In our 100th episode special, Byron has a conversation with Stephen Wolfram on the nature of reality, belief and morality itself.
Stephen Wolfram is a British-American computer scientist, physicist, and businessman. He is known for his work in computer science, mathematics, and in theoretical physics. In 2012, he was named an inaugural fellow of the American Mathematical Society.
Byron Reese: In my capacity as the publisher of GigaOm, I’ve had occasion to interview Stephen Wolfram twice before. One was back in 2015 and an be found here, then again last year Stephen appeared on an episode of Voices in AI.
In those two interviews, we covered a great deal of ground, and I thought long and hard about what to discuss this time around. Much of Stephen’s work is quite practical, such as with Mathmatica and Worlfram Alpha. But he also spends much time up in the intellectual stratosphere where fundamental questions of reality are explored. He is arguably our generation’s best bet to Figuring It All Out, finding the fundamental nature of reality and what makes the universe tick. It is these topics I wanted to explore. In addition, much of his thinking ends up being almost religious nature. His view of physics borders on philosophy and even religion, so I was eager to explore his thinking there. So this interview is a bit unorthodox, but then again, so is he, so sit back enjoy.
Welcome to the show, Stephen.
Stephen Wolfram: Thanks.
Do you believe in God?
Oh, that’s an interesting question. I’m certainly not adherent of any organized religion. However, it’s an interesting question. The things that I’ve done in science tend to intersect in strange ways with things that people have studied in theology for a long time. I mean, for example, it used to be the case. Back in the day, there was this thing that used to be called “The Argument by Design” although that subsequently got a different meaning. It was a question of, look at the universe. The universe could be completely without laws, but actually, that’s not what we see. We see a universe that’s full of definite laws and rules and isn’t as complicated as it could conceivably be. People said, “Okay, that very fact is a proof of the existence of God.”
I guess that since I’m in the business and I happen to be actively starting to work on this again, of trying to find the fundamental theory of physics and believing that that fundamental theory has at least a chance to be simple, then at least by the standards of the early Christian theologians or something, I have to be following the argument by design. In so far as I believe that there’s a simple rule for the universe then their version of an evidence for something – their argument, I would have to say that I subscribe to. When was it? I was visiting some country. Maybe India where they put – on the visa application, they insist that you fill in religion. I was going to put there “animist”. My children said, “Don’t do that. It will just cause trouble.”
Why would I do that? One of the things that is a consequence of a bunch of science that I’ve done is this question of, what has a mind? What things that exist can be thought of as mind-like, like our brains, we attribute minds to. Some version of this is statements like, “The weather has a mind of its own.” The surprising thing that came out of a bunch of science that I did is that – in fact, there’s this principle of computational equivalence that says that in many ways what the weather does it just as mind-like as what brains do. That’s the concept of things like the animistic religions is this idea that there’s spirits in everything so to speak. This notion, does the universe have – is the universe mind-like? This scientific result, this principle of computational equivalence implies that. Following through on that, I kind of have to say at some level that I would be – should be considered by some classification as an animist so to speak.
Given what you know about physics and the principle of computational equivalence, is there any method by which the human could survive the death of their body in a practical way?
Okay. What’s a soul? That’s kind of what you’re asking. Is there a soul? What might the soul be like? I think we have the experience with computers now to at least imagine what souls might be so to speak. I mean, there’s a – okay, thought experiment you might do. I’ve imagined I was going to years ago and I may finally when I get totally old and unable to do other things actually follow up on this, but I was going to write some pseudofiction book about interviews with famous scientists and thinkers of old so to speak. Imagining the person goes from today’s world, bringing their laptop and goes to visit Pythagoras or something. Then the question is, what does – you have that conversation with Pythagoras, what does Pythagoras think the laptop is so to speak? The obvious thing is, it’s a bunch of disembodied human souls. You start peeling that back and you say, well, no it’s not. I mean, it’s just a piece of electronics. It’s like, well, who created what that electronics does? It’s a bunch of people. Who made that software work that way? It’s the ideas of some particular person.
I guess there’s a question of what the distinction is between the output of the level of software we write, words we write, whatever, things we record about our lives, and the actual internal state of brains. For example, one thing I’ve wondered about, I’ve recorded lots of stuff about my life. Millions of emails, lots of other things, and so I wonder is there enough information about me to reconstruct a bot of me by this point. In other words, my brain has some number of synapses, some amount of memory in it, and if you were to just take its output over the last 30 years or so, and say, okay, can we now reverse engineer what’s inside this brain? I don’t know what the answer to that. I don’t know if it’s possible, but I don’t think it’s obviously far from possible. There will come a point at which you can perfectly reasonably have something where it’s a – where you should be able to get a bot of me that will respond in more or less the same way that I’m responding in this conversation to you. Then we have to ask ourselves, is that me, is that something different from me?
I think that’s the point at which we have to start wondering about, is the bot of me the soul of me so to speak or not? There’s a question of whether you can do it with reverse engineering or whether you have to take a brain and dissect it and pull out all the data that’s stored at each synapse or some other thing like this. I think my answer is that the – I really don’t doubt that the soul in this informational sense of a person, I think the thing we’ve learned from the whole computational experience is that it’s extremely really certain that eventually that will be preservable digitally and independent of the biological manifestation of the human.
You and I have had a conversation before and I’ve probably never really expressed my question clearly enough, but I always come back to it when I think about it, and it goes like this. You know people who say they believe something like they believe in treating everybody nicely, but then you see them mean to people. You say, “Aha! You don’t really believe that”, or all kinds of things where people say they think one thing, but their actions sure imply they think something else. When that happens, we tend to think whatever they do really is what they believe. When I talk to you and you talk about the weather has a mind of its own and a storm cloud – a hurricane and the brain are the same. Then when I try to talk about consciousness you get dismissive and say, “That’s just a word.” Then you say things like, “It’s all just computation. Everything in the world is simple rules iterated over and over.”
All of this very impersonal non – it’s just a bunch of cranking numbers. A whole universe is just that and if we could see it well enough, that’s what we would just see is just a bunch of numbers, and yet, I know you to be like an emotional and compassionate person who loves things and doesn’t like other things. I see all kinds of ethics. You have an ethical code and a moral framework and all of this stuff. I have to look at it and say, that does not logically flow out of what Stephen says he believes. I can only really infer that you don’t actually believe it. It’s a good model for understanding certain things, but it isn’t actually your core belief because it’s so – you could imagine somebody who lived consistently with that view of the world and really said, “Nothing matters. A storm dissipating and a child dying are just the same thing”, but you don’t think that. I posit you don’t actually believe it. It’s convenient way to think of the universe, but it isn’t actually what you believe.
It’s an interesting topic. It’s like, I like chocolate. It gives me a good experience when I eat it. I could imagine deconstructing that whole process and realizing, “Gosh, it’s just some neural firing, etc.” My subjective experience of it is, “I like chocolate.” Therefore, since I live in my subjective experience, I do things which pander to my subjective experience. Now one of the things I might say about things I’ve discovered in science is I don’t necessarily like all the things I’ve discovered in science. The concept that, for example, the unspecialness of us as humans and so on. I don’t particularly like that. It’s just I pride myself on being a decent scientist and so I discover these things and that’s what I’m going to report so to speak. Rather than saying, “Well, I’d like to hide the fact that actually, there’s no real purpose to the universe. We’re not that special. We’re not that unique, etc.” For me personally, in terms of my subjective experience, yes, I like people. I find people interesting. I think people are – I’m interested in people person by person so to speak, and yes, in terms of the science I’ve discovered, makes absolutely no sense.
A lot of things I’ve done are in a sense deconstruct the meaning of things. They explain in a broader context how things work and they show that something is not as special as we might at first assume that it is. I don’t think this idea that that means that – does that affect my subjective response to these things? No, I suppose I could whip myself up into the frenzy where I would say, “I don’t care about anything. It’s all just computation all the way down”, but that is not my human subjective reaction. That is, what I’ve discovered in science and what I report as being a good scientist so to speak.
It almost sounds like you’re agreeing with me there. You’re saying this is like a useful model to understand the universe, but I’m not going to live that way. I’m going to live as if people are special. I’ve never known you to get emotionally attached to a hurricane. You do get emotionally attached to people, and so you live as if people are special.
Living one’s paradigm is really hard. I’m always curious, when I see people, who’ve discovered things about the world, and you ask, do they in fact live that paradigm? Sometimes they do and it leads them into terrible trouble because that paradigm – and often they don’t. I think isn’t there a quote from Tolstoy about how “I’m not a very good Tolstoyan.”
When you see fields develop, intellectual fields develop, it’s a funny thing. There’s a generation that invents the field and then there are generations that come after. The generation that invents the field, still knows all the things that are wrong, all the foundational things that they’re not really sure about, and so they’re a bit more tentative about it. By the time you’re at the fourth generation, they’re like, “Well, of course, it works that way.” We have this whole culture built up around, this is the way things work.
Now it’s certainly true that one could imagine – you asked about religion early on here. It’s certainly true one could take the things I’ve done in science and one could build something that many people would think of as being religion-like set of beliefs around it. Those beliefs would be very cold in many ways. They’d be very non-human. In terms of my subjective way I lead my life, that wouldn’t be natural to me. That doesn’t mean that I don’t think that these things are scientifically correct. It’s just a question of – just like I like eating chocolate, but it’s not that there’s something scientifically amazing about the chocolate molecule or whatever. It’s just that, the way that – actually, I think that – let me roll this back a little bit because I think there’s a – one of the things that does come out of the science I’ve done is the following observation.
You might think that what’s special about where we are as humans is we’re the only intelligent things in the universe, and that that’s what’s special about us, and we should be very proud of that attribute. What the science that I’ve discovered shows is that is not – if that’s what we’re proud about, then we are barking up the wrong tree. That’s not the thing that is special about us, but the thing that is special about us is lots of details. In other words, what this idea of computational irreducibility implies is the notion that, in order to know what happens in a system you just have to trace through what the system actually does. You can’t go and just look at the system and say, okay, I can jump ahead and tell you what’s going to happen in a million years, and so it is with human society. That if there wasn’t computational irreducibility, we could say, oh, look at human society, people are running around doing this and this and this, but the outcome is going to be blah. There’s no reason for these people to be going around and doing all these human things. It’s really just all a waste of time. In the end, the answer is 42 or whatever.
What computational irreducibility implies is that’s not the case. It affirms that something is achieved by the human experience. That is that it’s not the case that you can just take the universe that we live in and say, “Okay, the outcome is going to be this.” It’s like the actual – the living of life so to speak is the story. It’s not that this is just a piece of a calculation where the answer is going to be 42 so to speak. What I’m saying is that I think that in a sense the science that I’ve done, you might say it says it’s all pointless in the sense that there’s nothing special at the level of thinking about – there’s no big special thing. It’s not that we are the only mind-like things in the universe. What it’s saying is, there is a special thing and the special thing is all of our details.
I think at some level actually I’m going to disagree with myself and you here because I’m going to say that I think that point, as you really start to internalize that point, that the details of what happens are the things that we should – that are special about us and that we should think are important, that actually is a rather human-oriented view of things quite different from the cold view of, “It’s all just computation. Everything is computation.” Yes, that’s true, but what is relevant to us is the special computation that is us. That’s something where we can revel in the details of that. Even though we know that the whole phenomenon of computation is not – there’s nothing abstractly special about it. It’s something that is…
Yeah, I find that unsatisfying candidly because you could – beavers could say that too. They could say, “It’s the experiences that all of us beavers have building our dams that make us special.” A hurricane could say that. It could say, “It’s all the places I went.” Everybody doesn’t get a medal.
Why do you say that?
That’s just another way to say that nothing is special.
The point is that I and you, we’re all members of this collection of humans. I think it is correct that if you look at the beavers, the whales, the dolphins, the storms and so on, there is some sense in which each one of those is special. We just don’t happen to be one of those. We happen to be humans. I don’t think you can say in the – I think it’s funny - in the modern world where people are so concerned about equality of various kinds. This is a form of equality that people haven’t yet started thinking about. That is, who are we to say that we should be intrinsically any more special than the weather or than the beavers so to speak. I think that what the science is saying is we’re actually not any more special, but that doesn’t mean that in the conduct of our lives as humans, that we shouldn’t view what’s going on around us as humans as being something special.
It’s just that we can’t – it’s an almost religious claim that we would be more special than anything else, and people have imagined that the science would say “No, no, no. We’re genuinely special. We’re much more special than anything else.” My conclusion from the science is that’s not the case. For example, it makes me worry less about oh my gosh, what happens when the extraterrestrials arrive. It could be the case that I could have a big inferiority complex. We could all have. All the stuff we’ve done, all these things we’ve figured out and etc. It’s all for nothing because those guys over there on Alpha Centauri, they figured this out a 100 million years ago or a billion years ago, whatever. There’s nothing that we could possibly have achieved, but I think I’ve become a lot less concerned about that. I’ve become a lot more interested in a sense in leading my life and trying to figure stuff out and trying to find – it’s like I could say, “Well, what’s the point of trying to find the fundamental theory of physics? After all some entity in the universe has surely already found that. I’m going to be scooped by whatever.” I don’t feel that way actually. I feel in a sense by knowing that in some sense, “Yes, in some senses of the word…” because it’s not even clear what you mean by figuring out – figuring out the fundamental theory of physics is a little bit like saying – like making claims about how Columbus discovered America so to speak. America was already there. It was like he just made the link between Europe and America. Similarly, the universe is already there. It’s just what we’re doing by discovering the fundamental theory of physics is making a link between our human experience and what exists in the universe.
I’ll only say one more thing along this line because my list of questions is so long. What you said earlier about people not living their paradigm. I believe people are special. I have a teleological view of humanity, that we’re heading in a direction. Because of that, I believe in things like human rights. That’s why there are such things where there aren’t chair rights. There aren’t storm rights. It’s because of the specialness of humans. I live that paradigm. Now I’m imperfect in so many ways, but just with regards to this very narrow thing, I live my life as if people are unique. And see, that’s my point is, you seem to live in my paradigm as well.
Oh, yeah. I too. Yes, I do.
I have to believe you really believe it as well.
Okay. Let me ask you this question then. You probably believe in some level of animal rights as well?
To what extent do you believe in AI rights?
In practice, I think that if anything can feel pain, it gets rights.
All right. How do you feel about switching off your computer?
I have no reason to believe my computer feels pain and so it’s…
I think this idea of feeling pain is a bit circular. Let’s not try dissecting that one.
(Laughing) You see, it’s that sort of statement. You know what pain is. You avoid pain. You don’t want to feel pain and yet you have to say, “Well, there’s really no such thing as pain. It’s just computation.”
No, I understand, but the thing is, the subjective experience that one has is the real thing for one, but that doesn’t mean that – I mean, this is a sad fact that, I don’t know if it’s sad or not, but its subjective experience which is what you live in in living your life. It’s not the same as objective reality. That is, you can have – it can be the case that you know perfectly well that, I don’t know, something about the universe or something about the paths of human lives, but yet you can act in your subjective experience, you’re locally optimizing your subjective experience in a way that is completely oblivious to this global objective reality. I think that’s what one has to do because I think otherwise – another way to think about it – okay, let me make a general statement. It is your allegations about my lack of paradigm living. I do think about this from time to time, and I have a gut feeling that it’s not as far away as you think. I don’t know whether I really parsed it apart properly. One of the things I find peculiar in my life is that I look at stuff I’ve done and I say – and then years later I realize, Oh, that’s why I was able to do that.” or “That’s why I figured that out.” For example, I’ll give you an example. Early in life, I was a physicist and physicist typically work in the following way. They say, “Let’s look at the natural world as it is. Let’s try and drill down and find out what the primitives inside that natural world are, and let’s do it in this reductionist type way.” One of the things that I did early on when I started to learn cellular automata and that eventually led to A New Kind of Science, was I said, let’s just start inventing universes and see how they work, and then see whether they have a correspondence to our actual universe. It was only years later that I realized why I had the intuition that that would be a reasonable thing to do. The answer is, I started doing that after I had invented my first computer language. When you invent a computer language, what you’re doing is you’re essentially making up this artificial universe, and then seeing how it plays out. I didn’t have that… The fact that I ended up having that slightly different intuition from the intuition that your average physicist would have about how to do science was a consequence of that. I feel like a bunch of the things that I do and I – it’s interesting and I can’t do it in real time, but I have this suspicion that a bunch of the things that I do make more sense than you think they make so to speak, or that I can see that they make right now. That is that there’s more connection between the things that I believe scientifically and the way that I lead my life than superficially appears to be the case. Although I’m not claiming – it’s not a causal relationship. As in the fact that I lead my life the way that I do is not a causal – is not a consequence of some scientific belief that I have.
When I run my company, you might think, oh gosh I must be using all kinds of analytical methods and so on that are derived from science that I’ve done. No, I’m not. I actually am quite a – I’m a gut-thinking type person when it comes to those things. I have the suspicion that there’s more correlation that one is immediately recognizing because there’s certainly – just like you, I have certain outlier tendencies. Maybe different outliers from yours, and I have a suspicion that this outlier tendencies are more correlated that one might think. I just don’t yet – I haven’t yet gone back and understood the history well enough and the themes well enough to see how those different tendencies are correlated. Superficially, your allegation of lack of paradigm living appears correct. My gut feeling is it’s not as correct as one thinks it is, but I can’t prove that in real time.
Fair enough. What is your explanation for the Fermi paradox? [Note to reader – the Fermi paradox poses the question of “If there is so much life out there in the universe, where is the evidence of it all?”]
There’s lots of intelligence out there. We just don’t recognize it. It’s not commensurate with our way of understanding things. Let me give a parable version of that. I mean, got involved in my friend Nova Spivack’s peculiar project to send little beacons expressing the achievements of our civilization out into space. One of them was on this Israeli spacecraft that encountered the moon in April, I guess. Which we thought had been smashed to pieces when the thing ran into the moon at what, 2,000 kilometers an hour, but actually it’s now clear although I still haven’t really written this up, but it’s now clear that the thing had a sufficiently grazing impact on the moon that it bounced and probably the payload is actually just fine and sitting somewhere. Happily sitting on the moon. Okay, the question is, what on earth is in that pay load, and if the extraterrestrials show up, will they understand what’s there? What could conceivably be there that would be “understandable” to the extraterrestrials as a beacon of the achievements of our civilizations?
What you start realizing as you start trying to unpack that is, it’s a philosophically doomed concept. In other words, this idea that there’s a way of expressing the achievements of our civilization. Again, you’ll criticize the fact that, here I am trying to produce works of some permanence and yet saying – like I said in something that I wrote about the project to put beacons out into the universe, I was saying, if it was a prospectus for an IPO and it has a risk section in the prospectus, that they might as well run that poem of Shelley's about Ozymandias, “I’m the king of kings. Look on my works, ye mighty, and despair.” It’s this broken statue somewhere in the desert. That that would be the appropriate risk section. Basically, my point of view about the Fermi paradox is yeah, there’s plenty of intelligence in the universe. It is in the form of things that we might say, “Oh, that’s just physics.” In fact, it’s even sobering to think about the distant future of our own civilization, and what aspects of that are likely to be recognizable – even to recognizable from us today for example. We look back at the archeological past. We say, “What were those people…” we know that they were people. We find their skeletons, whatever else. “What were those people thinking when they did this?” Okay, well, it’s all well and good because we can find skeletons and things like that, but one day, if it’s disembodied souls in the future as the future of our civilization, it’s going to be this very much more – just a bunch of bits flying around and it’s going to be very hard to know what – to be able to say, “Gosh, that’s so special and intelligent.” Actually, it’s just a bunch of bits flying around and to us, those bits or – to us, at that time, so to speak, those bits may seem incredibly meaningful, but in the abstract theory of bits, there’s nothing very special about them. That’s what I think is the case. I think the universe is full of intelligence, full of things that have evolved over long periods of time. Maybe things that have shown lots of the features of what we consider to be life and biological evolution and civilizations and all this kind of thing. In the end, it’s all computation so to speak, and you can’t really distinguish, you can’t really say – and it comes back to this whole question about specialness - you can’t really say, “Oh, this is…” I suppose the other question is, we represent some zone in the space of all possible computations that could be being done. Our civilization is this little point in the space of all possible civilizations, computations, whatever. There is a good question about how big is that space of possible civilizations and computations and so on. What is the likelihood that there’s a point close to us? Is it going to be the case like in Star Trek or something that you go visit another planet and there are a bunch of humanoids whose only difference is they have pointy ears and so on? I think it is likely that the space of possible computations is really very, very vast, and that the overlap so to speak between things that we recognize as, oh yeah, that’s a human-like civilization, and what’s out there in the world of possible computations that can happen in our universe, it’s likely not to be something where you can readily align and recognize it. My belief is there’s intelligence all over the universe.
Also, to your point about specialness, it’s like, do we care about it? We say okay, it’s very fine that these gas clouds are doing all these complicated things and physicists should go study that. We don’t go and say, “Gosh that’s an amazing civilization. Look at all those. The art that they produce and the whole complicated array of patterns of these streams of dust and so on.” We just say, “Oh, that’s just a physical thing. We don’t care about that.” It reminds one a little bit of some of the things that you look at in human culture where even between different societies, and different times in history, different kinds of activities that people undertake. There are things where you say, oh I don’t know, for example, I’m not a big person watching sports for instance. I see some sporting thing on television, it’s like, I have no idea what’s going on. I don’t know why people are doing this. It looks completely pointless to me. It is something where – is it intelligent behavior? I haven’t a clue. Its people running around and stopping and screaming and this and that and the other. I have no idea what’s going on. I think this question about how close do you have to be to recognize, to be really impressed with the intelligence, so to speak. The answer is probably quite close and the fact is the universe is the space of possible computational things that can happen is very big. There isn’t that much alignment that’s going to happen.
You don’t believe A, that we’ve been visited by aliens or B, that there’s likely to be intelligent life as we know it in our solar system.
In my version of aliens, we’re visited all the time. We’ve got all these radio emissions from pulsars. We’ve got all these things, all kinds of things that go out.
Right. But as everybody else in the planet would mean it?
What I’m saying is I think that the alignment of – this is what I’m saying. The question is, in the space of all things that can happen in the universe, is it the case that – have we been visited? Okay. I have a friend who has the belief that actually derived I think from some of the things that I’ve said that that the end point of intelligence is data transmitted through light or the electromagnetic radiation. That in the end, and may very well happened to our species that maybe we’ll all be – it will be the box with a trillion souls right here on earth. Maybe we’ll say, “Okay, let’s just transmit everything about our – this planet isn’t going to be around forever. We want to feel proud of ourselves and want to last forever. We’re just going to transmit all of this information about our existence to the whole universe through electromagnetic radiation” and maybe we’re going to find some way which doesn’t work with light as such, but some way to get – it actually could work although it’s a little bit of an ornate physics question to get the interactions that can have continued thinking going on…
Then it’s like, okay, there’s electromagnetic radiation. It’s all over the universe. It is a representation of our civilization and intelligence. You say, have we been visited by things like that? Yeah, the answer is there’s electromagnetic radiation from other parts of the universe coming to us all the time. A lot of it is – in my view of computational equivalence, a lot of that is as computationally and mind-like sophisticated as anything that we’re doing here on earth so to speak. In a sense, what I’m saying is, it’s like many of these science questions that seem mysterious. They end up getting answered not by frontal assault of saying… but by realizing that the question wasn’t really very well formed. That’s what I think is the case when we talk about aliens and extraterrestrial intelligence and so on. What we mean by intelligence is just not a very well-formed thing in that construction. What we immediately assume just isn’t quite right, and when we reformulate it, the answer is obvious. That’s my current point of view.
You have a new book out, Adventures of a Computational Explorer. It’s a book of consisting primary or exclusively of your well-researched blog post compiled. I want to read you just a sentence or two out of it. “If there’s a simple rule for the universe, what might it be? I’ve done a lot of work on this. Written quite a lot about it. One important thing to realize is that if the rule is simple, it almost inevitably won’t explicitly show anything familiar from ordinary everyday physics.” What does that mean? Second, how close are we to finding that rule? Are you going to do that? Are we going to do that in your lifetime, or is there no way to know that?
Right. I don’t know the answer. I’m actually finally after 40 years of thinking about it, I’m finally ready actually, really in these times to jump in and try and make my most serious assault on trying to find the fundamental theory of physics. My question is, is this the right century to do it? I don’t know. It’s not obviously not the right century to do it. The history of physics is, the current fundamental theories that we have were basically invented 100 years ago. They’re basically two strands. One is general relativity, the theory of gravity, and the other is quantum field theory, the theory of particle physics and small things. The framework of those theories. General relativity is from 1915, quantum field theory is from the 1920s. They’ve been embellished and developed over the years, but basically the frameworks come from those times. As I mentioned earlier, it’s like when you have a field that has gone through many generations, by now people would say, “Well, of course those are the frameworks. That’s the only way we could possibly understand the universe.”
We’re at generation five or something for physicists. I think that the – what I’ve been interested in is. “Okay, let’s imagine throwing all of that away. Let’s imagine starting from nothing. What is the most structureless structure that we can imagine from which we might be able to build the universe?” I guess that I have this experience of having gotten intuition from exploring the computational universe with simple programs and discovering that you might think a simple program would only do simple things, but you’d be quite wrong about that. Simple programs can do incredibly and in fact, maximally complicated things. That gives one at least the hope that maybe our universe would correspond to a simple program. Then the question is, can we go find it?
One feature of a simple program is you just don’t get to put into that simple program things that are familiar to us. For example, there are three dimensions of space that we commonly experience. You don’t get to put a number three. If you’re going to represent the whole universe as a few lines of code, the number three is very unlikely to appear in those three lines of code. Even if the number three occurred, the muon-electron mass ratio, 206 point watever it is. You don’t get to put that in there. You don’t get to put in the fine structure constant, 1/137 point whatever it is. You don’t get to put in all these other features of physics as we know it. Basically, if you have a rule for the universe that is simple enough, all of those things have to emerge from the operation of that rule. They can’t be things that you say, let’s stuff them into the rule to begin with.
Actually, it’s funny because it’s something that people point out to me with some regularity is, Einstein in his later years, to most of the physics community was off doing old fuddy-duddy stuff of trying to invent some unified field theory. It is amusing that in his later years he at least made one comment that is extremely aligned with the things that I’m trying to do about the fact that ultimately, will turn out that space-time is really discrete. We just don’t have the tools yet he said in the 1950s to go further with this idea at this time. Now we do have the tools. At least I think we have – we have tools that allow us to go further. Whether we’ll get to the end, I don’t know. This is a strange project for somebody like me because most of my life, most of the things I’ve done are purely building. That is, it’s like you start from nothing. You build a big computer system. You build a computational language. You build stuff and in a sense what I’m doing when I’m building those things is I’m building things that fit in to the human – they are things that are relevant. They are bridges to humanity so to speak. When you build a computational language, what you’re doing is making a bridge between the abstract computational universe and the stuff that us humans happen to care about. This project of trying to find the fundamental theory of physics is a different project. At some level, it’s a bridge project because at some level it’s like, well the universe just is there and the question is, can we turn it into something that we humans can understand about it. At another level, it’s not a project where we’re just building something. It’s a project where we might happen to have the right idea about how it works or we might not. It’s not something where – in any of these projects that I might do like building a computational language, there’s no sense in which the project might just outright fail. You’ll build something, whether it was useful or not as useful, that’s a different matter, but you’ll be able to build something. Finding a fundamental theory of physics, it could just be that there isn’t a way to find a fundamental theory of physics.
First question is, is there a fundamental theory of physics? That’s the first question and that comes back to the theological questions we were talking about at the beginning. It’s not self-evident that there should be a fundamental set of rules that govern the universe. It could be that the universe capriciously just does what it feels like. It’s full of miracles. It’s full of all kinds of things, but in fact, what we have come to believe - and it is a belief. In the early days, it was a theological belief, then it became a scientific belief – is that there are natural laws.
Then the next question is, is it a theory that we are close to being able to understand? That is, it could be the case that it’s a theory that is conceptually beyond what our human brains and what are civilization is capable of turning into a thing that we can tell as a story to ourselves. What do I mean by that? For example, in doing mathematics, we invented a series of different things. We invented mathematical notation. We invested arithmetic, algebra, etc. Over the course of a millennium or so we’ve invented layer upon layer of mathematics. If you look at modern, pure mathematics today, it’s built on this tower of concept that we gradually as a civilization, gotten them to understand. It’s similar to the way that we invent words and human languages.
The idea of a, I don’t know, a podcast for example. Now we know what a podcast is, but back even 30 years ago, somebody would say, “Well, it’s this thing. You record it and then you transmit it.” Then whatever. We didn’t really have a way to talk about it and think about it. Eventually, we have this concept podcast, the word podcast and then we can start reasoning about it. There’s a question of whether we have the abstract apparatus to construct a theory about the universe that we humans can reason about at the current stage of the development of our civilization and it’s not obvious. I think that there is a – the reason that I’m interested in trying to find it is that I think we failed to look for it for many years now. It could be that it’s quite low-hanging fruit. We just don’t happen to have bothered to go out and look for it.
To drill down on that, you made the comment about Einstein saying we didn’t have the equipment to do whatever, but you said you got this brainwave that tiny simple programs could produce complex results. Pascal also had a notion of that with Pascal’s triangle and all that. I guess if you think about Newton, he didn’t have equipment, right? He just had that apple falling on his head. Einstein, mainly his stuff was conceptual. He said, “I saw a triangle by and I wondered what would happen if you turned on a flashlight.” I’m curious to what extent do you think that this theory, if it exists, and I suspect – you suspect that it does, is it something that we could have come across a thousand years ago if somebody had just thought about it for long enough and had been the right person, or is it something that’s actually going to require equipment that will give us some key insight that we’re missing now.
The thing to realize about what’s happened in science and mathematics is it’s built on a lot of layers. It’s an inexorable thing. Somewhat similar to this phenomenon of computational irreducibility that I’ve talked about in connection with life in general. It’s something where if you say, “Could you deliver, if you went back and delivered a modern mathematician or scientist or something into ancient Greek society, would people just say, ‘Aha now I understand everything’.”? The answer is, I think there are a few things where that would we the case but there’s an awful lot of stuff where there are towers of understanding that are quite deep that you have to have, and that are, in a sense are things that become part of society. There are things that people take for granted and experience from the earliest times in their lives that make things seem much more obvious. I’ve wondered with my things like cellular automata, my all-time favorite discovery is this complexity from simplicity in the Rule 30 cellular automata. I have certainly wondered for a long time, will we one day dig up an archeological artifact that is just a picture of Rule 30. There’s nothing about Rule 30 that appears to be conceptually require the 20th century so to speak to get to it.
What I think does tend to happen is yes, there’s a big waterfront of possible things to think about and the things people think about end up being things that somehow fit into a narrative that has been societally developed. There just wasn’t any narrative that something like Rule 30 could have fit in to. Nobody thought about it because there’s a big collection of things that could be thought about. Now if you go back – for example, when I published this New Kind of Science book, there are plenty of people who were saying, “I just don’t get it. I don’t understand what the point is. I don’t understand how – why do we care? How does this fit in to anything?” Now it turns out that in the time since I wrote that book, there’s been this transformation that’s happened after 300 years of people using mathematical equations to model things, now the most common way to model stuff is with programs, which is what I was talking about in that book. Now it’s like, well of course, people use programs to model things. Why wouldn’t you? It’s the obvious thing to do.
It requires this societal context to make it all make sense. For example, one of the things that an interesting one, a historical issue is the idea of the universal computation. The idea that there are programmable – that it is possible to make programmable computers or programmable things. Leibniz almost had that idea in the late 1600s but didn’t quite get it. He didn’t get it crisply. When Turing had that idea… Well Gödel had the idea, but he wrapped it up in so much mathematical complexity that the crispness of that idea was not at all clear from what he did. Turing had the idea more crisply in 1936, but he also didn’t know… He didn’t know that it was that significant. It only became significant when it coincided with a lot of other things that the civilization produced.
Will we say, ‘oh gosh, this is the theory of the universe. That should have been obvious. It should have been obvious to Newton that this is how things work’.” I think it’s extremely unlikely that that’s the way it will play out because of this phenomenon that the – what is a plausible fundamental theory for physics now will be something in the sense extremely low level from which there has to be – it’s many layers of emergence above that before you get to things that are recognizable to us from either our everyday experience or even the physics that we’ve discovered. I think that those layers of emergence necessarily involve both quite a lot of abstract work and abstract concepts and also quite a lot of actual practical computation. I think it’s not going to be – I think we already know that the fruit is not so low-hanging as that we can pick it and be embarrassed that it wasn’t obvious. I think it is the case that if it turns out to be low-hanging enough that we can pick it, people including myself, will kick ourselves a little bit for, “Well, we could have done this a few decades ago.” I don’t think it would have been we could have done this a few centuries ago.
Artificial intelligence, you wrote in the book that it’s going to become obviously a bigger part of our life, that it’s going to be like government. It’s going to be something that we’re going to have to grapple with. What do you worry about and what do you hope for with artificial intelligence? Do you worry about that it’s going to disrupt employment and through automation take a lot of jobs? Do you worry that we’re going to automate warfare? Do you worry we’re going to use it to invade privacy? Do you think we’re going to use it to increase our lifespan? What are your hopes and fears at the edges of that technology?
Automation is a long story and AI is just the latest episode in the trend of automation, the going from the purely biological to the thing where we’ve automated. I suppose that the – gosh, it’s a big topic and it relates back to a lot of these questions about human purpose and the specialness of humans and so on. I think that AI, it allows us to leverage what’s possible from computation. What’s possible from computation is an awful lot of stuff. The question is what do we choose to get from that ocean of computation? What we choose to get are things that seem purposeful to us humans to get. In a sense, what AI is going to do is put a stronger mirror back on us about what is it that we actually want. We will be able to achieve a lot more. We will be able to by dipping into this computational ocean, we will be able to do all kinds of stuff very automatically and so on.
The question is what do we actually want? A large part of my life has been building our computational language and one of the points of that computational language is to provide a bridge between things that humans think about and what’s possible in this ocean of computation. I think that will be the big theme of AI, that “OK, we can do anything. What do we actually want to do?”
I think it puts the spotlight on how we as a society think about the human condition and so on. I was involved recently in this question about what I was calling automated content selection businesses, companies like Facebook and Twitter and Google and so on, where they’re going through billions of web pages or items or news stories or tweets or whatever that exist, and they’re deciding which ones should they show to a particular person at a particular time. The question is, OK people say, “Let’s get rid of hate speech. Let’s promote good, wholesome things.” Or maybe more commercially, “Let’s increase engagement by showing things to people that they want to click on” and so on. Then the question is how should we define the ethics of what we’re going to show? That turns into “What do we want for the world?”
People say things like, “We want to make the world a better place.” In Silicon Valley at events people will say, “We’re doing this and this because we want to make the world a better place.” It reminds me of religious benediction-type statements. That’s become the benediction, to make the world a better place. Now the question is, “What on earth do we mean by that?” We might say “Let’s just make an AI that can figure that out.” You show the AI the world and say, “Make the world a better place.” What on earth is the AI going to do? It could say, “I can simulate all of society and make it be the case that there are three people in the world who are happier than anybody could possibly ever be at any other time, but there are seven billion people who are kind of unhappy.” [Or] “I can make it so there are six billion people are pretty happy, but sorry a billion are going to be very unhappy” or some other distribution of whatever attributes, happiness, fulfillment, whatever you want to say. Then the question is “What do we want to do?” There’s no right answer to that question. In the world of political philosophy and ethics and so on, people have been debating what one might do for thousands of years. It’s not just that they failed, but there isn’t a right answer. There’s no mathematical theorem that says this is how the world should be. The only things we can say are things like, “If the world is this way, then everything is going to die out.”
If you have some culture where you have a principle that nobody should have any children, it’s not going to survive. Beyond things like that, I think there’s little that one can say. I think computational reducibility even means one’s ability to say “If you do this and this, society is going to blow itself up, so we shouldn’t do that.” I think that’s a hard thing to conclude. I got interested in this question – I sort of got backed into it because I agreed to testify in Congress.
Testify in Congress, yeah.
What happened there was they were talking about, “Gosh, let’s make sure that nothing bad happens with these AI systems. The way we’re going to do that is we’re going to have an inspection clause, so to speak, that says somebody can go look inside the AI system and check that it isn’t doing anything bad.” I was like “No, that’s not going to work. There are both practical reasons that’s not going to work and theoretical reasons like the whole computational reducibility business that says that’s just not going to work.” I was imagining I was just going to say that’s just not going to work. I was a little embarrassed because it was like I’m going to go talk to these guys and I’m going to be like the typical intellectual who comes and says “Nothing works.”
I want to be a bit more constructive, so I started thinking how could one make this work? I realized that it’s not that hard. It’s sort of an interesting synergy between humans and AIs. The fundamental question is what do humans want? The answer is there’s no one answer to that.
Somebody who says we’re going to make up the definitive Facebook ethics that will govern for the whole world and maybe they’ll have some algorithm that depends on who’s looking at them. It will be based on you’re in place X, so this is going to happen, etc. It is not the case… the idea that there could be such an ethics, I think it’s just a wrong idea. What I realized is there’s actually a way to deal with this that is very straightforward, which is to just say to people “You’re going to finally rank things that appear on your news feed or whatever or things that appear on your search results.” You can say in the end, “I want some trusted entity, call it a brand, to do that ranking for me.”
Obviously, they’re not going to have a collection of wise people sit down every moment and go through a million things a second deciding how they’re going to get ranked. That’s going to have to be represented in an AI. The idea is that different brands can take responsibility for ranking. People can expect that some brand that’s a very child-friendly brand is going to try and do this kind of thing. Some brand that’s very politically left wing is going to do this kind of thing. Some brand that’s very intellectually oriented is going to do this kind of thing.
The idea is that then something like Facebook, you’ve got a network effect that means it doesn’t make a lot of sense to say let’s just break Facebook into 20 different networks. It’s just not going to work very well that way. What you can do is to say let every person pick between their hundred possible ranking providers. Do they want the one that is branded with New York Times or do they want the one that’s branded with their favorite religious organization or their favorite celebrity or whatever else? Then the question is how do you take those brands and bottle up their essence, their preferences into an AI that can then be run in real time as a way to rank content to these services?
I realized that the whole set of things I’ve done with computational contracts and representations, symbolic computational representations, of every day discourse and so on, this is all closely tied into this question of how do you bottle up these preferences. What does it mean for something to be a credible news story? Different people would differ on that, and you might just say, “What has it meant to newspaper XYZ for something be a credible news story just based on what they’ve actually run?”
I’m reminded of a result my son, Christopher, got from looking at the front page of The New York Times. You might have thought the news cycle was getting faster in the sense that when an event happens like in the 1906 San Francisco earthquake, it would stay on the front pages less time now because the news cycle is getting shorter. He found the opposite. He found it’s getting longer. I was talking to a friend of mine who used to be the editor of a newspaper and was asking him about it [and he said] “Yeah, once you have a horse that’s succeeding, you want to keep running it, so to speak.”
My belief is that the right way to handle this is with sort of a synergy between human brand-like or tribal behavior and AI, that is being able to sort of bottle up some set of preferences of some group of some brand and be able to use that is something that people can select between. People then say, “Oh, my gosh. That will lead to more and more echo chamber-like behavior where people only listen to things that they already knew they wanted to listen to.” I have to say, I think that’s “Welcome to the species” type of thing. Personally, I would find it interesting, or at least I think I would find it interesting, to have some service like this that says, “80% of the time I’ll show you stuff like what you’ve seen before, but twenty percent of the time I’ll show you this wild, random stuff out of left field” because I’m one of these people who happens to be curious about just stuff that’s out there in the world. That’s just my personal meta preference, so to speak.
I fully recognize that other people have different ones. I think in a sense this idea of mine, which I consider to be a simple and obvious idea, although it seems like other people have not really expressed an idea like this. Possibly it’s because nobody has really happened to think about it. It’s sort of a piece of low-hanging fruit in a sense. I think the way I view it is it’s kind of a retreat from content totalitarianism, so to speak, in the sense that rather than saying you’re going to have one ethics for everything, this is recognizing the fact that it has worked better in human society to not have that. It could be the case that we would have long ago evolved to the point where there’s only one country in the world. There’s only one system of laws in the world. It seems to have been, so far as we can tell, healthy for human society, just like it seems to be healthy in biological organisms, to have some diversity of different things going on. By the time it’s a pure mono species, monoculture-type thing, that tends to be much more fragile than something that has more diversity. In a sense, these things are far away… You ask how do these things relate to my personal ways to live my life paradigms and scientific paradigms. This particular kind of thinking about how to deal with differences in the world is something yet again cuts across those paradigms in a strange way. We’ll see whether I actually end up doing a serious work in this area. It depends on what really happens with the dynamics of companies and so on. For me it’s sort of interesting because it’s another slice, another kind of paradigmatic area that sort of addresses different kinds of questions for me than other ones have done.
You just had your 60th birthday. I would like to invite you for an interview on September 9th of 2029. I would ask you what do you expect what will then be the prior ten years – what would you have done? What would you have accomplished? What you have hoped for?
It’s hard to say. What’s interesting about the stage I’m at right now is that I’ve worked on things for a long time. The things I’ve done, I’ve been working on projects for 30, 35 years and more. A bunch of those projects feel like they’re coming to fruition right around now. The whole computational intelligence, computational language thing feels like it’s coming to fruition in the sense that I feel pretty good about where we’ve managed to reach in terms of being able to encapsulate computational intelligence, being able to have a computational language that can express lots of kinds of thinking and so on. I’m feeling pretty good about that.
I feel pretty good about the in the trenches software engineering of being able to deliver all of those capabilities to lots of different places in the world. My goal, I suppose, is to take these ideas and make them ubiquitous. In today’s world everybody expects a computer will have an operating system and a UI and things like that. I’d like to make what we built with computational intelligence and computational language similarly ubiquitous. The dynamics of doing that are things I personally don’t like to spend my time on. There are lots of dynamics about business deals and mechanics of distributing things and so on.
I’m very confident that a bunch of things I’ve already done will be considered important and will be surviving ideas and pieces of technology. For me to predict exactly when they are going to be like everybody is doing that, everybody knows about that, it’s very hard for me to predict that. That depends on things in the world that I don’t know. The idea of things we call notebooks we had 33 years ago, 32 years ago. People have been using our technology for a long time, but finally a few years ago it was like oh, these are amazing. Everybody thinks they’re so cool. That was a 28-year gap.
There are things I’ve done in science, in technology, and the way we built our computational language where it feels inexorable to me that eventually these things will be important. There are also plenty of things where I say that’s amusing, but it’s not going to be a surviving thing. It’s very hard for me to predict in a ten-year span exactly which of these things are going to get to the point where the world says that’s wonderful.
Do I care about the world saying that’s wonderful? For my ego, I don’t particularly care. For what it means is possible in the world or even is possible for me to do, I do care. Those are the things that enable all sorts of things that I think are really interesting to happen in the world when something becomes more ubiquitous or whatever else.
For me personally, I’m going to try to write more. I’ve done a fair amount of writing over the last few years like the things that are in this Adventures of a Computational Explorer book. I feel like people seem to find some of what I write interesting. I find it interesting if I look at it again. I feel like I should write more. I’m hoping to write more.
There are a bunch of things that are – I’m a little disappointed that the world hasn’t just done these without me. There are things, for example, like being able to get people to learn computational X for all possible values of the field X, so to speak, without having to learn sort of the details of computer science. We now have the technology through computational language to do that, to let people not have to major in computer science just to be able to use computation in the world. I think it’s going to fall upon me to write the textbook for that and so on. Probably that’s going to be a worthwhile exercise because it will cause me to think more clearly about a bunch of things and to invent some things I wouldn’t have otherwise invented. It feels like a little bit more of a service task for the world than something that is really going out and creating new things. I still think it’s important and going to be worthwhile, and I hope to do it.
The two big projects that I’m hoping to do in addition to more writing and things like trying to define the path for computational X, one is make a serious assault on the fundamental theory of physics. The other is try and build this symbolic discourse language to represent all the kinds of things we humans tend to talk about in precise symbolic computational form. These projects are definitely century-scale projects in the sense that the fundamental theory of physics, it’s not like people haven’t been thinking about that for awhile. This is something where I’ve been interested in it for a long time. If you’d asked me 40-something years ago when I was still doing physics for a living, did I think I really had a serious idea about the fundamental theory of physics, I would have said, “No, not particularly.”
I’m doing physics where I’m pushing forward things based on the framework and the structure of physics that already exist. If you said to me, “Can you see how to make a truly fundamental theory?” I would say, “No. I don’t really have any good ideas.” In the intervening years I have had ideas, which I think have potential. I don’t know if they’re right. I don’t if it’s the right century to try and do it.
I want to put a certain amount of effort into that. I don’t want to spend all my time doing that because I think it’s a thing where I might spend all my time and it’s just a pure lose, although I think it’s probably inevitable that the mathematical structures that emerge will have some ancillary value as things like string theory have had – ancillary value from their mathematical structures from the point of view of physics, they haven’t necessarily gotten that far.
In terms of symbolic discourse language, that’s kind of a 300-year story. People last seriously tried to do this in the 1600s. I feel like it’s time to try and do that again. The main difficulty and the main reason I haven’t done that for 40 years since I first started thinking about it is I haven’t had good use cases. I think now there are some use cases like computational contracts, like AI constitutions, like this AI preferences mechanism. There are use cases which help give a concrete purpose to trying to do this. That’s something I hope to work on.
I have hobbies. A funny thing about my life is there are things that I do that start as hobbies and they wind up becoming real things. I used to do business as a hobby. Then I wound up as a CEO. I used to do various sorts of science things as a hobby, etc. I used to do history stuff as a hobby, and then I’ve ended up doing books about it and so on.
If I try to predict based on my current hobbies, one of my current hobbies is education. We’ve had our school for the last 17 years. We’ve had our summer camp for high school students for the last six years or something. I have personally every week have been doing little classes with middle school kids about computational kinds of things. I’ve had fun doing that.
Based on my past experience with hobbies, my hobbies tend to turn into real things. I certainly toy with starting some educational institutional structure, although I have to say the details of doing that don’t appeal much to me. It might end up happening. Another thing that will probably affect the course of my next decade is our company has been quite successful, which is nice. There are things that may happen that may produce big bursts of success and big windfalls and so on. If some of those happen, then they’re probably things that – it’s like if we have some large amount of money to burn, there are things I might consider doing that I wouldn’t consider doing otherwise.
Going to Mars or something like that.
I’m not into going to Mars. It’s sad and interesting to me because I was looking back. One of the things I decided to do for my 60th birthday, I thought how am I going to spend that day? I can’t do my standard meetings and standard things I normally do on that day. I decided I would break it into two halves. I would have half of it as a nostalgia day and half of it as a future day. My half that was my nostalgia day consisted of looking on a live stream just for fun.
I’ve scanned about a quarter million pages of documents that I produced in my life. I was just going and looking back at ones I made when I was 10 years old, 12 years old, things like that. I’ll tell you one shocking thing that I found; when I was 12 years old I gave a speech. I was looking at this thing, and it was not a terrible piece of public speaking. I’d written it all out.
What was really shocking to me was what the topic was. The topic was basically AI ethics. It was interesting because some of what I said I still agree with. Some of what I said I don’t agree with. It was just so bizarre to me that when I was 12 years old, that was the topic I ended up giving a speech about.
I also found all kinds of strange things. I found a little essay I wrote about climate change when I was ten years old. That was kind of bizarre. I had no idea that in 1970 anybody, let alone me, was thinking about that.
One thing I’ve noticed about you is on the one hand you’re a really patient guy. You are big about teaching and people ask you very basic questions. You answer them enthusiastically. You really try to meet people where they are. Then other times you express vitriol towards incompetent people, especially that you meet in business, that you’re angry at because of their gross incompetence. Who gets in column A and who’s relegated to column B in that line up?
Interesting question; it’s funny because a lot of my life has been about actually trying to get things to happen. There are situations in which I am sort of locked into getting something to happen. When I’m in that state and something gets in the way and it’s some horrible incompetence and so on, then I push really hard to get something to happen. When I’m talking to some kid and trying to explain something, it’s not like I’m trying to get something to happen. I’m just living in that moment, so to speak. I think that’s the distinction.
At some moment I lock in. It’s like I’m going to try to get something to happen. What I’ve discovered is it’s pretty hard to get things to happen in the world. I built a great company and structure for getting things to happen. I’ve gotten to the point where I’m well connected in the world. It’s still pretty hard to get things to happen. The only way to get things to happen is to push pretty hard. I think when I’m in the get things to happen phase mode, I push pretty hard
There’s a Far Side comic strip that’s got this guy sitting on an overstuffed couch that’s busted open. He’s wearing a t-shirt and eating a chili dog. There’s a single light bulb burning above him. He’s watching TV. The caption is “Giorgio Armani at home.”
I wonder about you. Do you do anything that’s mindless? Do you watch cat videos on YouTube. When I read this book it all seems about finding efficiency and doing things and getting things done. You’re clearly productive beyond any scale. Do you ever do anything that’s absolutely just…
I think I can answer this one. This is one of the few questions that I can answer in a short way. The answer is no.
In your list of things you carry with you, you put hand sanitizer. Are you a germophobe?
Not really a phobe. I like to be practical about those things. The only things that’s interesting about that is that I started doing that long, long before it was widespread. I figure what the heck? It’s easy. It’s one of these things where it’s easy to do something that improve one’s quality of life. Why not do it? If it was really hard to do, I wouldn’t do it. It’s easy, so I do it.
I want to come back to your question about mindless activities. I tell a silly story about myself. It is really rare that I’m not engaged in something that in – it’s very rare that I’m just hanging and thinking. To tell a story against myself, a few years ago I happened to end up getting some MRI. It was one of these cases where you’re lying in this MRI machine and you really can’t do anything except be there. I was realizing this was an incredibly rare event for me. Usually I’m talking to somebody. I’m typing on the computer. I’m somehow connected to some sort of means of output or something. I realized, “What on earth am I going to do?” I started trying to compute from first principles, various things. The MRI machines, the big ones, go thunk, thunk, thunk. I was trying to compute various characteristics of the MRI machine and trying to see what I could reproduce. I finally just managed to get there from Bohr magnetons and trying to work it out, remembering values of Planck’s constant and all this kind of stuff. I just managed to figure out what the frequency should be and the MRI was done.