00:00:14.000 Throughout history, mankind has struggled to understand life's mysteries.
00:00:21.000 Our guest is a multiple award-winning pioneer in the field of quantum computation
00:00:26.000 and argues that explanations have a fundamental place in the universe.
00:00:30.000 They have unlimited scope and power to cause change,
00:00:34.000 and the quest to improve them is the basic, regulating principle,
00:00:38.000 not only of science, but of all human endeavor.
00:00:41.000 This stream of ever-improving explanations has infinite reach.
00:00:46.000 We are subject only to the laws of physics, and they impose no upper boundary
00:00:50.000 to what we can eventually understand, control, and achieve.
00:00:54.000 He applies that worldview to a wide range of issues and unsolved problems
00:00:59.000 from creativity and free will to the origin and future of the human species.
00:01:05.000 We welcome Fellow of the Royal Society, a pioneer in quantum computing,
00:01:09.000 visiting professor of physics at the Center for Quantum Computational Expert University,
00:01:14.000 multiple TED talker, optimist, and author of the beginning of infinity,
00:01:19.000 explanations that transform the world. David Deutsch, welcome to the show.
00:01:26.000 I love what you say, wherever there has been progress.
00:01:29.000 There have been influential thinkers who denied that it was genuine,
00:01:32.000 that it was desirable, or even that the concept was meaningful.
00:01:36.000 You argue that all progress, both theoretical and practical,
00:01:39.000 has resulted from a single human activity, the quest for what you call good explanations.
00:01:44.000 That is the basis of the book, and that's why I wrote the book.
00:01:48.000 It's about good explanations, basically, and the many ramifications of that concept.
00:01:53.000 This idea has been contradicted in many different ways,
00:02:00.000 So an explanation is a statement about how the world is,
00:02:05.000 and how it behaves and why, and the why part always involves explaining
00:02:13.000 what we see in terms of what we don't see, explaining what we know is there,
00:02:22.000 And sometimes that changes our mind about what we know is there.
00:02:26.000 So that's an explanation, but most explanations are bad.
00:02:30.000 And in the book, I argue that good explanation is one that is hard to vary
00:02:38.000 while still explaining what it purports to explain.
00:02:43.000 And therefore, the epitome of a bad explanation is one that could explain anything.
00:02:49.000 If you say, for example, you know, conspiracy theory,
00:02:53.000 it's an example of a bad explanation, because whatever happens,
00:02:57.000 if the events turn out bad, you can say, well, the conspiracy made them bad.
00:03:02.000 If events turn out good, you can say, well, the conspirators are just
00:03:07.000 abiding their time and lulling us into a full sense of beauty, and so on.
00:03:12.000 Historically, religions have been quite fertile sources of bad explanations.
00:03:19.000 Also, a few good ones, by the way, I'm not entirely opposed to religious traditions,
00:03:33.000 they are the alternative to just plain prediction.
00:03:37.000 And I think plain prediction is not really science at all,
00:03:42.000 because if you say that X happens whenever Y happens,
00:03:49.000 It could be that X calls Y or that something else calls both of them,
00:03:53.000 or that since you don't understand what X and Y really are,
00:03:58.000 it could be that you're misinterpreting the whole thing.
00:04:02.000 On the other hand, when you have an explanation,
00:04:06.000 when you say that not only is malaria caused by living near to swamps,
00:04:13.000 but it's caused by the mosquitoes that live in the swamp
00:04:18.000 and put something into you when they draw your blood,
00:04:23.000 then you have an explanation that is not only useful,
00:04:28.000 What this book did to me was made me question the history of knowledge in a way.
00:04:37.000 is there is nothing more deceptive than an obvious fact.
00:04:41.000 And that got me thinking that knowledge is built on a starting point.
00:04:52.000 You'll have seen me quoting the philosopher Karl Popper.
00:04:55.000 And one of his maxims is that the starting point doesn't matter.
00:04:59.000 The reason that it doesn't matter is quite profound.
00:05:14.000 So if knowledge consisted of building up from a secure foundation,
00:05:19.000 then if there was anything wrong with the foundation,
00:05:21.000 the rest of the structure might fall down at any moment.
00:05:29.000 which is giant tottering edifice with some good stuff,
00:05:43.000 And what we do is we don't try to build it up more securely.
00:05:48.000 We try to find places where it seems to be wrong.
00:06:05.000 We guess that senseer would be better than we try it.
00:06:09.000 Sometimes that's trying it experimentally or practically.
00:06:25.000 So, I mean, it's very to jump on your initial thing.
00:06:50.000 even what you just said there is very empowering
00:06:52.000 because you say the misconception is that knowledge needs
00:07:04.000 but you're writing to understand your own thoughts
00:07:23.000 That's the great thing about this theory of knowledge.
00:07:32.000 Not only does the knowledge not require foundation,
00:07:36.000 that people with the knowledge have no authority.
00:08:51.000 no, you're not going to learn anything from that
00:09:26.000 your effort to make it simple for people to understand
00:09:35.000 and businesses and open to trying new business models
00:10:02.000 Everything has to revolve around conjecture and criticism.
00:10:23.000 And this is what the rational approach to politics
00:10:32.000 as ways of facilitating conjecture and criticism
00:10:48.000 that one should judge a political institution by,
00:11:06.000 We have to expect that our policies at any one time
00:11:15.000 And our institutions, the institutions that we use
00:11:26.000 and that kind of thing will also be riddled with errors.
00:11:51.000 Which is a maximal innovation or iteration in any case.
00:12:04.000 you have to be willing to let them skin their knees
00:12:35.000 and therefore trying to create an error-free system,
00:12:40.000 and it actually just entrenches existing errors.
00:12:46.000 warn children about the mad dog outside the gate,
00:12:52.000 try to make a system in which no bad thing will ever happen,
00:13:02.000 which is the better we come to understand phenomenon remote
00:13:06.000 the longer those chains of interpretation become,
00:13:09.000 and every additional link necessitates more theory,
00:13:14.000 Yes, so this is another refutation of the traditional idea
00:13:18.000 that science is about just seeing that being open
00:13:25.000 and to just induce the regularities from what we see.
00:13:42.000 and there's no such thing as direct experience.
00:13:49.000 And what we've done to gain our best access to reality
00:13:54.000 is we have not brought ourselves closer to experience.
00:14:00.000 We've put better and better theories in between us
00:14:06.000 That's how we know by getting some output traces
00:14:12.000 from a radio telescope that there are black holes and quasars
00:14:20.000 We can tell all that, even though the thing we're actually
00:14:24.000 looking at is a piece of metal and plastic that's here on Earth.
00:14:32.000 we know that the behavior of that particular metal and plastic
00:14:36.000 is intimately related to the behavior of things
00:14:44.000 to all our experience, yet we know about it via the theories.
00:14:50.000 encourages the building of the tools in order to be able to
00:14:53.000 get closer to the proving or disproving of the theory.
00:14:59.000 It's always disproving that theories are never going to be proved.
00:15:15.000 then the future of humanity in the world of artificial intelligence.
00:15:18.000 Because in some ways, we look at artificial intelligence and go,
00:15:22.000 that's going to replace us and we're all going to be jobless.
00:15:28.000 And it's why I said in the introduction that you're an optimist
00:15:32.000 That actually, we can almost outsource the road tasks of
00:15:40.000 experimentation or theory proving to artificial intelligence
00:15:44.000 and then actually work on the important part which is the thinking.
00:15:47.000 Yes, it's important to distinguish between two very different things,
00:15:51.000 almost opposite things, which are both called artificial intelligence.
00:15:56.000 One of them is the things we have today, like Siri
00:16:00.000 and the path and recognition and Google algorithms and all those things.
00:16:08.000 historical reasons because they do things that at one time,
00:16:16.000 But they are not what would be required to replace a human
00:16:27.000 That is, creative thinking, the creation of new explanations,
00:16:33.000 That is a completely different task which unfortunately,
00:16:39.000 I think that the theory is not yet good enough by a long chalk
00:16:43.000 for us to make an artificial general intelligence,
00:16:48.000 They will be made one day, but we don't know how to do it.
00:16:52.000 So, when I think of those things in two different categories
00:16:57.000 because they are almost opposite of each other,
00:17:00.000 a better AI is one that better meets your specification.
00:17:09.000 according to some criterion, or it plays chess better,
00:17:24.000 and can only play to win is not really human at all
00:17:31.000 And AGI will be able to play chess, perhaps if it wants to,
00:17:37.000 if it wants to, may or may or may or may not want to,
00:17:44.000 insisting on playing to win all the time is crazy.
00:17:52.000 The top 10 players in the world are something like that.
00:18:04.000 For most people, to play a game like chess or whatever,
00:18:13.000 something beautiful, something that you will learn from.
00:18:21.000 It might be learn how to understand opening theory better.
00:18:26.000 It might be learn how to construct a beautiful chess puzzle.
00:18:32.000 It might be learn how to explain the real meaning of chess position.
00:18:38.000 There's a million different ways of human or an AGI.
00:19:07.000 this is one of the things we don't teach people how to learn.
00:19:13.000 and wrote information can be done by artificial intelligence.
00:19:18.000 The future of education or education, as it should be.
00:19:22.000 The existing assumptions behind educational systems
00:19:27.000 are that the purpose of education is to transmit valuable knowledge
00:19:34.000 faithfully from one generation to the next, basically.
00:19:39.000 From people who already have that knowledge to people who don't.
00:19:42.000 So the knowledge is conceived of as a kind of valuable fluid,
00:19:49.000 which you pour from one generation to the next.
00:19:54.000 And if it doesn't work, it can only be because either you haven't poured it carefully enough,
00:20:09.000 It assumes that there's an authority for knowledge,
00:20:25.000 then you're allowed to form your own new theories of something,
00:20:29.000 or attack the problems at the edge of the subject.
00:20:32.000 But as I said before, there is no edge of subject.
00:20:35.000 My old boss, John Wheeler, used to say about physics,
00:20:51.000 or in the interior, which is the noun's essentials,
00:21:00.000 can be a good point, but also the previous generation passing on
00:21:04.000 its knowledge to the next, and expecting the next generation
00:21:08.000 Where I was getting out at the start when I mentioned
00:21:17.000 does not want to be attacked by the next generation.
00:21:26.000 actually, you know what, this may not be the best way
00:21:29.000 because the environment has changed so much that the establishment
00:21:35.000 but the previous establishment or those that are where the authorities
00:21:38.000 in that establishment don't want the next generation to attack us.
00:21:52.000 They are not open, they're not open to improvement.
00:21:56.000 And similarly, there are companies and institutions
00:22:04.000 But there are other, I don't think that's a tall true
00:22:07.000 of all companies or all participants in the economy.
00:22:16.000 to fulfill a vision of how things might be better.
00:22:26.000 If the company is founded by a person with that kind of thing
00:22:31.000 look, you could make more money if you do it this way.
00:22:44.000 You know, I've got enough gold-plated bathroom tabs.
00:22:56.000 to make a better computer, or how to make a better
00:23:01.000 transport system, or how to make a better space rocket,
00:23:08.000 Just like in science also, there are scientists
00:23:16.000 and enjoy their position of telling other people
00:23:27.000 But there are other scientists who don't care about that,
00:23:31.000 who only care about the problem that they're involved in.
00:23:48.000 A physicist, he said, is a ruthless opportunist.
00:23:58.000 What you're there for is to try to understand things better
00:24:01.000 and to find a problem that's interesting and beautiful
00:24:08.000 If you don't solve it, you'll still have had a fun time.
00:24:12.000 The problem should come first and the problem should be worthwhile.
00:24:18.000 Never mind the economics and the money and even the techniques
00:24:24.000 There should be something that would be worth existing
00:24:37.000 and setting up a particular kind of institution,
00:24:39.000 then all that is conditioned by the original problem.
00:24:44.000 Of course, the original problem can also change.
00:24:48.000 You may find a better problem, in which case you change to that.
00:24:51.000 One of the core ingredients for all this is creativity
00:24:57.000 because imagination and creativity are the core ingredients
00:25:01.000 to the future of humanity, but also business or innovation in any sense.
00:25:05.000 I agree, so creativity is human level creativity
00:25:22.000 It may have a sort of mechanical kind of creativity
00:25:31.000 but it's not creative in the same sense that humans are creative.
00:25:34.000 It's not creative in the sense that one can make progress with.
00:25:41.000 We know that it's some kind of computation going on in brain.
00:25:47.000 We know a lot about indirect, for example, as I said,
00:25:51.000 we know that it consists of conjecture and criticism.
00:25:57.000 So we create new conjectures, we create new criticism.
00:26:05.000 We don't know how to do it well enough to program it.
00:26:11.000 And that is the thing on which everything depends.
00:26:17.000 the way to make progress is to create better explanations.
00:26:22.000 In other words, to root anything that's unsatisfactory
00:26:27.000 about the world, any problem, any evil and so on,
00:26:37.000 And then think of ways of changing them to improve them.
00:26:46.000 Einstein found that he couldn't understand what would happen
00:26:53.000 if you rode on a photon, which is traveling at the speed of light
00:26:56.000 because Newton's law said one thing, and Maxwell's electrodynamics said another thing.
00:27:02.000 And so he tried to unify those and he succeeded with a different theory,
00:27:06.000 not at the first attempt, but probably something like the 500 attempt
00:27:11.000 and not just wild guesses, but successive improvements.
00:27:18.000 I mean, I haven't read a biography of Einstein, but it must have been like that.
00:27:22.000 And then he found that that theory, which is called special theory of relativity,
00:27:32.000 So he had to make further changes, and he worked on that for years and years,
00:27:37.000 and there I know that he went through several theories,
00:27:47.000 And by the way, the story that I have heard about this,
00:27:50.000 is just one of these apocryphal physics stories
00:27:53.000 that my physicist history rather than actual history.
00:27:57.000 But the story I heard was that when Einstein was trying to discover
00:28:03.000 the general theory of relativity, that is to unify special relativity with gravity,
00:28:09.000 he found certain problems, mathematical problems,
00:28:13.000 and he gave a lecture, and the great German mathematician David Hilbert
00:28:18.000 was in the audience, and he heard these problems,
00:28:21.000 and he went back home and he wrote down what we later called Einstein's equation.
00:28:26.000 But the lesson of this, but they're not called Hilbert's equation,
00:28:30.000 and the reason is that Hilbert did not know what he was doing.
00:28:35.000 He only knew the mathematics of what Einstein had asked for,
00:28:39.000 but he didn't understand why or what the symbols were meant,
00:28:44.000 and why that criterion rather than some other criterion.
00:28:49.000 So as the way I heard the story, that happened in 1913,
00:28:53.000 Einstein didn't actually come up with it himself until 1950.
00:28:57.000 I don't know why Hilbert didn't send in the answer,
00:29:09.000 It is really impossible to make progress in fundamental physics
00:29:14.000 without an explanatory understanding of the problem that you're trying to solve
00:29:19.000 and of the thing your proposing reality is like.
00:29:24.000 If you just work with the mathematics, you never get anywhere,
00:29:28.000 and I fear that many theoretical physicists don't follow that maxim.
00:29:40.000 so that the 1% inspiration theory, 99% perspiration,
00:29:45.000 the Edison quote, is a misleading description of how progress actually happens
00:29:49.000 because the perspiration phase can actually be automated,
00:29:57.000 we have been doing exactly that for tens of thousands of years.
00:30:07.000 we can spend our time doing this stuff we want to do
00:30:14.000 and, you know, at some point somebody invented shirts,
00:30:25.000 the shoes were doing a lot of the perspiration work
00:30:30.000 that the person would have had to do if he didn't have shoes,
00:30:34.000 like, you know, being very careful where you step,
00:30:40.000 and so on, and not being able to do certain things at all,
00:30:44.000 because what one can't do the relevant perspiration work,
00:30:50.000 that's the same, you know, if we come up to the 19th century,
00:30:55.000 then we were beginning to have sewage installed,
00:31:00.000 so people didn't have to do the chamber pot thing,
00:31:03.000 which itself had been an innovation at some time.
00:31:06.000 The sewage system does a tremendous amount of perspiration work,
00:31:21.000 There's that marvelous documentary by Mark Williams
00:31:40.000 So we'd all be going commando if it wasn't for the house.
00:31:46.000 We'll watch it and you'll see how disgusting it used to be.
00:32:02.000 and once enough knowledge has been embodied in the lunar colony,
00:32:05.000 our natural inclination to want to devote our thoughts
00:32:15.000 things like underpants that we take for granted,
00:32:48.000 I do not think about the vast amount of sophisticated effort
00:32:54.000 that went into making the road between here and there,
00:33:13.000 And it makes every day live less perspirational.
00:33:22.000 And so we can devote our efforts to the inspiration.
00:33:35.000 And it's something that I talk about on this show,
00:33:41.000 there's no search thing as passengers were all crew.
00:33:44.000 So we have to actually respect the ship we live on.
00:34:13.000 even though they're kind of contradict each other.
00:34:25.000 and it has this miraculous life support system,
00:34:56.000 is that when we've strained it past a certain point,
00:35:08.000 Now, I think this is wrong from beginning to end.
00:36:23.000 And so that's why billions of humans are now alive.
00:36:28.000 That's why not only are billions of humans alive now,
00:37:11.000 so that the total number doesn't increase exponentially.
00:37:24.000 people are not having children at the replacement level.
00:37:31.000 every couple has to have two points something children.
00:37:39.000 it would be exactly two that they would have to have.
00:37:46.000 some people don't reproduce for all sorts of reasons.
00:38:03.000 turns out that when people get prosperous above a certain level,
00:38:07.000 they don't want to have two points something children.
00:38:13.000 and a lot of people are satisfied with having one.
00:38:26.000 the central core of that idea is that the earth
00:38:43.000 Because of evolution and because of lucky accidents
00:39:04.000 from which people derive much the same conclusion
00:39:12.000 which says that there is nothing special about humans.
00:39:26.000 on the outskirts of a nondescript galaxy and so on.
00:39:30.000 This is trying to contradict the old religious idea
00:39:41.000 I think by the way, morally at the center of the universe
00:40:10.000 our explanations, our concept of what an improvement is.
00:40:22.000 There's no reason to believe that it can succeed any further
00:40:32.000 hunting game on the savanna and that kind of thing.
00:40:48.000 because our brains evolve for this other purpose
00:40:52.000 and we can't expect this extension to go on forever.
00:41:00.000 And so, therefore, we can't expect endless improvement.
00:41:07.000 then the priority is not error correction anymore.
00:41:15.000 The priority then becomes the orderly management of stasis.
00:41:30.000 The orderly management of stasis is not much better.
00:41:44.000 They both lead to this kind of anti-human attitude
00:41:51.000 What dawned on me when I was reading the book was,
00:41:57.000 I was thinking of your work and I was going to go and imagine
00:42:05.000 So my interpretation of it is based on everything I've read,
00:42:12.000 Imagine I'm actually formulating what I see based on what I know
00:42:18.000 because I'm really working on this part of my own world is to go,
00:42:29.000 I'd love to be able to remove the veil naturally.
00:42:32.000 And that's what I love about the idea of the multiverse
00:42:34.000 is that you can, at least you can be aware that there is a veil of reality.
00:42:44.000 and even more generally what human thought does.
00:42:59.000 of the people sitting in the cave with the shadows,
00:43:05.000 that they are caused by people outside the cave.
00:43:08.000 You can't see those people, but your best explanation
00:43:15.000 why they always happen just before they deliver food to you
00:43:31.000 Now there that's making life that isn't in here and so on.
00:43:38.000 It's explaining the scene in terms of the unsee.
00:44:00.000 The link is that there is an explanatory theory.
00:44:18.000 we had this idea was, was Owen Schrodinger a few years earlier
00:44:22.000 but Everett's the one with credited with the name
00:44:34.000 Unusually, we knew a lot about quantum phenomena
00:44:40.000 in terms of being able to calculate the answer.
00:44:50.000 But when I say crazy, I don't mean counterintuitive.
00:44:59.000 I mean they were crazy in the sense that they didn't make sense.
00:45:05.000 that when you become consciously aware of something
00:45:18.000 And okay, that's, in the point of view of physics,
00:45:50.000 from what if you say that the first observer observes something
00:45:53.000 and this is called the paradox of vignus friend
00:46:05.000 and he imagined that the friend was himself described
00:46:09.000 This being our most fundamental theory of nature.
00:46:24.000 many ways in which attempted explanations made no sense.
00:46:27.000 And ever produced an explanation that made sense.
00:46:32.000 So that's how the power of universe theory began.
00:46:36.000 And now we can use it to explain all sorts of other things,
00:46:41.000 like quantum computers or why quantum computers are so powerful.
00:46:48.000 When they do exist, why they will be so powerful.
00:46:52.000 It's because they share the work of computation
00:46:56.000 between many universes, vast numbers of universes.
00:47:10.000 it's not even like each chapter could be a book of this book.
00:47:14.000 Each paragraph could be a book in its own right.
00:47:16.000 The amount of research and the amount of knowledge
00:47:21.000 to help the reader understand these complex concepts.
00:47:25.000 But I loved the line, and I pulled this line out of the book
00:47:34.000 An imaginable, numerous environments in the universe
00:47:37.000 are waiting out there, and we humans could be that spark.
00:47:49.000 And this is how I envisage not just the human species,
00:48:11.000 that there is there are only two possibilities.
00:48:21.000 That would be that we are destined to run into a brick wall
00:48:28.000 which means that no human thought is ever affected again.
00:48:33.000 In the book, I don't just argue that that would be bad
00:48:37.000 therefore it can't be true. I argue from very fundamental
00:48:49.000 thinking there's a boundary to potential human knowledge
00:48:53.000 is the same as blind belief in the supernatural.
00:49:02.000 Therefore, if you don't want to be completely irrational,
00:49:05.000 you have to accept the idea that unlimited progress is possible.
00:49:26.000 and give us the best possible options for the future
00:49:31.000 And David, if people want to find out more about your work,
00:49:39.000 then there's my website, which has got lots of links
00:49:49.000 And then there's the construct of theory website.
00:49:54.000 but that is actually what I'm mostly working on
00:49:56.000 on the moment, a kind of the new theory of physics.
