments? Not on specific things I'm working on, but in general, I do worry about, like, how much, you know, you obviously want to, like, stay in charge and, you know, be in control of your experience. Still be the one driving change.

2:11:43Aravind Srinivas:and have a lot of agency for yourself. So I do worry that, like, it's all about, like, making sure everybody's upskilled and understanding, like, where the future is headed and not being, like, fed only, like, dangerous apocalyptic messages. And because it's very essential that human beings retain their agency and staying curious, right? Like, so if that stops being the case, if you start subscribing to the vision that, okay, your jobs are done, you don't really have any meaning in the world, and we'll pay you some dividends and you just sit at home and chill. That is not a good thing. And I feel like there are not enough voices in AI that are actually saying anything different to that.

2:12:23Aravind Srinivas:And I like when Jensen was here, I think he was a little different. I think he tried to give a more positive version where he said, okay, like the radiologist thing. Okay, if all radiologists can take away, they start doing different kind of work. so I think we need to start looking at like okay like okay first of all guys relax you have a lot you have one premium skill your curiosity so let's figure out ways to channelize that let's change the way work is done in companies let's change the way educational institutions run let's change the incentive structures and let's help you build new ideas and new companies and explore things that are not even being considered and the government should obviously support all these initiatives.

2:13:07Aravind Srinivas:So that's what needs to happen more. But what's happening actually right now is, okay, hey guys, you're all losers. You're going to lose your jobs. And don't blame me because I told you so. And still give us money because we're still going to do it anyway. So that's what's happening more and I think we should stop doing that. That's my opinion. Well, the problem it's kind of a self-fulfilling prophecy and if you tell people that they're going to be a loser and you're going to yeah their life is over they're going to think that way instead of giving them an understanding of like look this can open up new doors for you yeah this can and anytime there's any sort of disruptive technology there's always the the fear that it's going to go badly yeah this was the case with the locomotive this was the case with uh when the printing press was invented yeah by the way like i i did some research on this where um and the industrial revolution happened um people got new ideas um okay like for example uh when the industrial revolution happened um who came up with the idea of a steel plow john deere until then we were using wooden plows to like for farming no farmer complained that hey like we need fewer farmers now because Steelflower is able to do it more effectively.

2:14:27Aravind Srinivas:No one complained. You actually had more farms and more productivity, more crop yields, and you're happier. But isn't that just a regular tool as opposed to AI? Sure. AI is different. It's not overnight going to become something that's capable of just running an entire multi-trillion dollar company on its own. There are a lot of things that AIs cannot do. There's a lot of tacit knowledge in every company that AIs don't quite understand. and there's a lot of new directions that you can just start working on that AIs are not well-equipped to do because it doesn't have full knowledge about it, and the knowledge about it is yet to be captured, and some of that requires human-to-human work and collaboration.

2:15:07Aravind Srinivas:So we obviously have to gravitate towards what is scarce. When AI makes the current labor that's considered scarce because that's where the money is going in commodity, then we have to gravitate towards what is scarce. and the only way to do that is to seek things that we don't know about, which is only something we can discover through our curiosity. There's nothing else. Whatever we don't quite understand well, whatever we don't know how to do well yet, even with the current capabilities of AI, that's where we should pull our labor and workforce into. So it needs more responsible messaging and that's not quite happening right now.

2:15:48I think it needs responsible messaging And then in the future, what it needs is like real direction in terms of like letting people find their curiosity and find these paths of interest and find something to do with themselves. It doesn't involve whatever their previous occupation that's irrelevant now. That's true.

2:16:12Aravind Srinivas:I think like passion for people is something that not a lot of people would be able to answer out of the box. Like if you go and ask them, what is your real passion? And the only thing they've known in life is to just climb up career ladders and make more money. That's going to actually take them a while to even discover. Right. Which is why it's so important to get kids off on the right start. Yeah. That's the hope. That's our hope for the future is the kids. The kids are born curious. They don't need to change themselves to be curious. Right. The adults who probably already are like, because of this knowledge work thing, who kind of curb their curiosity and try to fit into the existing system, it might be a little hard for them to adapt.

2:16:55Aravind Srinivas:But the kids, I think they don't have this problem. So I'm actually optimistic about the future long term because the future is all centered around whoever is very young today. What do you think about this idea that universal basic income is going to be required? Some form of it is good. So it's like a dividend. I almost think of it as a dividend. If a lot of spend that most companies are currently doing today on like payroll, which is paying a knowledge worker for a certain task. Think of knowledge work as basically taking information and transforming it into an artifact. Right. And it's messy and complicated.

2:17:34Aravind Srinivas:Let's assume that's being done by AIs. So obviously, companies will start spending more on compute instead of payroll. It's just a reallocation of, like, spend, a budget. Similar to, like, what happened in advertising industries where most of your advertising budgets went to, like, television and, like, billboards, and then now it's starting to go to Google and Instagram and YouTube and all that. So when that happens, obviously, like, the AI companies are going to make a lot of money, and people who helped be part of creating it, either directly or indirectly, would want to have some role to play in that ecosystem.

2:18:14Aravind Srinivas:And a good way to involve them is through giving them some ownership in the company. So as shareholders, if you get dividends from the profits generated by the AIs, it's not a bad thing. But that shouldn't be the only thing. Right, so this is similar to like people that live in Alaska, they get a check. Correct. Alaska does this, and it's not a bad thing. As long as they are doing some other things. Right. alongside. It could lessen the burden. Correct. Yeah. Yeah. And if people are interested in still being part of the AI industries, they go do things that AIs are not able to do today. And that's been the case before.

2:18:53Aravind Srinivas:Like when Industrial Revolution started, the United Kingdom actually started like projects around building railroads. And that gave a lot of people who are in the cottage industries new jobs. So there are going to be a lot of new projects to just, okay like what if we want to reimagine the government itself where the government runs largely on ai yeah that was my next question yeah so then we need people for that yeah because this is a legacy industry it's not it's not about the capabilities not being there it's about working through the legacy and bureaucracy to like actually deploy and implement this inside the most like like largest institutions in the country and that's going to need a new set of skilled workers to go do that.

2:19:41Aravind Srinivas:So some people who might be working at Microsoft or something today might actually end up working for the United States government because Microsoft may not need them, especially for, like, you know, internally deploying AI or selling AI to their customers, but the government needs them. And if the government can pay them well and it's a fulfilling job to find some meaning for, like, doing something good for the country, it's not a bad thing. So I think, like, just like in the Industrial Revolution where we had new projects because the demand for AI was so big, we're going to start seeing some new projects being created in AI as well when the capabilities advance enough that they can replace knowledge workers.

2:20:17That's the rosy scenario.

2:20:20Aravind Srinivas:It's not as rosy. Like, the real world is messy. A lot of things are still done through trusting other human beings. Nobody is, like, blindly trusting AI. AI still makes a lot of mistakes. I know a lot of people are hesitant to the idea of AI running government and I get it. But also, look at what the people are doing. Look at how much corruption there is, how much fraud and waste. Imagine if all fraud, waste, and corruption was instantaneously eliminated. Yeah. I mean, that was what Elon tried to do with Doge, right? And then I think the bottleneck there was just discovering how slow it is to do things.

2:20:57Aravind Srinivas:He's not used to running that slow. Yeah. Also, how much resistance. Yeah, resistance. Because there was so much grift. Correct. Yeah. So honestly, like more than AI, the government is running a lot of legacy software stack because a lot of these legacy enterprise companies just have created these multi-decade or like year contracts that are hard to get out of. And the way they do that is to sell it at a much larger discount. And, you know, like if you're on like a specific OS, you're not allowed to change this for like 10 years. You have to use the same sort of software. where all these people you hired only know to use that tool.

2:21:37Aravind Srinivas:So it takes time to actually change and implement new things, leave alone AI. If you just wanted to move everybody from Windows machines to Mac machines, good luck with that. It's going to take a lot of time. That's the state of the system. And so that has nothing to do with technology. And so to do things in such messy systems, you still need people. still need people to navigate all these changes it's not about the capability of technology it's more about how the system is structured and that's why I still feel there will be new jobs maybe there's a lot of new projects to be done maybe some good leader actually wants to change the system and is willing to be patient about it over a 5 to 10 year horizon if you take 10 years to run the majority of the government processes on AIs it may seem slow to you today but in the grand scheme of things it's actually good for the country and that's still going to need a lot of nice engineers to go work on these projects so they're not going to lose all their jobs there's going to be some displacement there's going to be some new projects there's going to be new priorities but it'll keep going the system will keep going because that's just how historically things have been when you think about the future of AI and you think of this so what do you think about AGI in particular?

2:23:01You think about something that could potentially make better versions of itself self-replicating and then how far does it go?

2:23:10Aravind Srinivas:Yeah, so that is the ultimate form of I think some people in Silicon Valley have started calling that as ASI, so when you see the word ASI being thrown around people kind of think of ASI as an AGI that can recursively self-improve itself. So that's going to be no limits to how smart it can get. And I used to think that ASI is bottlenecked by power because you need a ton of compute for this model to keep on training itself and running its own rollouts and collecting data and then going and updating itself. But you could imagine that once the algorithm is correct, the ASI could be tasked with just making itself more efficient too, where improvement doesn't just mean capability improvement.

2:24:00Aravind Srinivas:Improvement could also mean power efficiency. And that way the recursive safe ASI that is improving itself also makes itself more compact and more efficient, and it can run on less compute. So that would be the ultimate project in AI. Think of it as almost as the last project in AI is basically cracking recursive self-improvement. Once you crack that, you don't have anything else to work on. in practice I think what's going to happen is because information is so muddled and fragmented and living in disjoint systems just the way we have constructed our messy real world it's going to be hard to point even a recursively self-improving AI at some metric and say go improve this or go reduce inflation by 5 % that would be awesome if you can task an AI to do that if that's the job of the government to just reduce inflation have a deflationary effect on society and make goods and services a lot more abundant and efficient, it's going to have to deal with a lot of messy legacy systems.

2:25:01Aravind Srinivas:If the task is to go improve the health care, we're good luck. Who's going to deal with all the compliance of actually implementing these changes inside hospitals? Most hospitals are still using legacy software because the software provider has lobbied the government in a way where only they're allowed to do that. God, what a stupid bottleneck. Exactly. So a lot of the bottlenecks in actually having AIs just take over and massively improve the human society and our hospitals, our legal systems, our government systems, where most of the payroll is going into, is just bottlenecked by a lot of compliance and regulation.

2:25:41Aravind Srinivas:And so that's why I feel the human beings are still necessary to effect the change because these laws and regulations were built for us. It also seems like we have to demand that those systems be usurped. Sure, 100%. And we need the help of AIs to rewrite all these laws. It's going to be humanly impossible to go and change one specific line here and there. Right, and then you're going to have a bunch of these software companies that are lobbying to try to stop that from happening. That's why this messiness and this need for getting all people on the same page and actually steering the society in a positive way our jobs will probably be more steered towards that problem solving at a different level of abstraction maybe more need for eq more need for actually like understanding differences of opinion and still like a leadership quality ability to understand people and ability to convince people these these are the skills that and will be even more important in the world where like Actual work can be done by AIs, but effecting the change in our society, in our country, still needs human beings because the systems are messy.

2:26:59It's a weird world we're in right now. Yeah. It's never been weirder.

2:27:03Aravind Srinivas:That said, there's a lot of things that can still go wrong when you give so much power to specific companies. And they deploy all these bots. And then anybody can use them in weird ways. You don't even know if you're talking to a real person anymore. They're like people who just run AI responses and chat with like 500 people at once. And that's like a whole business. and so um i think it's gonna it's gonna take a lot of adjustment well the another piece of adjustment that a lot of people are coming to grips with is that this is a new part of our conversation and that in 2020 like when i first moved here ai was never discussed yeah it was not a thing yeah i mean we knew about it we knew about ai but it wasn't like you it wasn't a huge part of the cultural discussion of what the future holds for us.

2:27:58Now it is.

2:27:59Aravind Srinivas:It is central, yeah. And in that short amount of time, in just six years, it really makes you wonder, because we know how technology progresses exponentially, what it's going to look like six years from now. Yeah, the 2028, like, you're definitely, my prediction is 2028 election debates are going to be largely about AI. Wow. Yeah. AI, energy crisis, power, power. People are going to care about all these things. Because AI is no longer a thing that is new. It's part of all our lives. Everyone's using some form of AI in some ways. And it's not as dangerous as people thought. It's an amazing tool for doing work and asking questions and learning things and all these things.

2:28:47When used correctly.

2:28:48Aravind Srinivas:Yeah. It can also be used incorrectly. Like everything. Like everything. So it's far more powerful that incorrect usage can cause serious damage. Like, for example, kids who are using AIs for companionship. Right. Crazy things are happening there. Crazy things are happening there. Not good. Yeah. It's as dangerous or probably more dangerous than social media. and it's also scary that social media companies want to build more of these kind of like companionship apps because they know that okay their only job is to get you engaged more and that's the only way to sell more ads and make more money and clearly companionship is a way to get you engaged more yeah and so that's dangerous if ads start being part of like AI chats.

2:29:39Yeah.

2:29:39Aravind Srinivas:Because then if that ends up working, then all these chatbots are just going to be sycophants that just tell you stuff that you want to hear. It's also an indistinguishable facsimile to a real person. Like they communicate like a real person. Right. So you really think you have a relationship with this. Right. And it truly screws with your mind. And it's hard to decouple and it takes a lot of time to recover if you want to unplug. And so the business model incentives are not well aligned to humanity. Did you see that AI companion that they developed that was at the Consumer Electronics Show in Vegas this year?

2:30:22Which one? It's like a hot Asian lady. I see. Yeah.

2:30:27Aravind Srinivas:These are the weird kind of projects that are going on. Yeah. It's a hot Asian lady that talks to you. Yeah. And, you know, she talks to you through AI. And right now it's just a kind of a crude sort of robot. But you can see where it's going. You can see where it's going. Ex Machina. Yeah. It's going right there. Yeah. Yeah, that movie was quite far ahead of its time. Really? Yeah. That's one of my top ten favorite movies of all time. It's underrated, actually, because reviews online say it's not as good. But I liked it. I loved it. Yeah. I thought it was fantastic. I like it better than her.

2:31:03yeah her I lost her after a while I shut it off yeah it lost my attention I'm sure it's good it was the wrong time for me to watch it yeah but Ex Machina I've seen it like five times I fucking love that movie yeah it's just so I don't want to give anything away but it's it's so incredible and so bleak and so yeah in the relationship that he has with the hot one yeah

2:31:26Aravind Srinivas:Yeah. You believe it. You're like, I'd be right there with them. You know, it's too confusing to our system to have something that looks exactly like the thing that you desire that is actually interested in you. It just happens to be. Because it pulls all your data. Yeah. Knows too much about you. Knows how to pull your strings. Yeah. Yeah. But listen, man, very fascinating discussion. Yeah. I'm glad we did it. Thank you very much. Really enjoyed it. Thank you so much. And thanks for having an awesome platform. Poplexity has been great. We really love using it here at the show. It's made the show more interesting.

2:31:56It's cool.

2:31:57Aravind Srinivas:Thank you. It's very fulfilling because we want the app to be used by curious people. We want to lift the ceiling of what our population can be. Not everyone is fully curious all the time, but we're all born with it. So at some point in time, the system curbs it from us. So there should be more apps that get us back to what we're naturally good at. It's a fascinating tool for technology or for curiosity, rather, because it's seamless the way we use it on the show because there's always a question. It comes up so often. Throw it in perplexity. Let's find out what's up. So it's been great for us.

2:32:35So thank you. Thank you. All right. My pleasure. Bye, everybody.