David Epstein in Conversation with Malcolm Gladwell

00:00
David welcome to the 92nd Street Y is
00:01
this your this is your first time on
00:03
stage here is it I did the Q one time is
00:06
my first a oh really yeah the DA is more
00:11
fun than the Q I think I think so yeah
00:14
we I thought we would start by talking
00:17
about how we know each other yeah I
00:19
think that’s a I’ve wanted to do that
00:21
like at the end to make sure we did that
00:23
so can I say how we know each other yes
00:26
what we’ll give each give our version of
00:29
the events because I suspect it might be
00:31
different but you go first
00:32
okay my version is that in our
00:37
relationship our first date was I guess
00:39
me criticizing some of your work in my
00:41
first book yes and and and I remember
00:47
when you know not expecting that book to
00:49
do much I was at like a very small event
00:52
in Greenwich Village and somebody came
00:54
by and said you know I just saw reading
00:55
your book at a cafe Malcolm Gladwell and
00:57
I was like oh darn
00:59
I didn’t I didn’t think it would get on
01:00
your radar yeah and then our second date
01:04
was you critiquing me back in The New
01:06
Yorker also being very positive but also
01:08
critiquing me back well you know I hold
01:10
on ironic I wrote an article for the
01:13
magazine which was I mean it was the
01:15
warmest sweetest face of your book yes
01:17
and then I did a separate piece for the
01:20
website where I gently pushed back
01:26
against some of your more outrageous
01:27
assertions okay
01:31
and then pick up the story from here
01:34
okay then I was I am so gentle you are a
01:38
book tour with almost no book why was I
01:40
in well I was in Washington DC and I’m
01:42
going into NPR and then you come
01:45
swinging through the door me you did did
01:48
you not remember this we had it was in
01:50
the movies they would call this meeting
01:51
cute and then for some reason we started
01:55
running together yeah well no you
01:57
skipped over our third date yeah which
02:00
was the first time we actually met in
02:01
person which was at the MIT Sloan Sports
02:03
analytics conference oh that’s right
02:05
that was the first time we met in person
02:06
we were invited to do a debate that was
02:08
10,000 hours versus the sports gene yeah
02:11
and and in some ways my preparation for
02:15
that debate because you’re very clever
02:17
and I’d never met you I didn’t want to
02:18
get embarrassed yeah so I did a lot of
02:20
homework and that that debate in some
02:24
ways seeded some of the ideas for for
02:27
this for this book but but what I really
02:30
want to say about that debate was you
02:33
could very much have just like tried to
02:36
you know crush me or like using your
02:38
literary crowd or whatever it was clout
02:40
but instead we ended up having a great
02:42
conversation not only that but when we
02:44
came off the stage you told me what you
02:47
thought my good points were and said
02:49
when we’re back in New York tomorrow why
02:51
don’t you go running he said this was a
02:53
great idea you should explore that more
02:54
and for me this was in many ways like
02:59
the people I write about in chapter 10
03:01
of range where this could have become a
03:02
zero-sum thing which frankly with some
03:04
other authors that I came into conflict
03:06
with it did become a zero-sum thing but
03:09
in this case it wasn’t like those like
03:12
the foxes in chapter 10 you were willing
03:14
to update your mental models and I
03:15
learned from that and I think in some
03:17
ways it empowered me to take on a more
03:20
amorphous and ambitious book in this
03:21
project that I that I know isn’t perfect
03:24
but that I was willing to do because of
03:26
that and that’s sort of openness in
03:28
exchange I think made me better and I
03:32
think if that happened more you know we
03:34
can see in society there I think there
03:36
are too many conflicts that are viewed
03:38
as zero-sum ideas
03:39
we were both disincentivized from
03:41
agreeing about
03:41
thing in many ways but it made me better
03:44
that’s for me it’s kind of a model of an
03:45
intellectual relationship so I really
03:48
appreciate we do I mean I feel sorry
03:50
you’ll see because I have a I helped
03:56
this whole theory of love bombing which
03:58
was when you’re someone criticizes you
04:01
the only appropriate response is to love
04:03
them back even if you’re doing it
04:05
cynically because it completely disarms
04:07
us the last thing they’re expecting but
04:09
in your case I started out thinking all
04:11
this love bomb him and then I realized
04:12
actually he has convinced me so started
04:16
cynical and ended up totally idealistic
04:18
in the sense that I was like oh he’s
04:19
told me right I tried to love bottom and
04:22
failed because he actually won me over
04:24
so by the way not to put a critique one
04:27
of your theories but I’ve seen your your
04:30
responses to like like some of Christian
04:33
breeze criticism and I wouldn’t call
04:34
that love bombing I was worried I was
04:36
maybe a teeny we all stray off the
04:39
straight and narrow
04:40
but I do in the main I like to love by
04:43
critics but wait we have to get to the
04:46
point
04:46
what was it so we had this discussion
04:49
wasn’t a debate it was a discussion the
04:51
MIT Sloan conference and you said it’s
04:53
soda seeds for this book what was the
04:55
seed so in trying to anticipate what I
04:59
thought you would have to argue in this
05:01
debate I I said well you’ll have to
05:04
argue in favor of early specialization
05:06
in sports and so I went and looked at
05:09
all the research I could find about the
05:11
development of athletes and it showed
05:13
that this pattern that athletes who go
05:15
on to become elite have a sampling
05:16
period where they play a broad range of
05:18
sports they they gain these this broad
05:21
general skills that become a scaffolding
05:23
for later skills they learn about their
05:24
interest they learn about their
05:25
abilities they delay specializing until
05:28
later than their peers who plateau at
05:29
lower levels and and it’s not even just
05:32
a selection effect because when you
05:33
match kids in studies where they’re
05:35
matched for a certain ability level a
05:37
certain age and tracked the ones who in
05:39
a certain age do more variety of
05:40
different sports improve more by time to
05:42
basically and so I sort of brought that
05:44
up and you know in some ways that was
05:46
incompatible with with some aspects of
05:48
the 10,000 hour theory and so when we
05:51
were walking up the stage and we framed
05:53
it as the Roger verse tiger prop
05:54
right so pause on that point okay
05:57
build-out Roger versus Tiger because
05:59
there’s a beautifully simple way of
06:00
illustrating this argument okay
06:02
so Tiger Woods probably even even for
06:05
people who don’t know his story you’ve
06:07
probably absorbed at least the gist of
06:08
it which is 7 months old his father
06:10
gives him a putter not trying to train
06:11
me to be a golfer but just gives him a
06:12
putter
06:13
he starts carrying it around in his baby
06:15
Walker at 10 months he starts imitating
06:17
a swing he was physically precocious two
06:19
years old he’s on national television
06:20
two years old the CDC development
06:23
benchmarks are stands on tiptoes and
06:25
kicks a ball and he went on television
06:26
and showed his driving off in front of
06:29
Bob Hope basically by three his father
06:32
was media training him at four he
06:34
started hustling people basically you
06:36
know he’s famous as a teenager by 21
06:38
he’s the greatest golfer in the world
06:39
Roger Federer and maybe the most famous
06:42
development story in the history of
06:43
anything Roger Federer meanwhile played
06:47
about a dozen different skiing
06:49
skateboarding badminton tennis
06:50
basketball soccer all these things
06:53
mother was a tennis coach refused to
06:55
coach him because he wouldn’t return
06:56
balls normally she said it was no fun
06:58
when his coaches tried to bump him up a
07:01
level he declined because he just wanted
07:03
to talk about pro wrestling with his
07:04
friends after practice when he finally
07:07
got good enough to warrant an interview
07:08
with a local newspaper and the reporter
07:10
asked him if he ever became a pro what
07:12
he would buy with his first paycheck he
07:13
said in Mercedes and his mother was
07:15
appalled and asked if she could hear the
07:17
interview recording and and he’d
07:19
actually said Mayer CDs and a Swiss
07:20
German accent he just wanted more CDs
07:24
and so then she was like okay we’re
07:27
doing okay his father had no rules just
07:29
said don’t cheat don’t care anything
07:30
else and he specialized year he
07:33
continued playing badminton basketball
07:34
soccer specialized years after what is
07:37
Roger Federer really only playing tennis
07:40
mid teen years basically where he’s only
07:42
doing tennis but he still continues to
07:43
non formally play soccer even though
07:45
he’s doing that and in other informal
07:48
sports continues with them even after
07:50
that and the question basically was
07:52
which one of these models is the normal
07:54
but it’s one should we extrapolate over
07:55
why this is Tamayo is the fascinating
07:58
question so we have these two two of the
08:01
greatest athletes of the last yeah 50
08:03
years represent diametrically opposed
08:06
models of development one
08:08
one unknown yeah wasn’t story we’re in
08:10
love
08:11
with the tiger model if I pulled the
08:13
audience most of them would say the
08:15
tiger implicitly is is the model that
08:17
leads to greatness you’re arguing no
08:20
it’s the Roger model why it doesn’t
08:23
I wouldn’t thing I’ve never understood
08:24
is why did we fall in love with the
08:25
targa model and not like the Roger model
08:28
wait I thought you made us fall in love
08:30
with the tiger model don’t blame me
08:32
you’re I can just I did not write a book
08:35
about I just didn’t I no no no that’s
08:38
true that is that is very true that
08:41
ideas that you started became outrageous
08:44
in other hands in many cases that’s it
08:47
but I remember the Time magazine article
08:49
that was like unrecognizable about yeah
08:52
well it was that point I was positing
08:54
that there was another Malcolm Gladwell
08:55
walking around curly hair who had a set
08:58
of views that I would there were some
09:00
unknown to me but but in terms of tiger
09:02
as I think to steel it’s dramatic it’s
09:03
incredibly dramatic there’s a video of
09:05
him on YouTube at age two it makes a ton
09:07
of intuitive sense it’s very easy for a
09:09
prescription to tell people and I think
09:11
as you said we’re obsessed with
09:12
precocity right you said these child
09:14
prodigy videos are human cat videos and
09:16
I think that’s true and I’m mad I didn’t
09:18
think of that line for my book but is
09:20
that an is that enough though because
09:23
it’s also clear that Tiger pays an
09:26
extraordinary price for his precocity in
09:31
a way that Federer does not right in
09:33
fact it’s not difficult to reach the
09:35
conclusion that one of the reasons Tiger
09:37
had a kind of meltdown for many years is
09:39
that he really has been a prisoner of
09:41
golf since he was this high and one of
09:43
the reason Federer seemed so
09:45
well-adjusted is that he’s he had a
09:47
normal childhood he did he completely
09:49
had a normal childhood his his the
09:51
writer who probably knows the family
09:53
best called his parents pulley not not
09:56
pushy so he did have a very normal
09:57
childhood yeah so so even given the fact
10:00
that the Tiger model is costly we still
10:03
embrace it yes because well we’re
10:07
obsessed with excellence and I think so
10:08
if one of the themes in range I think is
10:11
that there are and maybe this doesn’t
10:14
apply to golf and we can talk about that
10:16
but that there are things that you can
10:17
do that cause head starts that actually
10:19
systematically honor
10:20
long term development but I think that
10:22
is a deeply counterintuitive idea and
10:26
when push comes to shove our intuition
10:28
is that getting ahead is getting ahead
10:30
and that that prodigious performance in
10:33
a child is a trajectory not just a
10:35
cross-section but that’s that’s often
10:37
not the case but but it’s also it’s just
10:38
it’s admirable to see someone want to
10:41
work that hard like I respect that in
10:43
them but and it’s intuitive that that
10:46
would work but also that’s you know one
10:48
of the reasons we do science is because
10:50
our intuition always figured it out so
10:53
let’s walk through the reasons why the
10:56
tiger model doesn’t work and as far as I
10:58
can tell from reading your book there’s
11:00
at least at least three if not more but
11:03
starting with explain walk us through
11:05
the match argument which is a really
11:07
interesting one which had never occurred
11:09
to me the so match quality is this term
11:11
that that economists use to basically
11:13
describe the degree of fit between an
11:16
individual’s abilities their interests
11:18
and the work that they do it turns out
11:19
to be incredibly important for
11:20
motivation for their performance right
11:24
and even their apparent grit so you get
11:26
good fit and it’ll look like grit when
11:28
someone does something when they’re when
11:29
they’re in something that fits correctly
11:30
and the problem is in in sports
11:33
selection this dovetails with something
11:35
you’ve written about the earlier it goes
11:37
the less likely you you optimize
11:39
someone’s match quality so one of the
11:41
things that happens when you delay
11:42
matching is you give people a chance to
11:44
get more signal about what they’re good
11:45
at and they end up picking better
11:47
matches for themselves and not just in
11:48
sports so so you know one of the other
11:51
studies and range looks at timing of
11:52
specialization in higher education and
11:54
the question the Economist asks is who
11:56
wins the trade-off the early specializes
11:57
or the late specialized errs and what he
11:59
finds is the early specialize errs do in
12:01
fact jump out to an income lead after
12:02
college but by year six the later
12:05
specialized errs who have picked a fast
12:06
a better match have a faster growth rate
12:08
fly past them and the early specialized
12:10
IRR start quitting and much higher
12:11
numbers because you know it’s like if we
12:14
treated those those decisions the way we
12:17
treated dating we would never pressure
12:18
people to settle down that quickly
12:20
before they took some more data about
12:21
things and so so so to pause on this
12:24
because I think this is a crucial point
12:27
the parent who says who observes of
12:30
their six-year-old that you know Lucy
12:33
enormous tea is really well coordinated
12:37
and flexible I want to make her a
12:39
gymnast yeah the mistake they’re making
12:42
is that you don’t know at 6:00 whether
12:44
Lucy is best is best cut out for
12:47
gymnastics and if you wait until 12 you
12:50
message may be a bad example here but if
12:52
you wait longer you have a better
12:53
likelihood of figuring out what her
12:55
skills match up with definitely six is
12:57
just too soon yes gymnastics is female
13:00
women’s gymnastics specifically is a
13:01
weird example because it requires a pre
13:03
puberty PDF what’s better but because
13:06
you know female gymnast has shrunk from
13:07
five foot three to four foot nine on
13:09
average in the last thirty years because
13:10
it makes their power to weight ratio
13:12
better and lower moment of inertia so
13:13
that’s a whole different advantage but
13:16
but you’re absolutely right so in Tiger
13:18
Woods by the way he said in 2000 his
13:20
father never asked him to play golf it
13:22
was him asking his father to play is the
13:24
child’s interest that matters and so I
13:27
think the idea that he was like father
13:28
manufactured from the get-go like you
13:31
shouldn’t be worried about missing the
13:32
next Tiger Woods because if there’s that
13:34
like incredible incredible sort of
13:36
outlier display of interest like that’s
13:39
not something that his father
13:40
manufactured from the get-go yeah so so
13:42
I think people are worried about missing
13:43
that but really what you should be
13:46
oriented toward his match quality and
13:47
there’s all sorts of reasons so you’ve
13:49
written about the relative age effect
13:50
right so I was just looking at the
13:53
breakdown of the birthdays of soccer
13:56
players in the u-17 under-17 European
13:59
Championships 47% of them were born in
14:02
January February and March and 6% in the
14:05
last three months of the year and that’s
14:06
because as we put selection earlier and
14:08
earlier all the coaches are selecting
14:11
four is the kids that are effectively a
14:12
year older and they are actually
14:14
biologically mature and they’re
14:15
mistaking that for talent and then
14:16
they’re in the pipeline you’ve
14:17
deselected the other kids and it’s
14:19
getting more and more exacerbated where
14:21
we’re picking for things that have
14:22
nothing to do with the traits you
14:23
ultimately want because we’re driving
14:24
selection earlier yeah so that’s one
14:28
argument that by and I can see actually
14:31
it’s it’s kind of fascinating to apply
14:33
that outside of sports as well so the
14:35
equivalent would be to observe of your
14:38
six-year-old that she has a facility for
14:40
counting and to put her immediately into
14:43
a pre math ph.d program that’s that
14:46
would be that but that’s exactly what
14:48
yes what parents are doing and in fact
14:49
what they’re encountering is a good
14:51
example because that what these things
14:53
that parents do are usually based on is
14:55
the observation of what’s called a
14:56
closed skill something like counting or
14:58
the kid walks early or something like
14:59
that and those kinds of closed skills
15:01
that aren’t these more general pieces of
15:04
scaffolding that that are good for long
15:06
term development there’s a fade-out
15:07
effect on those kinds of skills whether
15:09
they’re in sports whether they’re in
15:11
math lots of lots of academic programs
15:13
that are meant to give kids a boost
15:15
early on to get them on a different
15:16
trajectory and it does initially because
15:18
the way you can give them the fastest
15:20
improvement is by teaching them closed
15:22
skills that have to do with procedures
15:24
that are used over and over and over and
15:25
there’s a ubiquitous fade out effect
15:28
which is actually just other people
15:29
catching up because everyone’s gonna
15:30
learn that skill eventually and it
15:32
ceases to become an advantage and so we
15:34
make choices based on precocity in these
15:35
closed skills in many cases that are not
15:38
really in the long term in advantage so
15:40
second argument is that in order to
15:43
excel at a complex skill in the long
15:46
term you need to build a broad base yeah
15:48
so walk us through that both I’m
15:51
interested in this one this one’s even
15:52
more relevant outside the sports realm
15:54
yeah yeah but give us both sports and
15:57
non sports in this instance yeah so
15:58
wonder if this is me we should introduce
15:59
the issue of the kind and wicked
16:01
learning environment basically which are
16:03
which are terms taken from a
16:06
psychologist named Robin Hogarth and and
16:08
one of the reasons there’s a real lack
16:10
of study in Gulf which is interesting
16:11
but one of the reasons I can believe
16:13
that early specialization may in fact
16:14
work in golf although the best player in
16:16
the world right now Brooks kepta picked
16:18
up golf later because he got in a car
16:20
accident and his parents didn’t wanna do
16:21
contact sports anymore but is a time
16:25
learning environment is where all the
16:27
information is available next steps are
16:30
totally clear people often wait for each
16:32
other to take turns patterns repeat
16:35
feedback is automatic and totally
16:38
accurate after everything you do so golf
16:40
is almost like an industrial task you
16:42
try to do known movements over and over
16:43
and over with as little deviation as
16:45
possible that’s a kind learning
16:46
environment so is give me other examples
16:48
chess is a kind learning environment so
16:50
it’s based the grandmasters advantage in
16:53
chess is basically patterned recognized
16:55
and that also is the reason why it is so
16:57
amenable to automation because computers
16:59
are even better at pattern recognition
17:00
so the the kinder in environment is you
17:03
know Golf is like entertainment and we
17:05
still watch people playing chess because
17:06
it’s entertainment but the more of a
17:08
kind environment the skill is the more
17:10
easy it is to automate which is why you
17:12
know now that your iPhone app can like
17:15
beat Garry Kasparov on the wicked and
17:18
our challenge is where the the rules may
17:22
not be clear people are acting in
17:24
real-time they’re more dynamic you may
17:26
or may not get feedback after everything
17:28
you do next steps aren’t always clear
17:29
the feedback may be delayed or may be
17:31
inaccurate so hogarth use this example I
17:33
love of a famous New York City physician
17:35
who became renowned because weeks before
17:37
patients would develop typhoid by
17:39
palpating their tongue feeling around
17:40
their tongue with his hands he could
17:42
predict that they would get typhoid
17:43
right again and again again and as one
17:45
of his colleagues eventually observed he
17:47
was a more productive carrier of typhoid
17:48
than typhoid mary using just his hands
17:50
and so
17:54
yeah that’s a really wicked learning
17:57
environment because the feedback yeah
17:59
wait that wasn’t even another joke
18:02
the the feedback teaches exactly the
18:05
wrong lesson and so he gets famous for
18:07
this feedback loop that teaches him the
18:08
wrong lesson most environments aren’t
18:09
that wicked either but the closer you
18:11
are to the wicked end of the spectrum
18:13
the more you have to do what’s called
18:14
transfer where you take knowledge and
18:15
skills and have to apply them to
18:17
situations you have never seen before so
18:19
this more repetitive using procedures
18:21
knowledge then can become an impediment
18:23
because you’re stuck doing the same
18:25
things when you really have to transfer
18:27
to situations you haven’t seen before
18:28
starting a business would be wicked
18:30
starting a business would be would be
18:32
wicked and I think that’s one reason why
18:34
like if there was some recent research
18:36
from LinkedIn that showed like people
18:38
who who become successful executives one
18:40
of the best predictors is the number of
18:41
job functions they’ve worked across
18:43
within an industry or again to go to
18:45
this obsession with precocity when Mark
18:47
Zuckerberg was 22 and he said young
18:49
people are just smarter and MIT
18:51
northwestern in the Census Bureau just
18:53
has research out showing that the
18:54
average age of a founder of a
18:56
blockbuster startup on the day of
18:57
founding not even when it becomes a
18:59
blockbuster is about 46 yeah but like
19:02
the tiger story we just focus on the
19:04
Zuckerberg story but actually people
19:05
have to zigzag usually quite a bit
19:06
before they find that that grant cuz the
19:09
goal isn’t initially clear like it is in
19:11
in kind learning environment yeah yeah
19:13
so it’s odd that the the the kind of
19:17
myth of precocity and the the idea that
19:21
the tiger bottle is so important is
19:22
relevant only in the areas that were
19:24
least interested in that’s you right I
19:28
should have talked to you before I wrote
19:30
some of the lines in the book yeah no
19:32
that’s that’s yeah very clever way to
19:34
phrase it and and that’s so one of the
19:35
things that I was critiquing in range
19:37
was you know books in in the did we both
19:41
read sort of in the performance genre
19:43
have used the tiger model is the most
19:46
popular model from which to extrapolate
19:48
to everything else in the world like
19:50
literally I think talent is overrated
19:51
the back cover says Tiger Woods the
19:54
polgár’s the chess family and this is
19:57
what works for anything that you care
19:58
about in fact it’s that leap that is a
20:01
problem it may work in golf but it’s the
20:03
extrapolation where we’ve made
20:05
mistake yeah yeah the the let’s let’s
20:10
talk a little bit about this this notion
20:12
of in sports so what’s a good example of
20:16
a wicked sport I don’t think any sports
20:19
are that wicked I think because they’re
20:21
all rule-bound so what Hogarth said is
20:23
either tennis is more wicked than in
20:25
golf
20:26
yeah because tennis set is dynamic you
20:28
have to you have to use so-called
20:29
anticipatory skills where the sport is
20:31
actually happening too fast for you to
20:33
react to so you need to learn to pick up
20:35
cues and a player’s body and you know
20:37
the ball and things like that to act
20:39
faster than you could otherwise so it’s
20:41
much more dynamic but it’s still you
20:44
know in soccer really one of the reasons
20:46
why applying like Moneyball stuff to
20:48
soccer has been so difficult because
20:49
it’s so fluid and the game changes so
20:51
much that analytics hasn’t made nearly
20:53
as big an impact there is like baseball
20:55
where something happens and it stops and
20:57
something happens and it stops and so
20:59
analytics have made a much better impact
21:02
but hogarth then says what we’re really
21:04
mostly playing in the world the things
21:06
we most care about our Martian tennis
21:07
where you can see people doing something
21:10
but nobody’s told you the rules it’s up
21:12
to you to deduce them and by the way
21:14
they can change at any time and that’s
21:16
that’s these these other challenges that
21:18
we mostly are about what about I was
21:20
thinking that there’s a third reason why
21:22
you would want to take a generalist
21:27
course supposed to a specialization
21:29
course and that is that you it is only
21:32
through taking a generalist approach
21:34
that you could have novel skills not
21:37
only but the chief reason you’d want to
21:39
need one it like so I’m thinking of this
21:41
in basketball that you know every now
21:43
and again there’s someone like a Hakeem
21:44
Olajuwon or Steve Nash these brilliant
21:47
basketball players who have strong
21:49
grounding in soccer yeah and that’s very
21:51
rare among basketball players but we say
21:54
of those who come to basketball late
21:55
from soccer that they have certain
21:57
skills that are unusual yeah or at least
22:00
they have developed certain skills much
22:02
more than their peers that’s right come
22:05
from soccer and that is what gives them
22:07
their their their special advantage
22:10
their comparative advantage so it’s it’s
22:13
quite conceivable that had he not play
22:15
soccer Steve Nash would not
22:16
been a superb mba player because what
22:20
what sets him apart as an NBA player is
22:23
the fact he brings an unusual skill set
22:25
it just so happens that I was emailing
22:28
with Steve Nash about this last week
22:30
Canadian royalty yes and there was this
22:35
HBO real sports segment about sport
22:37
development in Norway because like
22:38
Norway blew everybody away and these
22:39
last Winter Olympics and it’s all this
22:40
unstructured stuff they’re not even
22:42
allowed to have formal games until you
22:43
know competitions until until they’re
22:45
sort of mid or later teen years and and
22:48
Steve’s a big soccer fan you know and
22:50
France which just won the World Cup
22:51
overhauled its development pipeline
22:54
starting decades ago to incorporate this
22:56
so a French soccer player young soccer
22:59
player plays about half as many formal
23:01
games as an American soccer player of
23:02
the same age and and they have this
23:04
saying there’s no there’s no remote
23:05
control meaning the coaches aren’t even
23:07
allowed to talk to them most of the time
23:08
they want to do this like freeform
23:09
unstructured stuff so I actually think
23:11
that so Steve Nash didn’t even get a
23:14
basketball till he was 13 by the way and
23:15
I liked him as an example because he’s
23:17
relatively normal-sized like he’s not
23:20
that big for those she was plus for
23:21
those of you who don’t know Steve it
23:23
occurs to me I’m so deep inside
23:25
basketball nough stood I get there other
23:27
people or not right Steve Nash is
23:29
someone who sort of physically resembles
23:32
me yeah he and happens to be one of the
23:40
10 greatest point guards of the last 50
23:43
years two-time NBA MVP for sure oh yeah
23:46
yes I mean he’s three legendary
23:48
basketball player and he’s a he’s a
23:50
skinny guy from Canada I mean you could
23:53
just the number of parallels between him
23:56
and me are astonishing right right
23:58
this Venn diagram is you in any but um
24:07
but so so go back to silly emailing with
24:10
Steve oh yeah and so we watched this
24:11
real sports and he’s actually exploring
24:12
starting an academy to incorporate these
24:15
principles of unstructured play because
24:17
I think in some ways multi-sport is
24:19
actually just a proxy for movement
24:20
diversity really because if you go
24:23
around to you know Brazil the kids
24:24
aren’t even playing soccer they’re
24:25
playing futsal this game that’s like
24:26
small ball stays on the ground they’ll
24:28
play on a shape this
24:30
sighs one day sand one day cobblestones
24:31
the other day so they’re playing futsal
24:33
but it’s really a different game all the
24:35
time it’s different involving different
24:36
anticipatory skills and and so he’s into
24:39
this and so he wants to start an academy
24:41
because he realizes that his imprimatur
24:43
you know Steve Nash’s name will allow
24:45
parents to say okay maybe we will do
24:48
like what the science says we should do
24:49
instead of going to early specialization
24:51
and Judy Murray Andy and Jamie Marie’s
24:53
mother has done the same thing in the UK
24:55
where she basically facilitates
24:58
unstructured development and people send
24:59
their kids to her camp they won’t let
25:02
the kids do this stuff on their own but
25:03
if Judy Murray says it’s okay like then
25:04
it’s yeah it’s information yeah but this
25:06
is another thing that’s wrong with
25:08
really specialization which again sounds
25:11
like it’s specific to sports but applies
25:13
and I want to talk about that this
25:15
notion and you pointed out that when you
25:17
specialized early and you’re doing the
25:19
same repetitive movements over and over
25:21
again your risk of injury later in life
25:22
starts to increase oh yeah so but this
25:26
is there’s a beautiful parallel to this
25:28
in non-sporting thing yes which is this
25:31
notion of burnout yeah and I wonder
25:34
whether that’s not a really crucial that
25:37
somehow there is something about an
25:38
early specialization that leaches the
25:40
joy out of an intellectual activity and
25:43
limits it far too early I think that’s I
25:47
think that’s probably true for a lot of
25:49
people but by the way I want to say one
25:50
interesting about the injury issue which
25:51
is Cirque du Soleil lots of Olympians
25:55
they looking at this kind of day they
25:58
have a ton of physiology data decided to
26:00
have their performers learn the basics
26:02
of other performers skills not because
26:03
they were gonna perform them but to see
26:05
if it would make them more creative and
26:06
subjectively they thought it did but
26:08
they measure their injury rates next to
26:09
Canadian gymnastics and drop their
26:11
injury rates by a third so something
26:13
about doing that makes people less
26:14
fragile and I have theories about what
26:15
that is but it doesn’t matter the fact
26:17
is it works but but I think you’re
26:19
absolutely right so like when I started
26:20
I had to write about music and range of
26:22
course because probably then the next
26:24
domain that’s most associated with early
26:26
specialization and when you look at
26:28
those studies the main reason that
26:29
people they’re promising musicians quit
26:31
is that they report a mismatch between
26:34
the instrument they play and the
26:35
instrument they wanted to play and if
26:37
you look at the pattern of their
26:38
development they will usually so the
26:41
ones who come on to become the best
26:42
typically have a sampling
26:43
just like the the athletes so like even
26:47
yo-yo ma who actually you know did focus
26:49
very early had a sampling period went
26:50
through it a heck of a lot quicker than
26:52
most people because he didn’t like the
26:54
first two instruments that he was
26:55
playing and this what what the ones
27:01
going to become exceptional they early
27:02
on spread their early practice across a
27:04
larger number of instruments
27:05
whereas what it looks like for the ones
27:07
who plateau at lower levels and or quit
27:09
they have their first instrument where
27:12
they get tons of practice and and
27:14
someone kind of tells them you know you
27:16
can’t switch now you have a head start
27:18
you’ll get behind so it’s you know some
27:20
cost fallacy kind of thing and and they
27:21
end up quitting so I can Battle Hymn of
27:23
the tiger mother you know in the first
27:25
page she says here the secrets to
27:26
successful children and assigns one of
27:28
her children violin and and presides
27:31
over five hours a day of practice and
27:32
and and to the author’s credit later in
27:34
the book she acknowledges the daughter
27:36
says you picked it not me and quits
27:38
right people that part of the message
27:41
didn’t get his famous five hours of file
27:45
in a day is just the most bananas idea
27:47
I’ve ever heard not just but the child
27:48
who has to play five hours but for the
27:50
parent who has to listen to five hours
27:52
no why would anyone do that to
27:53
themselves I I my parents drove idea
27:57
they violated by the way what that one
27:59
week it lasted what my sampling curry
28:01
was one week so it’s great and I walked
28:02
away why why did I quit violin after one
28:05
week because my brother who’s older and
28:08
musical I say that in scare quotes I had
28:12
this child listen to him endlessly bang
28:14
away on the piano and I was like it
28:16
clearly it’s gonna take many many years
28:18
for him to even be remotely kind of
28:20
pleasing on this instrument why would I
28:22
put everyone else in our family through
28:24
the same painful so magnanimous of you I
28:27
don’t have that Reuben hey bris is not
28:32
about my family pathologies this is
28:34
about
28:35
something much fun but I want to talk
28:37
about this in terms of just talk about
28:39
this in terms of schooling what this so
28:42
if you’re if I make David Epstein czar
28:45
of American schooling this leave sports
28:47
aside for a while yeah I would like you
28:49
to redesign the curriculum of K through
28:51
12 to maximize people’s development as
28:56
human beings I see not even K through 12
28:58
came through the end of college tell me
29:00
what you would do in light of what
29:02
you’ve learned from range geez what a
29:05
question the first thing I would do is
29:07
but before I would just overhaul the
29:10
system from the bottom I would start
29:12
with things that we actually could do at
29:13
no cost today which is so so chapter
29:16
four is called learning Fast and Slow
29:17
with apologies to Daniel Kahneman and it
29:21
it details these really well-known
29:24
findings in cognitive psychology about
29:25
learning that again are deeply
29:28
counterintuitive because they showed the
29:29
quickest way to demonstrate progress
29:31
actually undermines long term progress
29:34
so the probably the single most
29:35
surprising study in the book to me was
29:38
this one done at the Air Force Academy I
29:40
love this one is amazing so because you
29:42
could never do this any other place
29:43
right so you have an Air Force Academy
29:44
that brings in you know whatever a
29:45
thousand students every year they all
29:47
have to take a sequence of three math
29:48
courses calculus one calculus two and
29:50
they are randomized to professors for
29:53
calculus 1 re randomized for calculus 2
29:55
re randomized again and so you have this
29:58
incredible experimental condition and
30:00
and these researchers who wanted to see
30:02
the impact of teaching in this
30:04
incredible natural experiment and so
30:07
they followed thousands of students and
30:09
a hundred different professors and what
30:10
they found was that the students so the
30:13
the student and the students
30:14
characteristics coming in were evenly
30:16
spread across classes the students who
30:19
over performed compared to the ability
30:21
that came in with the most in calculus
30:23
one then systematically underperformed
30:28
in all of the follow-on courses the
30:30
professor whose students did sixth best
30:33
in calculus 1 out of 100 and got the
30:35
seventh best ratings from the students
30:37
finished dead last in how his students
30:39
performance was in their follow-on math
30:41
courses after that there was almost an
30:42
inverse relationship between how well
30:44
students over performed in calculus 1
30:46
and how much they then under
30:48
in the follow-on courses in between how
30:50
well they rated that first professor and
30:52
it turned out what those professors were
30:54
doing was they were teaching using
30:56
procedures knowledge they were teaching
30:57
a narrow curriculum that worked really
30:59
well for the calculus one test but did
31:01
not set up these broad frameworks that
31:03
allow you to scaffold later knowledge
31:05
and so again that’s so deeply encounter
31:08
intuitive that you could do something
31:10
that causes this kind of short-term
31:12
progress everyone had to take the same
31:14
tests of course and somehow undermines
31:16
long term development and so I think you
31:18
know you can you can kind of see what
31:20
I’m getting at this fact that the way we
31:22
use testing is evaluation can be a real
31:24
problem if you’re incentivizing people
31:27
to impart using procedures knowledge
31:30
that can make kids do the best on the
31:31
test but it’s not the best for their
31:32
long-term development that’s a problem
31:34
testing is wonderful but for learning so
31:37
there are three in that chapter sort of
31:38
three strategies testing interleaving
31:41
and spacing testing is just quiz
31:43
yourself right you want to force someone
31:44
to generate an answer before they know
31:46
what they’re doing because it’s the
31:47
attempt to generate an answer that then
31:49
Prime’s your brain to remember something
31:50
when you were told the answer so you
31:52
want to test before people are ready
31:53
interleaving means doing tons of
31:56
different kinds of problems the way that
31:57
math study usually works in u.s. is you
31:59
do a type of problem do it do it do it
32:01
do it problem ay-ay-ay
32:02
bbbbb ccccc and that leads to using
32:05
procedures knowledge what you want to do
32:07
is never show the same exact problem
32:08
twice and what that forces the learner
32:10
to do is to match a strategy to a
32:12
problem instead of learning how to
32:14
execute a procedure that’s called
32:15
interleaving where you mix up these
32:17
problems third spacing you don’t want to
32:19
do we usually do you do one thing you
32:23
wait and then you just move on to
32:25
something else whether this is school or
32:26
professional development what you really
32:27
want to do is do something do some other
32:29
things and then come back to that thing
32:31
and so you have this you’re repeatedly
32:33
coming back to things so a famous
32:34
spacing study Spanish vocabulary
32:36
learners they were taught one group was
32:39
taught eight hours on one day the other
32:41
group four hours on day one four hours a
32:43
month later all the same total training
32:45
eight years later when they were brought
32:47
back group two remembered two hundred
32:49
and fifty percent more with no study in
32:50
the interim
32:51
right same amount of study so the first
32:53
thing I would do is incorporate testing
32:55
spacing testing for evaluation not
32:57
testing I mean sorry testing not for
32:58
evaluation testing for learning
33:00
before people are ready spacing and
33:02
interleaving in everything we do because
33:03
that’s no cost stuff that that scaffolds
33:07
learning in a totally different way
33:08
where you learn this this knowledge is
33:09
called making connections knowledge
33:11
instead of but this but instead of
33:14
procedure underneath all that is this
33:16
really fundamental insight which is that
33:18
the sometimes the very best teachers are
33:21
those who disadvantaged us in the short
33:23
term yeah yes and I mean that’s one of
33:26
the themes of the book is that the
33:26
things that you can do they look the
33:28
best in the short term in order to be in
33:31
your terms in order to be the best at X
33:32
it seems intuitive that you should just
33:34
start doing X as soon as possible but
33:36
that turns out not to be the right thing
33:37
more more in a more conceptual level if
33:39
I were the schools are I think there’s
33:42
there’s something to talk about that the
33:44
there’s a section book I talked about
33:45
the army and their failure to retain
33:50
their most talented officers and first
33:53
they tried to throw money at them and
33:54
that the people who are gonna stay
33:56
stayed anyway the people were going to
33:57
leave left anyway and that was a half
33:58
billion dollars of taxpayer money and
34:00
then they started something called
34:02
talent based branching where instead of
34:04
saying here’s your career track up or
34:06
out someone goes in say here’s a bunch
34:08
of career tracks you can sample a couple
34:09
will pair you with a coach and after
34:11
each one they’ll help you reflect on
34:14
what you learned about your own
34:15
abilities what you learned about your
34:17
own interests and you’ll keep
34:18
triangulating until you get this better
34:20
match and I think I would take that
34:21
conceptual approach to kind of
34:22
everything where you help people who’s
34:25
one of my favorite quotes in the book is
34:27
from Herminia Barrett who studies how
34:28
people find careers that fit them she
34:30
says you learned we learned who we are
34:32
in process in practice not in theory and
34:34
what she means is there’s this whole
34:35
industry that tells you can just
34:36
introspect and decide who you are but in
34:39
fact the only way we learn about
34:41
ourselves and our options is by doing
34:42
stuff and reflecting on it so I think I
34:44
would want to build that kind of talent
34:45
based branching where the teacher or
34:47
coach is someone who helps a person
34:49
reflect on what they’ve learned about
34:51
their own abilities and interests from
34:52
these multiple experiences you know what
34:54
the enemy of what you’re describing is
34:56
is self-knowledge I’ve always thought
35:00
that self-knowledge was overrated why is
35:02
it so important to know the kind of
35:04
person you are and what you’re
35:05
describing is the benefits of not
35:08
knowing
35:08
so people who say I’m not a math person
35:11
I’m
35:13
I’m very X or Y are precisely the people
35:16
who would who would who would object or
35:20
who would great would have a problem
35:22
with the kind of course of action you’re
35:24
describing right you’re asking people to
35:26
sample widely yes outside their areas of
35:29
specialty right or their areas of
35:31
interest or their areas of another
35:32
interest their areas of imagined
35:34
interest and imagined specialty right on
35:36
the on the grounds that they don’t know
35:38
what it is it’s that will that they’ll
35:40
either thrive at or what they need to be
35:42
good their insight into themselves is
35:45
constrained by their roster of previous
35:46
experiences period yeah and and that’s
35:49
an important thing to know and not only
35:51
that but this concept this concept
35:52
they’re right about the end of history
35:53
illusion right which shows that we are
35:56
at every time point in life we realize
35:58
we’ve changed a lot in the past our
36:00
preferences our values what we think our
36:02
strengths and weaknesses are the friends
36:04
we prefer the things we like to do for
36:05
fun and at every time point we then
36:08
underestimate how much we will change in
36:09
the future we keep saying like man I
36:10
changed a lot from these experiences
36:12
there but now I’m pretty much done and
36:13
we say that at every time point so it
36:15
leads to these really weird results like
36:17
when people are asked how much they
36:18
would pay to see their favorite band
36:21
today ten years from now the average
36:22
answer is 129 dollars and when they’re
36:25
asked how much they would pay wait that
36:26
wasn’t even a joke when they’re asked
36:28
when they’re asked how much they would
36:29
pay today to see their favorite band
36:31
from ten years ago the average answer is
36:32
eighty dollars and so we really
36:35
underestimate how much we changed this
36:36
idea that we we come like fully formed
36:40
with with insight into ourselves is is
36:42
not supported by any other work in them
36:44
you know this reminds me of in my this
36:46
season of my podcasts I have all these
36:47
episodes about Jesuits
36:50
it’s the theme is how to think like a
36:53
Jesuit because I love the way Jesuits
36:54
think and the Jesuits had this really
36:56
lovely notion and by the way if there’s
36:58
a Jesuit in the audience and I get this
37:00
wrong just raise your hand and correct
37:02
me I’m not a Jesuit that’s why I’m like
37:05
but they have a notion what’s called
37:07
disordered attachments and the idea this
37:10
is an idea from Saint Ignatius is that
37:12
you cannot approach a problem if you’re
37:14
bringing with you attachments that get
37:17
in the way of seeing listening clearly
37:20
seeing the nature that so that guy I was
37:23
talking to gave me this example of when
37:25
he was a novice
37:26
and he was supposed to be sent out to do
37:29
your training and he said to his senior
37:32
he said you know I’ll do anything you
37:34
want I’ll go anywhere to do my training
37:36
just don’t send me to a hospital because
37:38
I just can’t stand beside the blood and
37:40
people dependent and the guy said you’re
37:42
going to a hospital and it’s exactly
37:44
your point oh he was observing that he
37:46
had a disordered attachment to the idea
37:48
that he was someone who could deal with
37:50
who could not deal with that particular
37:53
set of problems and that set of problems
37:55
were not useful to the direction he
37:57
wanted to go on and this the senior
38:00
priest understood that no that’s you’re
38:03
25 years old you can’t have a definite
38:06
self definition that rules out this
38:08
massive area of your presumed
38:11
responsibility which is people who are
38:13
suffering right we’re physically
38:14
suffering I mean especially at at 25 at
38:17
the the period of fastest personality
38:19
change over your life is 18 to 28 so
38:21
making those decisions for a 25 year old
38:24
is it’s making a decision for someone
38:25
that you don’t really know yeah
38:27
and for a world that you can’t really
38:28
conceive yeah and I don’t think that’s a
38:30
good stretch it’s go radical for a
38:33
moment what if for example would you go
38:36
for the following idea what if we if we
38:38
had students choose their – when you
38:41
when at the point in high school when
38:43
you start choosing your own courses what
38:45
if we took that away and we just had
38:47
people assigned people courses randomly
38:51
I think that would probably be fine as
38:53
long as we again I think we should
38:54
implement this kind of but I love that
38:58
you without even a moment like mister
39:02
stop we’re in the Upper East Side of
39:04
Manhattan we’re within walking distance
39:06
of Dalton
39:07
we’ve just suggested that the premise on
39:09
which all of this high-priced education
39:10
is based yeah it’s just nonsense well I
39:13
mean yet the evidence no but but the
39:16
evidence for that is actually pretty
39:17
clear though yeah like quiet because
39:21
school outcomes come from their
39:23
selection of students and those students
39:25
others background factors and things
39:26
like that African stuff they learned in
39:27
school the equivalent would be a
39:28
hospital that boasted of its success in
39:31
curing people and only admitted the
39:33
healthiest people right they basically
39:36
just admitted members of the US Olympic
39:37
team and then turned around and said oh
39:39
my God look at our look at the the
39:42
health status of our of our patients
39:45
right extraordinary they’re setting
39:47
World Records that’s how much we cured
39:48
them right right things like yeah no
39:51
wait so this is so you’re in favor of
39:54
randomization what was your caveat that
39:57
I think especially at those early ages
40:00
yeah I just want to make sure anything
40:02
they do like when I when I first started
40:05
in training to be a scientist when I did
40:07
my first lab work I thought here’s where
40:09
I’m gonna learn that this is what I want
40:11
to do for the rest of my life and I
40:13
learned that maybe this wasn’t what I
40:14
wanted to do for the rest of my life and
40:16
I wasn’t happy about that but that was a
40:17
very important signal to get but I think
40:19
you want to make sure that that you help
40:21
them maximize their learning from these
40:22
experiences and and some people do that
40:24
on their own so-called self-regulatory
40:26
learners and the thing they do the most
40:28
of is they stop and reflect and so so I
40:32
think you want to help them and make
40:33
sure that they’re getting signal from
40:34
whatever it is that they’re doing and I
40:36
don’t think you would lose much by
40:37
randomizing anyway because the fact is
40:39
like when Jim Flynn who I think we we
40:41
both know studied college students both
40:44
in in the UK and in the u.s. he found
40:46
that the skills that they were able to
40:47
use to get good grades at elite colleges
40:49
what had a I think it was about a point
40:52
zero three correlation with their
40:54
abilities on a test of critical thinking
40:56
that really matters in the world so
40:58
we’re clearly in viewing people with
41:00
skills that are no good for critic
41:02
analyzing the actual world so I don’t
41:04
think you stand to lose very much in
41:05
most cases what about this brings up a
41:07
second notion which is does this
41:10
argument suggest that you may learn more
41:14
from situations where you are relatively
41:17
speaking performing badly than
41:18
situations where you’re performing well
41:21
well so so Kaiser no relationship no I
41:24
mean the cognitive psychologist Nate
41:26
Cornell’s say if your difficulty is not
41:30
a sign that you aren’t learning but ease
41:32
is and I think that’s a that’s a good
41:34
thing to keep in mind if something is
41:35
too easy then maybe you like to do it
41:38
but that doesn’t mean that you’re not
41:39
learning much right like you can go to
41:41
the gym and lift the same weights the
41:43
same number of times every day and you
41:44
won’t slide backward but you also won’t
41:45
won’t cause adaptation yeah and and so
41:49
no I think there’s I think there is
41:51
something to that so why why then do we
41:54
persist what is the value of assigning
41:56
grades to academic performance in high
41:58
school I assume part of the goal is to
42:02
give people a sense of how they’re doing
42:03
but if you really wanted to
42:05
what self-regulatory learners really
42:07
need or the kind of the best learners is
42:08
very fast feedback so when you actually
42:10
give them a test and then show them the
42:11
answers like weeks later that’s not good
42:13
for learning you would want to do it
42:14
like right away yeah or work with them
42:17
right away not just give them the answer
42:18
so I but I think it’s for because we
42:22
need a system that likes if it’s people
42:23
for colleges you know and give them some
42:26
ostensibly is to give them some kind of
42:27
feedback about how they’re doing and how
42:29
they’re learning is but I think in but
42:31
why can’t that be exactly why is it I
42:34
don’t know I mean I just said playing
42:37
with it strikes me that the more you
42:39
think about your argument the more sort
42:42
of subversive it becomes because if you
42:45
really wanted to redesign an educational
42:47
system from scratch based on these ideas
42:50
yeah you really are sweeping away a
42:53
whole lot of the kind of rituals that
42:55
surround formal education I think there
42:59
are things that are done well in formal
43:00
education – I should say by the way like
43:01
so chapter 2 is about these steadily
43:03
rising IQs across the world and I think
43:06
some of that has to do with things that
43:07
that are learned in school in that sense
43:09
you think thin effect is so the Flynn
43:12
effect is this observation that in
43:15
the developed world at least and also in
43:17
parts of the yeah the non bubbler IQs
43:20
have been rising steadily over the last
43:21
hundred years I thought that was like
43:26
didn’t Steven Johnson famously argue it
43:29
was video games so it’s it seems to be
43:33
this well it’s controversial but seems
43:35
to be this more abstract thinking basic
43:37
because that all the games have been on
43:38
the most abstract type of tests like not
43:40
the stuff that’s taught in school
43:41
yeah that’s an argument that school
43:42
hasn’t helped but it’s it’s like things
43:44
like Ravens progressive matrices we get
43:46
these abstract designs you have to fill
43:47
in the missing one so this was supposed
43:49
to be the test that like if Martians
43:50
landed it could tell how smart they were
43:52
and in fact it’s the one that has to had
43:54
the most change and so I don’t think
43:55
it’s from things that are explicitly
43:56
taught in school but the way that that
43:59
we’ve learned to as Flynn would say look
44:00
through scientific spectacles to do
44:02
classification and abstraction I don’t
44:05
think that’s necessarily taught on
44:06
purpose but I do think it is is it
44:08
ingrained in what happens in school and
44:10
I don’t want to rag on schools too much
44:12
cuz I point out in the high point not
44:15
stopping you I mean this is the perfect
44:17
place to rag on school David you have no
44:19
bow but but I mean I do point out that
44:22
everyone thinks that education has
44:24
gotten worse since their day right but
44:26
actually so I I put in range some
44:28
questions from you know like the 60s
44:31
that sixth graders would say said I
44:32
thought it was said now patients got
44:34
more expensive since their higher
44:36
education I think right but so this
44:37
would be like middle school education
44:39
basically and there was without question
44:40
middle schoolers have a better grasp of
44:42
basic concepts than their forebears did
44:44
without question but if you look at the
44:46
test questions that test the same level
44:48
in the 60s it was like you know rate
44:50
times time problem just apply it and and
44:52
now it’s like these complex word
44:54
problems that require multiple steps and
44:56
so the challenge has gotten much harder
44:58
because we’re not training people to do
45:00
repetitive tasks anymore because we’re
45:01
not in that same kind of industrial
45:02
world so school is actually doing better
45:04
it’s just the challenge has outpaced
45:06
that yeah yeah in any case that the main
45:09
thrust of your argument is not so much
45:12
that what schools are doing is wrong but
45:16
that the the the the techniques that
45:19
students and parents use to navigate
45:21
education are wrong we’re making
45:24
different oh you know we’re just making
45:27
the wrong choice
45:28
is within a stable environment that’s
45:29
really the kind of yeah I mean one
45:31
program that I learned about was
45:33
researching is called career academies
45:35
that targets kids who you know are are
45:38
by traditional measures not not really
45:40
headed to college and gives them some
45:41
sort of vocational training basically or
45:43
early exposure to types of work and
45:45
surprisingly even when they often do not
45:48
decide to go in to do anything with that
45:50
career they still do better overall like
45:53
in income wise going to do something
45:54
totally totally different and and I
45:57
think a little bit of that has to do
45:58
again they’re getting more significant
46:00
sort of signal about themselves and
46:01
about match quality than you often do in
46:03
traditional classes yeah yeah
46:05
when is um speaking of match quality
46:08
presumably you couldn’t keep searching
46:11
forever I mean I have no idea what I’m
46:13
gonna do when I grew up I literally have
46:14
no idea what I’m gonna do that now like
46:16
no idea I mean when I was a teenager
46:20
that was maybe the Air Force Academy be
46:21
a test pilot and be an astronaut and
46:22
I’ve gotten like linearly less long term
46:24
goal directed I don’t know whether
46:29
whether you’re your particular position
46:32
right now is a best-selling author is
46:33
generalizable to the general public no
46:35
no I mean but but this was but in the
46:38
Dark Horse project in the book the the
46:41
common trait of people who find
46:42
fulfilment in their careers is it
46:43
focused on short term planning and and
46:46
that resonated with me so much such that
46:47
I ended up as a subject in the study
46:49
which I disclosed in the book what they
46:52
do is they all came in and would say
46:53
well you know don’t tell people to do
46:56
what I did because I came through this
46:58
weird path I thought I was gonna do one
47:00
thing then I tried I didn’t like it so i
47:01
zigged and zagged and they all view
47:03
themselves as having come out of nowhere
47:04
which is why the researchers called it
47:05
the Dark Horse project and their common
47:07
trait is this short term planning where
47:08
they don’t look around and say here’s
47:10
who’s younger than me and has more than
47:11
me they say here’s Who I am right now
47:13
here my skills and interests here the
47:14
opportunities in front of me I’ll try
47:15
this one here’s my hypothesis about what
47:17
I’ll learn and a year from now I’ll
47:18
change because I will have learned
47:19
something new and they just do that
47:20
until they get to a spot where they can
47:22
kind of uniquely succeed and feel
47:24
fulfilled and so I’ve totally abandoned
47:26
that that longer-term
47:27
planning in favor of these short-term
47:29
proactive experiments and and why would
47:31
you have to stop you can keep doing that
47:33
year old yeah does your if you were
47:34
running a company based on these
47:37
observations would you
47:41
put that observation into practice
47:43
that’s to say right now companies do a
47:45
version of this right they silo people
47:47
from the get-go
47:48
hmm you start out in marketing you stay
47:50
in marketing unless you’re one of the
47:51
very very few to rise to the very top
47:53
and then maybe you get a shot too
are you thinking you would do much more
cross specialization with it even at
four people in their 30s and 40s
absolutely absolutely the you know one
of my favorite character knew but kind
of got her first like real job when she
was 54 basically but and in a lot of the
characters did that Oliver Smithies was
another of my favorite characters right
who he he he started in med school and
then does chemistry and becomes because
he sees a lecture and loves it and it
becomes a biochemist when that was not a
thing now that’s its own specialty at
the time it was this weird hybrid and
then his 50s he decided to take a
sabbatical two floors away like every
couple years he goes in new domain two
floors away from his own lab to learn
DNA and then at 60 does his work that
wins in the Nobel Prize and lower like
on drag I’m the only in the last chapter
the only scientist who’s won both the
igg Nobel Prize for the silliest
research and the Nobel Prize for and he
says which one did he win first ignoble
for levitating frogs with magnets and
why is that ignoble that sounds like a
really interesting thing they they ask
you if you’re willing to accept it
because of the reputational thing first
and he was he was like happy to accept
it and he likes to say it’s
psychologically unsettling to switch
gears but he likes to say I don’t I like
to say I don’t do research I only do
search and I sort of love that he and
you and we were just talking about Bill
Simmons before this and the ringer who’s
who’s bounced around and had some huge
successes and some failures in some of
the work that he’s done and one of the
things is now these people who are
becoming famous at the wringer went in
with totally different jobs and he
49:28
allowed them to come on a podcast or to
49:30
write a story or do this other thing and
49:31
now they are famous yeah because he
49:33
allowed this sort of internal mobility
49:34
and then for to try things like you were
49:36
you were just calling someone a genius
49:38
who was hired as an online editor and
49:39
now as a famous podcaster right and
49:41
that’s not what she was hired for and he
49:43
allows people to try different things
49:45
yeah it was one of like the happiest
49:46
workplaces that visited do it well it’s
49:48
interesting because the what that
49:50
reminds us is that we have going back to
49:52
this question of math
49:54
we have way too much confidence in the
49:57
accuracy of the match mechanisms with
50:00
that are in place for sure like so you
50:01
you know the there is no reason for
50:04
someone who is 25 or 28 or 30 years old
50:08
to believe they have the system has
50:10
successfully matched them with what
50:11
their what they ought to be doing yeah I
50:13
mean or you can you mean you can always
50:15
be looking to make that match better
50:16
right and again this is what the army
50:18
realized where they said are our
50:19
traditional tests are not doing as good
50:22
a job as this talent based branching
50:23
base yeah that you have to actually do
50:25
some experimentation and maybe that’s
50:26
annoying but it should be viewed as an
50:29
investment in long-term development yeah
50:31
we have questions number and I’ve been
50:34
remiss and not looking at them oh here’s
50:40
a good one that points out that there’s
50:42
a difference between the u.s. in and the
50:45
English educational systems and to our
50:49
credit we generalize longer and yeah the
50:54
Brits specialized earlier yes so would
50:57
you they must your your argument would
51:01
suggest they’re paying a price for that
51:02
correct in fact the economists who study
51:05
that was got interested in it because he
51:07
was gonna go into the British school
51:08
system at the last minute decided he
51:10
wasn’t sure what he was gonna do and and
51:12
decided to come into the u.s. school
51:13
system instead and decided to that got
51:15
him interested in studying
51:16
specialization timing so he looked at
51:18
for example the school systems in
51:19
England and Scotland which are very
51:20
similar except for specialization timing
51:22
where Scotland allows some more sampling
51:24
in England you know mid teen years you
51:26
already have to be like applying to
51:27
programs in college and he said who wins
51:29
this trade-off the earlier the late
51:30
specialized errs and what he found is
51:32
it’s usually the late specialized errs
51:34
that you know there’s a million ways to
51:35
you have to get to performance but that
51:37
the late specialized errs have higher
51:39
growth rates because they match better
51:40
and they and they end up quitting less
51:41
because they match better do you think
51:42
that that is that is more of an issue
51:44
with people of high ability that is to
51:46
say do people of high ability require
51:48
longer to find their match yeah I don’t
51:51
know I think you know I don’t know the
51:55
answer to that I think you could make
51:56
the argument that they have more options
51:58
to choose from so for example like if
52:01
you look at something like the the study
52:03
of mathematically precocious youth that
52:04
has these five cohorts that it started
52:06
tracking from age
52:06
well and so some of these people are
52:09
middle-aged now the it takes these kids
52:12
who scored eight hundred on the SAT when
52:13
they’re 12 on the SAT math and and the
52:15
girls have score 810 to also score like
52:17
super high on the verbal if you score
52:19
high on one you tend to score high on
52:20
the other but a lot of the girls who
52:21
scored 800 a math score very high in
52:23
verbal and they tend to have this wider
52:25
variety of careers whereas you know if
52:28
some of the boys are have this high
52:29
ability tilt then they’ll go toward that
52:31
tilt but the people who are more even
52:33
tend to have these more options so they
52:35
spread across this larger numbers are
52:37
lovely if I’m remembering this this
52:39
research correctly they make this lovely
52:42
observation that boys to find what they
52:44
like is what they’re good at and girls
52:47
don’t they separate those two traits and
52:50
this is why they were trying to
52:51
understand why there was so much there
52:54
were all these brilliant girls who were
52:56
brilliant in science and math who were
52:59
leading science and math and they they
53:01
thought they had scrubbed out all the
53:02
bias and scrubbed it all the and and
53:06
they were so puzzled by this and what
53:08
they realized in the end was it was this
53:09
simple this difference in definition
53:11
it’s a it’s a matching gift definition
53:14
boys match to things they were good at
53:16
thinking that that would that correlated
53:18
with what they would like indras girls
53:19
never made that ascent in many ways that
53:23
the girl position is superior to the boy
53:25
position don’t you think typical yeah
53:31
hold on how does being a generalist
53:36
protect one against AI yes so that’s
53:39
kind of like the topic of the first
53:40
chapter in some ways so I’ve kind of I
53:44
think it’s and I didn’t make this up in
53:46
AI researcher gave me this this idea to
53:48
think about AI on a spectrum from from
53:51
chess where it’s based on there’s a huge
53:53
database of previous knowledge they’re
53:56
you know very constrained rules it’s not
53:58
changing and so computers have made
54:02
totally explosive exponential progress
54:04
done to self-driving cars where we’ve
54:08
made huge progress and very constrained
54:10
rules but there are sometimes unfamiliar
54:12
situations and it turns out we were not
54:14
as close right like like Elon Musk keeps
54:17
pushing it out two years seven two years
54:19
– the others – when we’re gonna have all
54:21
the way to something like scientific
54:23
research or cancer research we’re ibm’s
54:26
watson has been such a big flop that
54:28
some of the AI research as i talked to
54:30
were worried that it would damage the
54:32
reputation of AI in healthcare and as
54:34
one of the oncologists that quote had
54:36
said the reason watson did well on
54:37
Jeopardy and and not in cancer research
54:39
is because we know the answers to
54:41
jeopardy and so i think in those those
54:44
challenges that are more repetitive yeah
54:47
those are much more amenable to
54:49
automation if you look at things like
54:50
James Bessemer a good example of the ATM
54:53
when that came in and was supposed to
54:54
obviate bank tellers and in fact we it
54:56
caused more bank tellers because it made
54:58
every branch cheaper and so banks could
55:00
open more branches and they could hire
55:02
more tellers overall but it totally
55:04
changed the job from someone who had
55:06
these very specific procedural skills
55:08
that had to do with transactions to
55:10
someone who has this much more amorphous
55:11
human behavior marketing customer
55:13
service kind of orientation to these
55:15
much more sort of softer skills and even
55:18
even where AI is really good like in
55:22
chess I mean it was Garry Kasparov who
55:25
recognized when he played deep blue more
55:28
of expects this idea that humans and
55:30
computers have opposite strengths and
55:31
weaknesses and he realized the computer
55:33
was far superior tactics which is
55:35
patterns that’s you know that’s most of
55:38
chess and grandmasters patterns study is
55:40
their advantage but humans are good at
55:42
strategy this bigger picture planning of
55:44
how to manage the little battles to win
55:45
the war and so some of his efforts led
55:47
to this freestyle chess tournament where
55:49
humans could play with computers or
55:51
whatever they wanted and the the a
55:54
couple of amateur humans with normal
55:57
laptops beat the best humans they beat
55:59
the best computers and they beat the
56:00
best grandmasters with the best
56:02
computers these so-called centur teams
56:03
these were people who were moderate
56:06
amateur chess players who knew something
56:07
about computer search who could hand a
56:09
lot of glowing information so suddenly
56:11
overnight this stuff that Kasparov has
56:14
spent his life learning is outsourced to
56:16
the computer and he’s no longer the best
56:18
where when the game when the humans are
56:21
doing the thing that humans are uniquely
56:22
good at yeah by the way parenthetically
56:24
I’ve never understood why people got so
56:27
worked up about the fact that a computer
56:29
could be
56:29
in chess it’s like saying psyche getting
56:31
worked up about the fact that a car can
56:33
beat a human in a race well like yeah
56:37
it’s a car
56:37
I mean why we have races because we
56:41
raced people who are like ourselves
56:42
right you don’t race a car normally
56:45
because it’s in a different class has an
56:47
engine uses gasoline for yourself so
56:50
then in chess there like they import not
56:53
just a different species but an actual
56:56
like machine I don’t like whoa
56:59
the machine could beat me well I mean
57:01
anyway just crazy
57:03
it was IBM marketing it was actually I
57:06
think that IBM the we’re almost out of
57:10
time but we have about a few more
57:12
questions a few more minutes what has
57:15
been your book has been out for mmm how
57:19
many days to to put it I guess it has a
57:24
three you’re too modest to say this it’s
57:26
everywhere and having this kind of
57:30
sensational effect and when you and I
57:32
you and I did a David and I did a
57:34
similar kind of thing back in March
57:37
March sports centric that words and I
57:40
have to say as you were describing this
57:41
idea to a roomful of was at this this
57:44
sports nerd conference called Sloan and
57:47
it was like two thousand nerds and they
57:50
were so riveted you could have heard a
57:53
pin drop as you’re explaining this there
57:55
is something about this idea that seems
57:57
to grab people by the lapels what why
58:02
why do you think everyone is so kind of
58:05
powerfully attracted to this argument I
58:07
think I think well I think there have
58:12
been very strong arguments that people
58:13
perceive going in the other direction
58:14
for one no reason but you look at that
58:19
David see anyone but I think this is a
58:25
topic how broad or how specialized to be
58:28
that is important to everyone whether
58:32
they discuss it explicitly or implicitly
58:34
that that the signal is very strongly on
58:37
one side to only do one and that we talk
58:40
about all the time but that we talk
58:41
about purely with intuition
58:42
and so I think my goal was to not to be
58:46
the final word on this because I don’t
58:48
even know ultimately how you could ever
58:49
have the final word on this but to look
58:53
and see what research was out there and
58:55
bring to those conversations some
58:58
concrete information that I hope could
58:59
make those discussions much more
59:01
productive and interesting for people so
59:04
I hope it’s that this important
59:06
discussion that’s only been grounded in
59:07
tuition maybe we’ll go a little bit of a
59:10
different place now and that plus so
59:14
much of it was deeply counterintuitive
59:15
for me again this idea that you can do
59:17
things in the short-term that undermine
59:18
your develop in a long-term and so I so
59:21
I think you know apparently
59:23
counterintuitive nough sometimes works
59:25
in writing yes it does it does thank
59:28
David thank you very much David will be
59:30
by the way before you don’t clap yet’
59:31
David you will be signing books we’re
59:34
downstairs I’m not sure somewhere
59:36
somewhere in this area David will be
59:39
signing your books I encourage you all
59:40
to read this book I genuinely think it
59:44
is an eye-opening and much needed and
59:48
beautifully written
59:49
well book and you’re to be commercially
59:52
and and you know I want to thank you for
59:55
your support of it because like I said
59:56
this has been like a model sort of
59:58
intellectual relationship for me you
60:00
make me sound like I’m be an old guy
60:02
like this is Karate Kid none and your I
60:05
take it all back take it off back all
60:08
right thank you very much David thank
60:09
you
60:10
[Applause]