Joe Rogan - Sean Carroll on His Problems with The Singularity & Ray Kurzweil


6 years ago



Sean Carroll

3 appearances

Sean Carroll is a cosmologist and physics professor specializing in dark energy and general relativity. He is a research professor in the Department of Physics at the California Institute of Technology. His new book "Something Deeply Hidden" is now available and also look for “Sean Carroll’s Mindscape" podcast available on Spotify.


Write a comment...


Hello freak bitches. I do think we could be living in a simulation as Elon Musk famously suggested. We could be all living in a computer simulation. I don't think it's likely but it's completely plausible given what we know right now. Right, the real issue is that one day we most likely will have something as long as technology continues to exponentially advance, will one day have something that's indiscernible from the reality. They'll be able to interface more than likely. To be able to interface with your senses with the way the mind perceives reality and create something that passes that uncanny valley and literally feels like real life. Like the matrix. I think we might be very far or we might be pretty close. I think it's hard to tell because there's so much we don't know right now. It's not like this is coming in the next 10 years. But it could be a thousand years. Yeah, absolutely. The thought experiment that helps people accept this is a single neuron in your brain. You have something like 100 billion neurons in your brain. But every single one is not that complicated. It takes in some inputs, it puts in some outputs. It might be work to figure out exactly what it does but it's not mysterious. So we could imagine replacing one neuron in your brain with a solid-state micro device that does exactly the same thing as that neuron. And you would be the same person, roughly speaking, right? So okay, so if you believe that then do that with two neurons and do it with three neurons and you're gradually building up who you are but just replacing your brain with something that is machine-made. And if you believe that can happen there's no reason to think that machines can't be as human as we want them to be. Do you follow any of Ray Kurzweil's stuff? A little bit. I don't think that he's an especially deep thinker but he's a good robotic viewer. How dare you? He's right now counting on his desk. That could be. This motherfucker. You don't think he's an especially deep thinker? I mean he's a brilliant guy, he's invented a bunch of really fascinating things. Yeah he likes to extrapolate without limit pretty fearlessly. Well he must have irritated you for you to jump right out with I don't think he's a particularly deep thinker. No, but I do think he serves a role. I think it you know there are people who make us think more deeply. I remember seeing a panel discussion with Murray Gell-Mann who was a famous physicist and Isaac Asimov. This was like 30 years ago and Asimov was still alive. And there's someone else who was on the panel, I forget it was, another physicist. And the amazing thing was the science fiction author was by far the most conservative thinker up there in terms of what he thought would really be happening a thousand years from now. The physicists had these way out ideas and Asimov was like yeah I don't think it's gonna be all that different right? And it's just hard to predict the future accurately and there's a role served both by trying to be as realistic as possible and careful and Bayesian and saying what are the probabilities and so forth. There's another role served by just being the provocative and saying well maybe this crazy thing is gonna happen maybe we'll cure death in the next hundred years and life will change for everybody dramatically. That seems fairly likely right? They're gonna figure out some way to reverse aging. Yeah I think that's plausible there's no biological reason why not. Kurzweil's got some weird motivations too. He literally wants to sort of rebuild his father. Yeah. Yeah he wants to be able to piece together some sort of an artificial intelligence version of his father and go back and see him. Yep. That's deep. That's a word for it. Yeah. He thinks that there's gonna come a time inside his lifetime hopefully that you'll be able to download consciousness. That consciousness is going to be something you'll transfer sort of like you know code. Yeah I think that there's there's a difference between you know what is potentially possible given arbitrary amounts of time and resources and what is realistic in the relatively near term. By the near term we mean 50 or 100 years right? And I think that people like Ray Kurzweil are far exaggerating what's gonna be possible the next 50 or 100 years because they underestimate how little we know about how the brain works how important it is for the brain to be in our bodies right? One of the breakthroughs in artificial intelligence over the last couple decades was to realize that if you try to build an artificially intelligent computer it becomes much more realistic if you give it a body if you give it a face and it can interact with people. I mean we underestimate the extent to which having a body is an important part of how we think and who we are and these this is just like such baby steps in understanding this stuff that to imagine that in the matter of decades we'll have it all figured out and have downloadable consciousnesses is not realistic to me. Do you think what do we have to absolutely understand the exact way that the human brain works in order to replicate its possibilities? No I mean I think that we probably won't that probably won't be the way that we make artificial consciousness or artificial intelligence like we won't just be reconstructing human beings like when we made cars we didn't reconstruct horses right we just did it in a very very different way and cars are much better than horses in in various ways not as good in other ways like going on pills going up hills but also you know in the early days of cars one big complaint was but if I'm drunk and passed out it won't take me home. That was a complaint. That's true. Because the horse would just take you home it knows how to get there. Exactly and finally we'll get the artificial. And nobody gets mad. Self-driving cars will be able to do it finally. If you're driving a horse drunk nobody gives a shit. Nope you're passed out. That's right. Yeah that'd be like a big adjustment for people. Yeah exactly. No more drunk. And so the idea that the way that we'll make artificial intelligence is to sort of mimic a human being is just crazy. That's not how it's gonna be. What do you think it's gonna be like? Well I don't know there are there are the problems that can be solved by things that we design are just a different set of problems than the thing that evolution naturally made us do right. Like evolution built a very very general purpose machine that is inefficient and irrational in all sorts of ways. Like anyone's pocket calculator since the 1970s can multiply numbers way better than your brain can right. Your brain has enormously more computational capacity than a pocket calculator. Why can't we multiply numbers? That ability was not important back when we were evolving these things right. So the set of things that it is easy to design is just very very different than what the brain does. So who knows I don't know exactly what it will be what it will look like. I actually think that the more important thing will be blurring the distinction between human beings and machines. You know the the crossovers. There's another one of Elon Musk's project is called the neural link. The idea of you know basically a neural lace something that is just interfacing with your brain very very very fast so that you have access to the entire internet or whatever peripherals you want in real time. So like Wikipedia is part of your memory essentially and that's just who you are now you walk around and you can multiply numbers as fast as you want to. That seems like very likely. Yeah that augmented reality. Yeah yeah and and some sort of a weird symbiotic connection to the net and to electronics whether it's a wearable thing or maybe even an embedded thing. Yeah absolutely I think it will be embedded. You think so? Yeah it's more efficient. Will you be one of the first adopters or are you gonna wait a little while? I'm never a first generation adopter even of iPhones much less you know things embedded in my body. But look I just adopted two kittens. They get microchip implants. Oh yeah. You adopt a cat that gets a microchip inside so you can't lose it. Yeah. And so to think that we're very far away from doing that I don't think is right. I think that we're gonna be doing things. I was just looking at a laptop bag. It's a laptop bag that also is you have a passport bag and a laptop bag and then a carry-on and then a check and stow you know on a for airplane luggage and all of it is Bluetooth and all of it is location coordinated with GPS so that if somehow or another your bag gets lost you literally can go on a computer and I'll show you where your bag is. Yeah. It's like whoa yeah how long we gonna when are we gonna do that with people? And I feel you know I I'm a big believer in privacy rights and so forth. Her dilemma do I let my employer microchip me? What is this? There you go. There's a company in Wisconsin that had microchip day on August 1st and they implanted microchips in people. First of all that dude with his ears and that those fucking ear plugs and the lower ear lobe thing you ain't doing nothing buddy. You ain't sticking that in me. Look what you've done to your ears fella. Wisconsin company becomes first US employee. What? And these are very simple just location tracking things right? Why would you allow the company to do that? I wouldn't allow my company to tattoo me. Why would I allow my company to microchip me? It's for tracking and you can also like they can buy stuff in the cafeteria. Yeah you know what else you could do use your wallet. You say that but you know. Do you turn off all cookies when you have your laptop on in your browser? Yes I do. No I don't. You don't because it's a little convenient. Yeah these guys say like they get they can buy their snacks with their microchip. They just walk up to the machine and get a snack. But here's where the separation. This laptop I can go like that and then I can walk out the door and that laptop stays here. All the cookies and all my browser history stays here. It doesn't it's not enough. That's just what they want you to think. Oh Jesus. Now you're freaking me out.