Transcript
Mayim Bialik (0:05)
Time. It's always vanishing. The commute, the errands, the work functions, the meetings. Selling your car. Unless you sell your car with Carvana. Get a real offer in minutes, get it picked up from your door. Get paid on the spot so fast you'll wonder what the catch is. There isn't one. We just respect you and your time. Oh, you're still here. Move along now. Enjoy your day. Sell your car today, Carvana. Pickup fees may apply.
Mayim Bialik (0:35)
Marshall's buyers are hustling hard to get amazing new gifts into stores right up to the last minute. Like a designer perfume for that friend who never RSVP'd wishlist topping toys for her kids who came, too.
Stephen Wolfram (0:47)
Mm.
Mayim Bialik (0:47)
Belgian chocolates for the neighbor, A cozy scarf for your boss, and a wool jacket for your husband that you definitely did not almost forget. Marshalls, we get the deals, you get the good stuff. Even at the last minute. Find a Marshall's near you. Hi, I'm Mayim Bialik.
Jonathan Cohen (1:06)
And I'm Jonathan Cohen.
Mayim Bialik (1:07)
And welcome to our breakdown. This is part two of our conversation with Dr. Steven Wolfram. He's the creator of Mathematica Wolfram Language. He's a pioneering figure in computational thinking for over four decades. He also was the science consultant on the movie Arrival, which is something we're going to talk about with him in. In Part two, we're going to talk about his thoughts on aliens and alien communication. We're going to talk about language and how human language is actually simpler than we thought, as well as trying to understand how computers communicate as a model for how maybe we are communicating with each other. We'll discuss free will and consciousness. And is time linear? We can't wait for you to hear part two of our conversation with Dr. Stephen Wolfram. Break it down. So let's go into aliens because not just for the flashiness of it, which we are sometimes interested in. You know, what do people who are thinking at these high levels think about things like this, but from also an astrobiological perspective, from a alternate perception of reality perspective. I mean, I don't think we're asking the question, are we alone in the universe? I think it's pretty clear that we're not. But tell us where you want to talk about aliens.
Stephen Wolfram (2:26)
Yeah, I mean, you know, I think it's a sort of philosophically muddled concept that we have minds that work in a certain way. There are many things in the universe that we might say do mind like things. I mean, a favorite thing people say is, you know, the weather has a mind of its own. Right, and what does that mean? Well, there's fluid dynamics processes in the atmosphere that are doing all kinds of complicated things, Just like in our brains. There are electrochemical processes that are doing all kinds of complicated things. How do we compare kind of the electrochemistry of our brains with the fluid dynamics of the weather? They're both doing complicated things. Can we say one is fundamentally simpler than another? Well, I happen to have done a bunch of science to do with the thing I call the principle of computational equivalence. That kind of says there isn't really a way of saying no, but the weather is somehow fundamentally simpler than a brain. They're kind of computationally equivalent things. Now we say, well, the brain, the weather, might have a mind of its own, but it's not a mind like ours, It's a different kind of mind. Well, lots of things in the physical universe I think do quotes mind like things. The issue is how closely aligned are they with minds like ours? How much are they minds where we can recognize them as being kind of as being minds like ours. And I think this is a thing, you know, you're talking about sort of spirituality and so on. And there's the question, you know, the spirit of the wind, so to speak, you know, is it a mind? Is the wind a mind like thing? Well, I think at a sort of computational level, yes, the wind is a mind like thing, but it's not a mind like ours. And so the question is really out in the physical universe, there are many things which are computationally as sophisticated as our minds. In fact, many things that are much beyond the computational sophistication of our minds in terms of having more components than we have neurons in our brains or whatever else. But yet the question is, are the things that are going on in those things, are they somehow aligned? Are they minds like ours? Are they sufficiently aligned with our minds that we will. That we can, for example, communicate with them? You know, communicating the weather, you know, tough, tough, tough thing. I mean, you know, it's not communicating with a dog. A little bit closer. You know, I think that this is so to me, this question of sort of, are there aliens out there, so to speak? The answer is there are aliens all over the place. There are alien intelligences all over the place. The issue is, are they close enough in the way I try to describe these things in what I call rulial space, the space of sort of possible rules for describing how things work? Are they close enough that we can communicate with them in some way? The universe as we were talking about before is quite big. It's a long way from here to the next nearest star and so on. You know, the physical universe that you might traverse with spacecraft is pretty big. The rulial universe, as I call it, the universe of all possible rules, is in some respects vastly bigger than that. So this question of can you get from a mind like ours to this alien mind, the distance you have to go is much greater than the sort of distance you'd need to go in a spacecraft to get to the next nearest star, so to speak. So in other words, the distance to that alien mind is very great. And for us to be able to kind of align the way we think with the way that alien mind thinks is a big challenge. Now, you know, we have a test case in terms of AI because, you know, AI is a thing that is an alien mind. We attempt to align it. I mean, a raw neural net doesn't do human like things. A raw neural net just hangs out and has random bits flying around. After you train the neural net, you can align it to do things that are quite human, like, like tell pictures of cats from pictures of dogs and so on. And then the question is, well, so now we've got kind of this mind that we can sort of experiment with, that we can align with our mind by training it in the right way. I've done some little experiments which are kind of like artificial neuroscience experiments where you take a trained neural net and you say, let's say it's a neural net that makes pictures of cats. You've told it to make a picture of a cat. The ordinary neural net makes a picture of a cat. Now you start sort of changing pieces of that neural net. So you make it go from a mind that was aligned with a human mind to sort of an alien mind by changing the innards of the network.