Latest / Merryn Talks Money / What Career Should You Choose in the AI Age?

Transcript

What Career Should You Choose in the AI Age? · Merryn Talks Money

• 0:02
  Speaker 1: Bloomberg Audio Studios Podcasts Radio News Music Welcome back to
• 0:19
  Speaker 1: Merrin Talks Money, the podcast in which people who know
• 0:21
  Speaker 1: the markets explain the markets. I am Merrin Somerset Webb.
• 0:24
  Speaker 1: This week, I'm speaking with Daniel Susskind, an economist and
• 0:27
  Speaker 1: one of the leading thinkers on how technology and artificial
• 0:30
  Speaker 1: intelligence are changing our world of work. Daniel is a
• 0:33
  Speaker 1: professor at Gresham College and King's College London and a
• 0:36
  Speaker 1: researcher at Oxford University and the Stanford Digital Economy Lab.
• 0:40
  Speaker 1: He has spent the last 15 years studying technology and work,
• 0:43
  Speaker 1: and his previous books include A World Without Work and Growth,
• 0:47
  Speaker 1: A Reckoning. Now, regular listeners to Merrin Talks Money will
• 0:50
  Speaker 1: know that I often ask our guests, what should our
• 0:53
  Speaker 1: children be studying? What kind of work will survive the
• 0:56
  Speaker 1: rise of AI? And how should young people prepare for
• 1:00
  Speaker 1: what is coming? And how can we help them prepare
• 1:02
  Speaker 1: for what is coming? Well, well done, Daniel. He has
• 1:05
  Speaker 1: written a whole book about exactly that. Not because we
• 1:08
  Speaker 1: asked him to, but just because I guess it's interesting.
• 1:10
  Speaker 1: It's called What Should My Children Do How to Flourish
• 1:12
  Speaker 1: in an Age of AI? It asks what the next
• 1:14
  Speaker 1: generation should learn, which skills will matter, and how much
• 1:17
  Speaker 1: we can prepare for a future we can't predict. Daniel,
• 1:20
  Speaker 1: welcome to Maren Took's Money.
• 1:22
  Speaker 2: Thank you so much for having me. Right.
• 1:25
  Speaker 1: We are very lucky because we have got the perfect
• 1:27
  Speaker 1: hook for our conversation, which is that the PISA education
• 1:30
  Speaker 1: numbers came out this morning. And I'm sure that you've
• 1:33
  Speaker 1: looked at them as well. We are talking, by the way,
• 1:35
  Speaker 1: on the 8th of September. They came out this morning
• 1:38
  Speaker 1: and they were really depressing. They showed a global decline
• 1:42
  Speaker 1: pretty much across the board in maths and reading. They
• 1:46
  Speaker 1: were even down from what we saw in 2022. And
• 1:49
  Speaker 1: when we saw depressing numbers for reading and mathematical skills
• 1:53
  Speaker 1: around the world in 2022, we went, nah, COVID, it'll
• 1:56
  Speaker 1: get better. It'll be fine. But it isn't. It hasn't
• 2:00
  Speaker 1: got better. It's got worse. And I know that all
• 2:03
  Speaker 1: generations look at the one below them and go, well,
• 2:06
  Speaker 1: they're useless. It's getting worse. And Normally it isn't, but
• 2:10
  Speaker 1: these numbers suggest that there is something going on.
• 2:14
  Speaker 2: I think that's right. I think that's right. I mean,
• 2:16
  Speaker 2: the numbers appear actually early on in the book, but
• 2:19
  Speaker 2: only up to 2022 because it's global data and it
• 2:22
  Speaker 2: takes the OECD a good few years to gather it together.
• 2:30
  Speaker 2: Many people have pointed to the particularly sort of precipitous
• 2:33
  Speaker 2: fall in literacy, but it's also numeracy as well. I
• 2:39
  Speaker 2: think it's a disaster in itself, but it's a particular
• 2:42
  Speaker 2: disaster for thinking about how we prepare the next generation
• 2:46
  Speaker 2: for the age of artificial intelligence. I think there's two
• 2:49
  Speaker 2: reasons why. One is, and this is a theme that
• 2:53
  Speaker 2: runs through the book, and you alluded to it in
• 2:58
  Speaker 2: your opening comments there, which is that it is incredibly
• 3:01
  Speaker 2: difficult and impossible to predict the future. The way in
• 3:05
  Speaker 2: which we have traditionally, I think, thought about preparing the
• 3:08
  Speaker 2: next generation for the future of work is to try
• 3:12
  Speaker 2: and think as cleverly and as deeply as we can
• 3:15
  Speaker 2: about what's going to happen in the future and what
• 3:17
  Speaker 2: jobs will have to be done and then reason back
• 3:20
  Speaker 2: and say, well, look, these are the sort of future-proof
• 3:22
  Speaker 2: skills that we need to teach the next generation. I
• 3:25
  Speaker 2: just think that's too difficult to do. We just can't
• 3:28
  Speaker 2: predict future. with confidence, particularly in this extraordinarily uncertain world,
• 3:32
  Speaker 2: exactly what more advanced skills, perhaps it's creativity, perhaps it's judgment,
• 3:37
  Speaker 2: perhaps it's problem solving, are really going to be in demand.
• 3:41
  Speaker 2: But I think what we do know is that whatever
• 3:44
  Speaker 2: those advanced skills turn out to be, they are going
• 3:47
  Speaker 2: to rely in some way on people having those basics
• 3:53
  Speaker 2: in place. Now, those basics are basics, not simply in
• 3:55
  Speaker 2: the sense that Literacy and numeracy are the most fundamental skills,
• 3:59
  Speaker 2: but they're also basic in the sense that they're the
• 4:00
  Speaker 2: building blocks. They're the foundations for everything else. And so
• 4:03
  Speaker 2: I think that is one reason that I think it's
• 4:07
  Speaker 2: particularly worrying, that whatever skills do turn out to be valuable,
• 4:13
  Speaker 2: they are going to rely on these basics.
• 4:16
  Speaker 1: Would you blame technology for that? this decline in literacy
• 4:21
  Speaker 1: and mathematical ability, would you look at that and say, well,
• 4:24
  Speaker 1: this is social media, this is children not paying attention,
• 4:28
  Speaker 1: or is it something's gone wrong in teaching? I mean,
• 4:31
  Speaker 1: there's always fads, aren't there, in the way that children
• 4:33
  Speaker 1: are taught and there's different ways that children are taught.
• 4:35
  Speaker 1: And if you drill down into the numbers that we've
• 4:38
  Speaker 1: seen globally, there's something quite interesting that's happened to England
• 4:42
  Speaker 1: where mathematical and literary skills have remained fairly static. And
• 4:47
  Speaker 1: there's been no devotion to any particular type of educational
• 4:50
  Speaker 1: fad in England. It's stuck to basic old-fashioned ways of teaching,
• 4:55
  Speaker 1: give or take. And so England now finds that it's
• 4:58
  Speaker 1: moved from number 20, 25 in the world in most
• 5:01
  Speaker 1: things up into being in the top 10 in almost
• 5:04
  Speaker 1: every educational skill. Whereas at the same time, we've seen
• 5:07
  Speaker 1: devolved areas of the UK where maybe there's been something
• 5:10
  Speaker 1: more faddish going on, Scotland, Wales, where we've seen declines.
• 5:14
  Speaker 1: So it suggests that that could be part of it globally,
• 5:16
  Speaker 1: a different way of teaching. But another part of it,
• 5:19
  Speaker 1: it seems fairly obvious to adults anyway, that it is
• 5:21
  Speaker 1: social media related.
• 5:22
  Speaker 2: I think there's two really interesting things here. One is,
• 5:24
  Speaker 2: I think technology does play a role. I don't think
• 5:29
  Speaker 2: it's a coincidence that this decline in literacy and numeracy
• 5:32
  Speaker 2: begins in 2009. 2009 is the year that young people's
• 5:36
  Speaker 2: lives start to become saturated with smartphones and social media
• 5:40
  Speaker 2: and they're sort of distracted and attention's fragmented. And in
• 5:48
  Speaker 2: the book, I am very critical of social media. I
• 5:52
  Speaker 2: have lots of worries about social media. I think many
• 5:55
  Speaker 2: of the ideas that are gaining traction for restricting children's
• 5:58
  Speaker 2: access to social media, things like no smartphones before 14,
• 6:05
  Speaker 2: no social media before 16, phone-free schools, lots of play.
• 6:09
  Speaker 2: I think all of these are terrific things. But I
• 6:11
  Speaker 2: think there is something really important though to remember, which
• 6:14
  Speaker 2: is that social media is not the same thing as
• 6:17
  Speaker 2: artificial intelligence. And one of the worries that I have
• 6:21
  Speaker 2: is that legitimate concerns about social media are being bundled
• 6:26
  Speaker 2: into conservatism and fears about artificial intelligence. So sensible restrictions
• 6:33
  Speaker 2: on social media are leading to knee-jerk bans and things
• 6:36
  Speaker 2: like chat GPT. And I think that's a mistake because
• 6:40
  Speaker 2: whereas I think social media does disempower and distract us.
• 6:44
  Speaker 2: Artificial intelligence, if used in the right way, has the
• 6:47
  Speaker 2: potential to make our lives go vastly better. And that's
• 6:51
  Speaker 2: part of what the book is. It's a big part
• 6:54
  Speaker 2: of what the book is about. There's another way to
• 6:58
  Speaker 2: put this, which is that I think at the moment
• 7:02
  Speaker 2: we have trapped ourselves in a debate about what the
• 7:06
  Speaker 2: right amount of screen time is. Do we want more
• 7:09
  Speaker 2: screen time or do we want less screen time? And
• 7:11
  Speaker 2: we're sort of collectively searching for the sort of Goldilocks
• 7:14
  Speaker 2: level of screen time, not too much, not too little,
• 7:16
  Speaker 2: just right. And my view is that actually that's the
• 7:18
  Speaker 2: wrong debate. What we should really be debating is what
• 7:21
  Speaker 2: is actually on these screens? How are we using this technology?
• 7:26
  Speaker 2: And if we are using it right and using it
• 7:28
  Speaker 2: to prepare the next generation for this world that they're
• 7:30
  Speaker 2: about to enter, then I think that's a really, really
• 7:32
  Speaker 2: good thing.
• 7:33
  Speaker 1: One of the reasons I think that people may be
• 7:35
  Speaker 1: confused or are concerned about the connection between AI and
• 7:37
  Speaker 1: social media is because, of course, AI has invaded social media. So,
• 7:43
  Speaker 1: you know, kids are now seeing AI slop all over
• 7:45
  Speaker 1: the place, which I think keeps them moving through reels
• 7:48
  Speaker 1: faster and faster and faster and through larger chunks of
• 7:51
  Speaker 1: time than they might have pre-AI. So there is a
• 7:54
  Speaker 1: connection there. You can see where the danger signals are
• 7:57
  Speaker 1: for parents and for teachers.
• 7:58
  Speaker 2: Absolutely. Clearly the distinction isn't a sharp one, but I
• 8:02
  Speaker 2: think it's sharp enough to say there is something materially
• 8:04
  Speaker 2: different from using AI as a personal tutor to browsing
• 8:08
  Speaker 2: on TikTok. I mean, one of the big worries that
• 8:12
  Speaker 2: I explore in the book is around technological distraction. It's
• 8:18
  Speaker 2: not a new worry, almost continually. as new technologies have arrived,
• 8:22
  Speaker 2: people have worried that they were going to distract us.
• 8:26
  Speaker 1: From the day of the first printing press.
• 8:27
  Speaker 2: Quite.
• 8:28
  Speaker 1: Books are dangerous for people. Books are dangerous. And then
• 8:30
  Speaker 1: the radio was dangerous. And then the television was dangerous, right?
• 8:33
  Speaker 1: And social media was dangerous.
• 8:35
  Speaker 2: That's right. You know, I went back in time and
• 8:36
  Speaker 2: looked at different technologies. So even the introduction of the stamp,
• 8:39
  Speaker 2: for instance, the penny stamp, which made postage something that
• 8:43
  Speaker 2: was far more affordable and no longer just something for
• 8:46
  Speaker 2: the kind of, you know, aristocratic class in Britain to
• 8:49
  Speaker 2: enjoy each other, you know. What was fascinating was a
• 8:53
  Speaker 2: worry that people were going to be distracted by all
• 8:57
  Speaker 2: this tittle-tattle in the letters that they received. You might
• 9:01
  Speaker 2: receive up to 12 deliveries of posts a day, and
• 9:05
  Speaker 2: people were terrified that there was just going to be
• 9:09
  Speaker 2: too much gossip and too much sort of idle reflection
• 9:12
  Speaker 2: in all of these letters. Exactly right. It continues up
• 9:16
  Speaker 2: to today. But the serious point is that I think
• 9:19
  Speaker 2: there is a real... part of the book is arguing
• 9:22
  Speaker 2: that we need to have an AI curriculum. Um, and
• 9:26
  Speaker 2: we need to be spending, I mean, just to give
• 9:29
  Speaker 2: you a flavor of quite how ambitious I think we
• 9:31
  Speaker 2: need to be. I think we need to be spending
• 9:33
  Speaker 2: about a third of our time in education, teaching people
• 9:35
  Speaker 2: how to use these technologies critically. But, but part of
• 9:38
  Speaker 2: that is also teaching them about distraction. Um, I think
• 9:42
  Speaker 2: it's one of the great challenges with these technologies. And, um, yeah,
• 9:48
  Speaker 2: one of the things that I think some people might
• 9:49
  Speaker 2: find surprising is that the, paying attention or not being distracted,
• 9:54
  Speaker 2: the more we learn about it is, that it is
• 9:56
  Speaker 2: a skill that can be learned. The other thing is,
• 10:01
  Speaker 2: of course, that these technology companies have done an extraordinarily
• 10:09
  Speaker 2: effective job of leveraging the body of work that we
• 10:12
  Speaker 2: have been building up over the last few decades into distraction,
• 10:16
  Speaker 2: habit forming, and so on, and sort of turned it
• 10:18
  Speaker 2: against us. with these technologies. And so I think part
• 10:22
  Speaker 2: of it is also about educating the next generation on
• 10:24
  Speaker 2: the nature of distraction and how these technologies are catching
• 10:28
  Speaker 2: their attention and not letting it go.
• 10:47
  Speaker 1: The other thing, when we're talking about teaching AI and
• 10:50
  Speaker 1: how to use AI, when we go back to the
• 10:52
  Speaker 1: basics you were talking about earlier of critical thinking, it
• 10:55
  Speaker 1: would seem to me that that is one of the
• 10:58
  Speaker 1: most important skills that we need as long as we're
• 11:01
  Speaker 1: dealing with AI in the form of LLMs as opposed
• 11:04
  Speaker 1: to anything else, because LLMs, as we know, are very
• 11:06
  Speaker 1: error-prone hallucinations, we call them. So It is very, very frequent,
• 11:12
  Speaker 1: and particularly the longer you go on with your prompts
• 11:14
  Speaker 1: and your tokens, the more likely you are to have
• 11:16
  Speaker 1: hallucinations and errors come through. So if we are encouraging
• 11:20
  Speaker 1: people to use LLMs to help them study, we also
• 11:24
  Speaker 1: need to be sure that we give them the tools
• 11:26
  Speaker 1: that they need to understand how and where these errors
• 11:29
  Speaker 1: will creep in and how best to recognize them. That
• 11:32
  Speaker 1: seems to me to be a huge part of AI learning.
• 11:35
  Speaker 2: I think that's exactly right. And, you know, that again,
• 11:38
  Speaker 2: that is a big difference with the calculator, which is
• 11:40
  Speaker 2: that you don't worry the calculator is going to run
• 11:42
  Speaker 2: out of steam and make a mistake in the fourth
• 11:44
  Speaker 2: decimal place. You know that, you know, what you get
• 11:47
  Speaker 2: if you put the question in correctly is the correct answer.
• 11:50
  Speaker 2: That's not the case with AI. And so we have
• 11:54
  Speaker 2: to be teaching people to make critical use of these technologies.
• 11:57
  Speaker 2: And again, this isn't just, you know, the book is
• 11:59
  Speaker 2: called What Should My Children Do? But it's far more
• 12:01
  Speaker 2: than that. It's a book about how we prepare everyone
• 12:04
  Speaker 2: to flourish in this new technological world. But I think
• 12:12
  Speaker 2: there are kind of playful and interesting and imaginative ways
• 12:16
  Speaker 2: that we can teach that sort of thing. I was
• 12:19
  Speaker 2: thinking about, and I do this in the book, I
• 12:22
  Speaker 2: spend some time thinking about what would it look like
• 12:24
  Speaker 2: to teach subjects with AI, perhaps subjects outside of my domain.
• 12:27
  Speaker 2: So I'm a mathematician and an economist, and I have
• 12:30
  Speaker 2: a pretty good sense of what it would look like
• 12:32
  Speaker 2: to teach AI those subjects. But what would it look
• 12:35
  Speaker 2: like to teach, say, poetry? One of the consequences of
• 12:40
  Speaker 2: being a dad of three children, eight, five, and two,
• 12:43
  Speaker 2: is that I have very little time now to focus
• 12:47
  Speaker 2: on a novel. Large stretches of concentrated time. I found
• 12:53
  Speaker 2: myself over the last five, six years reading an awful
• 12:57
  Speaker 2: lot of poetry. The thing about poetry, of course, is
• 13:00
  Speaker 2: that you really need to read it in a group
• 13:03
  Speaker 2: with other people, have a kind of lecturer directing it.
• 13:06
  Speaker 2: It's often quite difficult to make sense of. And that's
• 13:08
  Speaker 2: actually one of the ways in which I've been using
• 13:10
  Speaker 2: AI as a sort of study partner to help me
• 13:12
  Speaker 2: make sense of it.
• 13:12
  Speaker 1: A one-on-one tutor.
• 13:14
  Speaker 2: Quite. And using AI as a personal tutor is a
• 13:21
  Speaker 2: big part of the book. But I was wondering, what
• 13:27
  Speaker 2: would it look like to teach poetry with AI? And
• 13:30
  Speaker 2: I thought, well, There's a very simple way where you
• 13:33
  Speaker 2: could study a familiar poem without AI, and then you
• 13:37
  Speaker 2: could use AI to help you understand an unfamiliar poem.
• 13:41
  Speaker 2: But perhaps there are more creative ways. Perhaps you might
• 13:44
  Speaker 2: be able to use AI to rewrite a poem in
• 13:48
  Speaker 2: the voice of another poet, or perhaps the same voice
• 13:52
  Speaker 2: of a poet but change the topic.
• 13:54
  Speaker 1: This is not going to make you popular with poets, Daniel.
• 13:58
  Speaker 2: Well, hold on a second, because then you might use
• 14:01
  Speaker 2: AI to gather all the commentary on a particular poem,
• 14:04
  Speaker 2: which I did. And you might say, well, hold on
• 14:07
  Speaker 2: a second, how do we know that these commentaries exist?
• 14:08
  Speaker 2: And you might use that as an opportunity for thinking
• 14:10
  Speaker 2: about hallucination, whether or not these pieces of analysis really
• 14:13
  Speaker 2: did work. And you might then say, well, let's get
• 14:17
  Speaker 2: the AI to analyze this poem. And you might say, well,
• 14:19
  Speaker 2: hold on a second, this AI isn't a human being.
• 14:22
  Speaker 2: How does that shape its, you know, the sort of interpretation?
• 14:25
  Speaker 2: You might use that as an opportunity to think about,
• 14:27
  Speaker 2: you know, what it is that human beings are uniquely
• 14:29
  Speaker 2: placed to do and what it is that, um, I, I,
• 14:33
  Speaker 2: so I'm not, I'm not suggesting that those rewritten poems
• 14:36
  Speaker 2: are in some way of greater aesthetic quality, but I
• 14:38
  Speaker 2: think it prompts a really interesting conversation again, to your
• 14:41
  Speaker 2: point about the, this, when we're teaching AI, it's not
• 14:44
  Speaker 2: just about how do you write an efficient prompt, but
• 14:46
  Speaker 2: it's this sort of broader conversation about what this, what
• 14:50
  Speaker 2: this technology is, what it can do, what it can't do,
• 14:52
  Speaker 2: what it does things, what are its strengths, its limitations. And, and, um,
• 14:57
  Speaker 2: I mean, those were the sorts of things I found
• 14:59
  Speaker 2: myself grappling with and, and it's the sort of thing that,
• 15:01
  Speaker 2: you know, again, it's not my domain, but it's the
• 15:03
  Speaker 2: sort of thing that, um, I would love, uh, uh,
• 15:08
  Speaker 2: you know, people teachers uh in each of these different
• 15:12
  Speaker 2: areas to be thinking about because i think you know
• 15:14
  Speaker 2: with their expertise they could come up with incredibly exciting
• 15:17
  Speaker 2: and imaginative ways of putting this technology to use in
• 15:20
  Speaker 2: their particular particular discipline.
• 15:23
  Speaker 1: Finally something to cheer teachers up listen we um we've
• 15:26
  Speaker 1: spent quite a long time talking about what education should
• 15:29
  Speaker 1: look like yes what we would like or you would
• 15:31
  Speaker 1: like the future to look like back to basics and
• 15:34
  Speaker 1: test both, include AI education, et cetera. But it's not
• 15:38
  Speaker 1: there yet. That's not what we've got now. So if
• 15:41
  Speaker 1: you are a young person today, let's say you're coming
• 15:43
  Speaker 1: up to choosing A-levels, for example, what do you choose?
• 15:49
  Speaker 1: You say we can't predict the future. I absolutely agree
• 15:51
  Speaker 1: with you. We have no idea what's coming next. And
• 15:54
  Speaker 1: it's very hard to tell what very specific skills will
• 15:57
  Speaker 1: be useful. It's hard to say, well, this is future-proofed
• 16:00
  Speaker 1: and this isn't. Everyone keeps telling young people to go
• 16:02
  Speaker 1: out and learn a skill, be a plumber, be an electrician,
• 16:05
  Speaker 1: be a baker, whatever. But we don't even know that realistically.
• 16:08
  Speaker 1: So if you were sitting there, you're 15 and a half, 16,
• 16:13
  Speaker 1: looking at your A-levels, you've got to choose three subjects.
• 16:15
  Speaker 1: What are they going to be? Yeah.
• 16:18
  Speaker 2: So I think this question, what are they going to be,
• 16:20
  Speaker 2: is really interesting. I mean, there's so much advice I
• 16:25
  Speaker 2: would want to give. Give it. That's what we're here for.
• 16:29
  Speaker 2: I think one of the biggest pieces would be I
• 16:33
  Speaker 2: think the way in which we prepare people for work
• 16:36
  Speaker 2: is to say, what job do you want to do?
• 16:39
  Speaker 2: And then to say, because you want to do that job,
• 16:41
  Speaker 2: these are the skills and capabilities you need. And I
• 16:44
  Speaker 2: think actually that approach is going to lead to disappointment.
• 16:49
  Speaker 2: So what I write about in the book is that
• 16:50
  Speaker 2: actually you've got to ask the question, not what job
• 16:52
  Speaker 2: do I want to do, but what problem am I
• 16:55
  Speaker 2: interested in spending my life solving? If you go into
• 16:58
  Speaker 2: law because you want to be the sort of lawyer
• 17:00
  Speaker 2: you saw on Suits or you go into medicine because
• 17:02
  Speaker 2: you want to be the sort of doctor you saw
• 17:03
  Speaker 2: on House, you're going to be bitterly disappointed because what
• 17:07
  Speaker 2: those jobs look like are going to change dramatically in
• 17:11
  Speaker 2: the coming years. But if you go into these worlds
• 17:13
  Speaker 2: because you're interested in, say, improving access to justice or
• 17:16
  Speaker 2: solving legal problems, you go into medicine because you're interested
• 17:18
  Speaker 2: in solving medical problems and you're far more agnostic. about
• 17:23
  Speaker 2: what that job might look like, then I think the
• 17:25
  Speaker 2: future is a very exciting one because these problems aren't
• 17:27
  Speaker 2: going away. It's just the way in which we solve
• 17:29
  Speaker 2: those problems are going to look very, very different. And
• 17:33
  Speaker 2: I think in part, I've not been surprised seeing the
• 17:38
  Speaker 2: kind of angry protests from young, particularly in the US,
• 17:41
  Speaker 2: kind of graduating students who have finished spending a huge
• 17:45
  Speaker 2: amount of money and a huge amount of time and
• 17:47
  Speaker 2: a huge amount of effort preparing to be a particular
• 17:50
  Speaker 2: type of, you know, have a particular job. And then
• 17:53
  Speaker 2: some tech boss stands up and says, no, that world's
• 17:55
  Speaker 2: going to disappear because of AI. Um, and I, I
• 17:58
  Speaker 2: think part of the problem is that, that this idea
• 18:02
  Speaker 2: of a job is just an unhelpful way, an unhelpful
• 18:05
  Speaker 2: way of kind of anchoring your expectations about what your
• 18:07
  Speaker 2: future is going to look like. I think we should
• 18:09
  Speaker 2: be encouraging people, young people to be far more focused
• 18:12
  Speaker 2: on what sort of problems are they interested in solving.
• 18:15
  Speaker 2: And I kind of, I caught a glimpse of this. Um,
• 18:17
  Speaker 2: this isn't a new, um, new observation, but it's one
• 18:22
  Speaker 2: that I think has become more important in the last
• 18:24
  Speaker 2: few years. One of the things that made me think
• 18:27
  Speaker 2: that this was going to become increasingly important was a
• 18:30
  Speaker 2: paper written in 2017 by a team at Stanford. It
• 18:34
  Speaker 2: was announcing the development of a system, which if you
• 18:38
  Speaker 2: gave it a photo of a freckle, it would tell
• 18:41
  Speaker 2: you as accurately as leading dermatologists whether or not that
• 18:45
  Speaker 2: freckle was cancerous. Today, AI-enabled diagnostics is pretty widespread, but
• 18:51
  Speaker 2: back then, 10 years ago, this was a pretty dramatic announcement.
• 18:55
  Speaker 2: It caused a stir, but what was particularly interesting about
• 18:58
  Speaker 2: it was the final co-author on the paper, who was
• 19:01
  Speaker 2: a man called Sebastian Thrun, who some listeners might know
• 19:05
  Speaker 2: he is not, in fact, a medical doctor, but he's
• 19:07
  Speaker 2: the guy who developed the first driverless car. Here is
• 19:10
  Speaker 2: a guy who knows absolutely nothing about medicine at all,
• 19:12
  Speaker 2: and yet he's able to use these systems to solve
• 19:14
  Speaker 2: a problem that would have required an extraordinarily talented dermatologist,
• 19:19
  Speaker 2: decades of experience and training and expense and insight. And
• 19:23
  Speaker 2: he was able to design a system that could solve
• 19:25
  Speaker 2: this problem. There is a sense in which these technologies
• 19:29
  Speaker 2: just don't have a lot of respect for the traditional
• 19:32
  Speaker 2: boundaries that we have drawn between different subjects and different
• 19:35
  Speaker 2: disciplines and different occupations and different jobs. And I think
• 19:39
  Speaker 2: we need to be frank with people about that and
• 19:41
  Speaker 2: just say, look, you have to focus on the problems
• 19:44
  Speaker 2: you're interested in solving. and be open to the job
• 19:47
  Speaker 2: that you have and the skills that you're required to
• 19:49
  Speaker 2: learn as time passes. just looking very different from the
• 19:52
  Speaker 2: sorts of things that you required in the 20th century
• 19:54
  Speaker 2: let's go.
• 19:54
  Speaker 1: Back to a level does that mean that when you
• 19:56
  Speaker 1: look at a levels you are for particularly focused again
• 20:00
  Speaker 1: on anything that teaches you critical thinking as opposed to
• 20:02
  Speaker 1: rote learning so any critical thinking and maybe even linguistics
• 20:06
  Speaker 1: because you know you're a your relationship with your agents
• 20:10
  Speaker 1: is a linguistic one, right? Going back to effective prompting,
• 20:14
  Speaker 1: that kind of subject?
• 20:15
  Speaker 2: I think it's a balance of two things. I think
• 20:17
  Speaker 2: partly it's, you know, back to basics. Are these subjects
• 20:20
  Speaker 2: helping you, you know, refine and improve those basics, which
• 20:24
  Speaker 2: I think are so important. But there's also something just
• 20:26
  Speaker 2: more mercenary, which is, you know, what is going to
• 20:28
  Speaker 2: prepare me best for the particular, either subjects I want
• 20:32
  Speaker 2: to do at university or, you know, if you're not
• 20:35
  Speaker 2: thinking of university as for other paths. There's a more
• 20:38
  Speaker 2: sort of transactional approach to A-levels as well that you
• 20:40
  Speaker 2: just have to confront. Yeah.
• 20:42
  Speaker 1: I had someone on the podcast the other day, and
• 20:45
  Speaker 1: we were talking about what should you learn, what should
• 20:48
  Speaker 1: you think about. And we were talking about quantum computing.
• 20:52
  Speaker 1: And he said that what we should all be doing
• 20:55
  Speaker 1: is learning philosophy, spending a lot of time reading and
• 20:58
  Speaker 1: learning about philosophy because the existential crisis that is going
• 21:03
  Speaker 1: to come to us all When quantum computing is really
• 21:06
  Speaker 1: a go, combine that with the problems that AI will
• 21:08
  Speaker 1: solve and our minds will be so blown we're going
• 21:10
  Speaker 1: to need philosophy to fall back on.
• 21:12
  Speaker 2: Yes. So I'd say I'd go back to critical thinking,
• 21:16
  Speaker 2: which in a sense is sort of philosophy applied, which
• 21:20
  Speaker 2: is philosophy is critical thinking applied to the kind of
• 21:22
  Speaker 2: the biggest questions. But just the reason I wrote this book,
• 21:27
  Speaker 2: What Should My Children Do?, was because I've spent the
• 21:30
  Speaker 2: last 10 or 15 years, as you said, writing about
• 21:33
  Speaker 2: the impact of technology on work, and often from those
• 21:36
  Speaker 2: bigger macro questions. What does it mean for work? What
• 21:39
  Speaker 2: does it mean for growth? What should the state do?
• 21:42
  Speaker 2: What I have found time and again, as I travel
• 21:45
  Speaker 2: around talking to very different groups of people, very different places,
• 21:47
  Speaker 2: is that people come and say, look, that's all very
• 21:49
  Speaker 2: interesting about this coming quantum apocalypse, but I've got three
• 21:54
  Speaker 2: kids at home. What on earth should they do? What
• 21:57
  Speaker 2: I'm really trying to do is just bring some of
• 21:59
  Speaker 2: these ideas back down to earth and say what all
• 22:03
  Speaker 2: of us, parents, individuals, teachers, policymakers, politicians can do if
• 22:08
  Speaker 2: we're thinking about preparing the next generation. What about us?
• 22:12
  Speaker 1: What about me? I mean, we're all so obsessed with
• 22:16
  Speaker 1: the kids, right? The next generation, them, them, them.
• 22:19
  Speaker 2: What about me, me, me? What do the adults do?
• 22:22
  Speaker 1: What do the adults do.
• 22:23
  Speaker 2: To prepare ourselves? Right, because one of the chapters in
• 22:26
  Speaker 2: the book is Don't forget the adults. I know, I
• 22:29
  Speaker 2: read it. Yeah, yeah, yeah. Because I think it's so important. Yeah.
• 22:34
  Speaker 2: Because these challenges aren't simply for the next generation, they're
• 22:38
  Speaker 2: for this generation too. And just to come back to
• 22:41
  Speaker 2: a word we used at the start, uncertainty, there is
• 22:43
  Speaker 2: a huge amount of uncertainty about the future, about exactly
• 22:47
  Speaker 2: what jobs are going to have to be done, what
• 22:48
  Speaker 2: skills are going to be most important. I think this
• 22:51
  Speaker 2: uncertainty is perhaps the defining feature of the age of
• 22:53
  Speaker 2: artificial intelligence. But the best response any of us really
• 22:57
  Speaker 2: have to uncertainty is flexibility, a capacity and a willingness
• 23:02
  Speaker 2: to retrain and reskill, particularly later in life, with the
• 23:05
  Speaker 2: same intensity and seriousness that people engage in education at
• 23:09
  Speaker 2: the start of their life. And I just don't think
• 23:12
  Speaker 2: we're anywhere near that yet. So really the call in
• 23:17
  Speaker 2: that part of the book is really saying we just
• 23:19
  Speaker 2: need to take adult education far more seriously. We need
• 23:23
  Speaker 2: to sort out the quantity of of money that we're spending,
• 23:26
  Speaker 2: both individuals, both states, companies, in helping people retrain and reskill.
• 23:32
  Speaker 1: So what should I go and do a course in?
• 23:37
  Speaker 1: Philosophy aside.
• 23:38
  Speaker 2: One of the things that I find particularly interesting and
• 23:43
  Speaker 2: exciting is the how. If you were to step into
• 23:46
  Speaker 2: a time machine and go back 100 years and walk
• 23:48
  Speaker 2: into a classroom or a lecture theater, it looks remarkably
• 23:51
  Speaker 2: similar to how it does today. despite the existence of
• 23:54
  Speaker 2: lots of technological alternatives. But I think a turning point
• 24:01
  Speaker 2: has now come after the sort of disappointment of education
• 24:05
  Speaker 2: technology over the last few years, where the technology just
• 24:08
  Speaker 2: hasn't been there. But I think it is now able
• 24:11
  Speaker 2: to do something quite interesting. And it's this, which is
• 24:16
  Speaker 2: we know that one-to-one tuition with a human tutor is
• 24:19
  Speaker 2: extraordinarily effective. So an average student who receives one-to-one tuition
• 24:22
  Speaker 2: with a human being will tend to outperform 98% of
• 24:26
  Speaker 2: students in a traditional classroom setting. Incredibly effective. And I
• 24:31
  Speaker 2: spent the best part of a decade teaching economics and
• 24:34
  Speaker 2: mathematics at Oxford. I've seen how effective the tutorial system
• 24:38
  Speaker 2: can be where I sat with one, two, three students,
• 24:41
  Speaker 2: tailoring what I taught to the particular strengths and weaknesses
• 24:44
  Speaker 2: of of the people in front of me. But the
• 24:46
  Speaker 2: problem with that, of course, is that it's just unaffordable.
• 24:49
  Speaker 2: We haven't been able to offer that kind of personalized
• 24:51
  Speaker 2: tuition to everybody who wants it. Now with these systems,
• 24:54
  Speaker 2: you really can. And the kind of personalized interaction that
• 24:57
  Speaker 2: you can have with these systems in trying to understand
• 25:00
  Speaker 2: new and unfamiliar problems really does rival, in my view, 95%
• 25:08
  Speaker 2: of teachers that I've seen. And I include myself in that.
• 25:10
  Speaker 2: They are incredibly effective. helping you understand new problems and
• 25:14
  Speaker 2: new ideas. If you tell them who you are, what
• 25:17
  Speaker 2: you're trying to understand, what it is that you want
• 25:19
  Speaker 2: to do. And this is a roundabout way of saying,
• 25:22
  Speaker 2: find a problem or an issue or a question that
• 25:25
  Speaker 2: you're interested in and just run with one of these
• 25:27
  Speaker 2: systems and get it, treat it like a tutor and
• 25:31
  Speaker 2: ask it to help you understand. Perhaps it's in an afternoon,
• 25:35
  Speaker 2: perhaps you want it to design a program for you
• 25:37
  Speaker 2: over a few months or experiment and test and explore
• 25:41
  Speaker 2: with these systems because they really are remarkable in the
• 25:47
  Speaker 2: way that they can offer this personalized interaction. Yeah, I
• 25:50
  Speaker 2: advise to give people one piece of advice, whoever they are,
• 25:53
  Speaker 2: whatever they do, it's just to kind of glimpse what
• 25:56
  Speaker 2: these technologies can now do. It's to find a problem
• 25:58
  Speaker 2: that interests you or something you'd like to understand more
• 26:00
  Speaker 2: about and try and use these technologies to do that
• 26:03
  Speaker 2: with you. I mean, just an example of something that
• 26:06
  Speaker 2: I've been doing. um, uh, just in the last few weeks, it's,
• 26:10
  Speaker 2: it's not in the book, but it's something that I've, I've,
• 26:13
  Speaker 2: I've been, uh, I've, I've been spending my time doing, uh,
• 26:16
  Speaker 2: my grandfather had this beautiful watch, which, um, um, my,
• 26:21
  Speaker 2: my mom, his daughter took to a repair person and said, um,
• 26:26
  Speaker 2: can you repair this? And he said, no, it can't
• 26:28
  Speaker 2: be repaired. And she suspected that something was up that
• 26:30
  Speaker 2: he was just, you know, wasn't sort of being entirely truthful.
• 26:33
  Speaker 2: And so, I've taken it upon myself now as a
• 26:36
  Speaker 2: sort of, to become a horologist with the help of
• 26:38
  Speaker 2: chat GPT. So I, I, you know, sit there with
• 26:41
  Speaker 2: the system alongside me. It gave me guidance and all
• 26:44
  Speaker 2: the tools that I needed. It told me to, you know,
• 26:46
  Speaker 2: I ordered a sort of, uh, a, a simple, uh,
• 26:50
  Speaker 2: mechanism from China to experiment on. So I could learn
• 26:53
  Speaker 2: the basics of how the movement works. Um, I, and
• 26:57
  Speaker 2: now I, yeah, I've, I've now begun on this, this
• 27:00
  Speaker 2: watch that my grandfather apparently unfixable. And, I'm in this
• 27:05
  Speaker 2: weeks-long exercise, sending it photos, sending it photos through the microscope,
• 27:10
  Speaker 2: gradually tinkering and tweaking it. The idea that I could
• 27:13
  Speaker 2: become a horologist without going for a kind of four-week
• 27:16
  Speaker 2: residential camp somewhere in the north of England to understand
• 27:19
  Speaker 2: how watches work and house repentance, unimaginable until very recently.
• 27:22
  Speaker 2: And now it's something that I'm able to do. So
• 27:27
  Speaker 2: watch this space to see if I manage to.
• 27:29
  Speaker 1: Two points on that. One, worrying implications for employment globally
• 27:33
  Speaker 1: if everyone can become their own horologist as well as
• 27:35
  Speaker 1: their own accountant or lawyer.
• 27:37
  Speaker 2: We haven't yet seen if I can fix the watch.
• 27:39
  Speaker 1: I haven't got that.
• 27:40
  Speaker 2: So let's check in.
• 27:41
  Speaker 1: Second, I would like to voice the thoughts of every
• 27:43
  Speaker 1: single listener and viewer, which is he's never going to
• 27:46
  Speaker 1: finish that job with three kids under 10 in the house.
• 27:49
  Speaker 2: Yeah. Yeah. So we're making … It's never going to happen.
• 27:52
  Speaker 2: In full disclosure, we're making slow progress. So … We're
• 27:55
  Speaker 2: making slow progress, but maybe I'll post some videos of
• 27:58
  Speaker 2: my progress on social media.
• 28:00
  Speaker 1: The very fact that it's possible is fascinating. Daniel, one
• 28:03
  Speaker 1: last question before you go, because I've taken up too
• 28:05
  Speaker 1: much of your time. What are you reading at the moment?
• 28:07
  Speaker 1: You've written this wonderful book. Everyone's going to read it.
• 28:09
  Speaker 1: But what are you reading yourself?
• 28:13
  Speaker 2: One of the issues that really interests me is AI
• 28:17
  Speaker 2: creativity and intellectual property. Um, and in fact that I
• 28:21
  Speaker 2: have a new series of lectures in my role, I'm
• 28:24
  Speaker 2: the professor of business at Gresham college in, in, in London.
• 28:27
  Speaker 2: And I do a series of public lectures each, each year. Um,
• 28:32
  Speaker 2: last year was on AI and its impact on work.
• 28:35
  Speaker 2: This year is on AI creativity and copyright. I just
• 28:38
  Speaker 2: think it's completely fascinating, you know, set of issues and tensions. And, and,
• 28:42
  Speaker 2: and so I've actually been reading a lot from, uh,
• 28:46
  Speaker 2: the early internet era. uh, people like, um, Larry Lessig,
• 28:51
  Speaker 2: the Harvard law professor and Yochai Benkler also at Harvard who, uh,
• 28:56
  Speaker 2: you know, the copyright regime is a consequence of this
• 28:58
  Speaker 2: sort of ongoing battle between technology and, um, our ability
• 29:04
  Speaker 2: to create and share new ideas and be creative in society. And,
• 29:07
  Speaker 2: and there's a really interesting analogy, I think, between what's
• 29:09
  Speaker 2: going on in the world of artificial intelligence at the
• 29:11
  Speaker 2: moment and what happened at the start of the internet era,
• 29:13
  Speaker 2: particularly with respect to, you know, sharing music and, sharing
• 29:17
  Speaker 2: videos and the copyright battles then. So I'm just sort
• 29:19
  Speaker 2: of immersing myself back in that period a few decades
• 29:22
  Speaker 2: ago and in particular reading some of Larry Lessig's wonderful
• 29:26
  Speaker 2: early work on those issues. Excellent.
• 29:28
  Speaker 1: And can all our listeners come and listen to your
• 29:31
  Speaker 1: public lectures?
• 29:32
  Speaker 2: Yeah, absolutely. Not quite that public. No, no, no, absolutely.
• 29:34
  Speaker 2: So this is a wonderful thing about Gresham is that
• 29:37
  Speaker 2: I give six entirely public lectures a year. So you
• 29:40
  Speaker 2: can either come in person to Gresham College, which is
• 29:44
  Speaker 2: just in the city of London. If you can't attend
• 29:47
  Speaker 2: in person, you can watch them online as well. It's
• 29:50
  Speaker 2: a fantastic thing. And if AI creativity and intellectual property
• 29:54
  Speaker 2: is not your thing, there's a set of other professors
• 29:57
  Speaker 2: in fascinating... fascinating other disciplines too. Lots of great public lectures.
• 30:03
  Speaker 1: Okay, brilliant. Daniel, thank you very much.
• 30:05
  Speaker 2: Such a pleasure.
• 30:05
  Speaker 1: Great to have you on.
• 30:06
  Speaker 2: Thank you for having me.
• 30:14
  Speaker 1: Thanks for listening to this week's Meron Talks Money. If
• 30:16
  Speaker 1: you like our show, rate, review, and subscribe wherever you
• 30:19
  Speaker 1: listen to your podcasts. And keep sending questions or comments
• 30:22
  Speaker 1: to meronmoney at bloomberg.net. You can also follow me and
• 30:24
  Speaker 1: John on Twitter or XM at meronSW. And John is
• 30:27
  Speaker 1: John underscore Stepak. This episode was hosted by me, Meron
• 30:31
  Speaker 1: Somerset Webb. It's produced by Sam Asadi and Moses. And
• 30:34
  Speaker 1: special design by Blake Maples. And special thanks, of course,
• 30:37
  Speaker 1: to Daniel Suskind.