WEBVTT

00:00:00.720 --> 00:00:04.637
Welcome, welcome, welcome to Armchair
Expert, experts on expert. I'm Dan

00:00:04.720 --> 00:00:06.557
Shepard. I'm joined by Monica Mouse.

00:00:06.640 --> 00:00:07.837
>> Hi. >> How are you?

00:00:07.920 --> 00:00:10.637
>> Oh, we were saying that maybe my new

00:00:10.720 --> 00:00:13.997
animal is a It's like a mouse.

00:00:14.080 --> 00:00:16.637
>> A hyrax. >> Hyrax. >> You're a Hyrax.

00:00:16.720 --> 00:00:19.437
>> Monica Hyrax. >> Oh,
yeah. Monica Hyrax bad.

00:00:19.520 --> 00:00:22.237
>> Cuz you said your new
nickname is the boulder.

00:00:22.320 --> 00:00:24.877
>> Oh [__] Right. That was on
another thing that we didn't air.

00:00:24.960 --> 00:00:27.437
>> Yeah. >> Yeah.
That's important. People should

00:00:27.520 --> 00:00:29.757
recognize that I am now
going by the boulder.

00:00:29.840 --> 00:00:32.957
>> Yep. >> Because
what's more formidable, a

00:00:33.040 --> 00:00:36.877
boulder or the rock?
I think the viewer can decide. [gasps]

00:00:36.960 --> 00:00:41.037
>> This guest, I really hope people listen
to this interview because I'm going to

00:00:41.120 --> 00:00:43.837
put it in in into a handful of guests

00:00:43.920 --> 00:00:47.037
we've had where I I just was literally

00:00:47.120 --> 00:00:52.477
overwhelmed with the scope and and
breadth of their knowledge of how all

00:00:52.560 --> 00:00:53.997
things work where I was like, "Oh, wait.

00:00:54.080 --> 00:00:57.597
Does this person kind of understand
how everything works in concert?

00:00:57.680 --> 00:01:00.397
>> It's amazing to be
in front of someone like that.

00:01:00.480 --> 00:01:04.397
>> Yes. And that man is Eric Schmidt.
Now, Eric Schmidt is a technologist, an

00:01:04.480 --> 00:01:05.997
entrepreneur, and a philanthropist. He

00:01:06.080 --> 00:01:11.837
is the founder of Schmidt Futures
and former CEO of Google for a long time.

00:01:11.920 --> 00:01:14.557
>> Yeah. >> Uh I started
this interview wondering

00:01:14.640 --> 00:01:18.877
how would those two people pick someone
they thought was more qualified than

00:01:18.960 --> 00:01:20.797
themselves to run their company? Who

00:01:20.880 --> 00:01:23.197
could that person be? And we met him. He

00:01:23.280 --> 00:01:25.917
has a new book out called The Age of AI

00:01:26.000 --> 00:01:30.397
and Our Human Future. The Age of AI
and Our Human Future explores what society

00:01:30.480 --> 00:01:33.837
must be focused on right now to make
sure that we are prepared for a new

00:01:33.920 --> 00:01:38.237
reality where humans and AI must coexist.
He wrote this book with Henry

00:01:38.320 --> 00:01:40.637
Kissinger and Daniel Hutenlocker.

00:01:40.720 --> 00:01:42.637
Fascinating interview. Please enjoy.

00:01:42.720 --> 00:01:45.117
Eric Schmidt.

00:01:45.200 --> 00:01:49.837
We get support from Quint. So, we've
been talking about Quint a lot lately,

00:01:49.920 --> 00:01:53.677
and as we head into fall, it's the perfect
time to reset your wardrobe with

00:01:53.760 --> 00:01:58.317
their amazing pieces. I think it's so
fun to do when the seasons turn over to

00:01:58.400 --> 00:02:00.317
just spruce up your wardrobe with a

00:02:00.400 --> 00:02:03.917
couple specials, and Quint is the place.

00:02:04.000 --> 00:02:06.477
>> Totally. We're talking organic cotton,

00:02:06.560 --> 00:02:09.917
Mongolian cashmere, tailored chinos from

00:02:10.000 --> 00:02:13.117
just $60. Premium fit and quality you'd

00:02:13.200 --> 00:02:17.677
expect from a luxury brand designer.
I love Quint, you know. I like it

00:02:17.760 --> 00:02:22.877
understated and subtle and I like stuff
that has high quality and Quint checks

00:02:22.960 --> 00:02:26.157
all those boxes.
>> I love a lot of stuff from Quint

00:02:26.240 --> 00:02:31.277
honestly. They have incredible premium
denim which is very important for fall.

00:02:31.360 --> 00:02:35.437
It's pretty much all I wear in the fall
is denim. And Quinc has such a good

00:02:35.520 --> 00:02:39.997
selection of quality denim. Their stuff
is so versatile, timeless, and built to

00:02:40.080 --> 00:02:42.717
last that you just grab every
single day without thinking.

00:02:42.800 --> 00:02:45.677
>> Right. And all of that for 50 to 80%

00:02:45.760 --> 00:02:49.997
less than similar brands. They go
direct to Ethical Factories, cut out the

00:02:50.080 --> 00:02:54.237
middlemen, so you get all that
quality without overpaying for it.

00:02:54.320 --> 00:02:58.077
>> We love that. And they do way more
than clothes. They have bedding, cookware,

00:02:58.160 --> 00:03:02.237
bath essentials, travel musthaves,
same deal, and great quality.

00:03:02.320 --> 00:03:06.797
>> Upgrade your everyday. Download the
Quint app for app exclusive offers or go

00:03:06.880 --> 00:03:09.197
to quint.com/dax.

00:03:09.280 --> 00:03:11.037
Get free shipping on your order and

00:03:11.120 --> 00:03:14.157
365day returns. Now available in Canada

00:03:14.240 --> 00:03:20.120
and the UK, too. That's qui nce.com/dax.

00:03:20.960 --> 00:03:23.837
We get support from Apple gift card.

00:03:23.920 --> 00:03:28.717
Back to school is here and for many that
means some big life transitions. Whether

00:03:28.800 --> 00:03:32.557
they are heading into a new grade
or setting up a dorm for the first time,

00:03:32.640 --> 00:03:35.997
you want to give them something
that actually supports their growth and

00:03:36.080 --> 00:03:40.157
success. Apple gift card is a versatile
solution because it takes the guesswork

00:03:40.240 --> 00:03:44.877
out of gifting. It's not just about
a new pair of AirPods for focus. It's

00:03:44.960 --> 00:03:48.637
about giving them the invaluable
chance to choose the digital tools and

00:03:48.720 --> 00:03:51.837
subscriptions that suit their unique
journey. They can use it for

00:03:51.920 --> 00:03:56.077
productivity apps on the App Store,
movie nights on Apple TV, or extra

00:03:56.160 --> 00:04:00.877
iCloud storage for all those new memories.
With Apple Gift Card, it's so

00:04:00.960 --> 00:04:02.717
empowering to give the students their

00:04:02.800 --> 00:04:05.437
freedom to allocate their own resources.

00:04:05.520 --> 00:04:07.917
Celebrate their new academic chapter.

00:04:08.000 --> 00:04:11.400
Visit applegiftcard.apple.com

00:04:11.520 --> 00:04:17.320
to celebrate their milestones
and gift one today.

00:04:20.040 --> 00:04:22.060
[music]

00:04:27.360 --> 00:04:32.557
>> He's an So, first
of all, nice to see you.

00:04:32.640 --> 00:04:37.197
>> Great to see you again. >> Yeah,
I guess I would have a year ago if

00:04:37.280 --> 00:04:39.917
someone asked me, "Do you think you'll
ever meet Eric Schmidt?" I would have

00:04:40.000 --> 00:04:45.117
said no. And then here we are twice
face to face within four months.

00:04:45.200 --> 00:04:47.197
>> Illuminati. >> Okay. Okay.

00:04:47.280 --> 00:04:50.397
>> We've had so many people on from that

00:04:50.480 --> 00:04:52.317
conference event

00:04:52.400 --> 00:04:57.837
>> and I'm so jealous. And if Adam doesn't
invite me next year, I'm gonna be mad

00:04:57.920 --> 00:05:01.757
making that. >> Well, Jared
Cohen's the guy you want to

00:05:01.840 --> 00:05:04.077
be on your side in this scenario.

00:05:04.160 --> 00:05:06.237
>> He's getting an email from me.

00:05:06.320 --> 00:05:08.877
>> He uh he controls the world [laughter]

00:05:08.960 --> 00:05:14.317
as far as I'm concerned.
>> And and he's grown into a real person.

00:05:14.400 --> 00:05:17.117
You know, he started off as sort
of a kid in the State Department.

00:05:17.200 --> 00:05:20.797
>> Yeah. >> But he made a significant set
of accomplishments. You may not know this,

00:05:20.880 --> 00:05:24.877
but his most significant achievement has
been he worked tirelessly for the last

00:05:24.960 --> 00:05:26.877
few weeks to get all the Afghani

00:05:26.960 --> 00:05:29.917
refugees out. And what he did, which I

00:05:30.000 --> 00:05:34.237
didn't understand, is he simply called all
the prime ministers and yelled at them.

00:05:34.320 --> 00:05:36.157
>> Oh, really? >> You have
to do this. You have to do

00:05:36.240 --> 00:05:40.077
this. You have to do this because it was
the only way to get places to put these

00:05:40.160 --> 00:05:44.477
people. And so a couple thousand
people look like they got out because of

00:05:44.560 --> 00:05:47.037
Jared's direct intervention. >> Oh my god.

00:05:47.120 --> 00:05:48.557
>> And the help of a lot of other people.

00:05:48.640 --> 00:05:51.757
And that's a great day. >> Oh my gosh.
Yes. I don't know that I'll

00:05:51.840 --> 00:05:53.037
ever be able to make a claim that.

00:05:53.120 --> 00:05:56.797
>> No. >> That's so wonderful. Yeah.
That I actually got people to safety.

00:05:56.880 --> 00:05:58.797
>> Saving lives.
I mean, this is why people

00:05:58.880 --> 00:06:02.637
go into being doctors. Imagine walking
in and saving someone's life. It's

00:06:02.720 --> 00:06:04.477
really an extraordinary achievement.

00:06:04.560 --> 00:06:07.037
>> Yes. >> What what's
interesting to me is that we

00:06:07.120 --> 00:06:10.157
have problems in society
that we don't know how to solve.

00:06:10.240 --> 00:06:11.757
>> And I view these as the hard problem.

00:06:11.840 --> 00:06:13.997
And I'll give you an example of addiction.

00:06:14.080 --> 00:06:16.557
>> Another one is homelessness.

00:06:16.640 --> 00:06:21.277
and why my industry is not trying
to figure out how to solve these problems.

00:06:21.360 --> 00:06:23.437
Maybe because the science is too hard,

00:06:23.520 --> 00:06:25.757
but wouldn't it be amazing if there was

00:06:25.840 --> 00:06:28.877
some app that came along that could

00:06:28.960 --> 00:06:31.837
really address addiction, homelessness,

00:06:31.920 --> 00:06:37.117
or inequality of some form, you know,
unfairness in some form. We need to

00:06:37.200 --> 00:06:40.557
start working on that stuff.
>> Well, wouldn't you say conventionally

00:06:40.640 --> 00:06:44.317
the hurdle is that those aren't
profitable endeavors? that there's not a

00:06:44.400 --> 00:06:47.117
ton of incentive for any
industry to tackle those.

00:06:47.200 --> 00:06:48.957
>> But the interesting
thing about my world

00:06:49.040 --> 00:06:54.077
is it's now gotten so large and you have
so many young people entering it who are

00:06:54.160 --> 00:06:58.637
there really are principled. They really
do care about climate change and society

00:06:58.720 --> 00:07:02.637
and ethics and so forth and so on, maybe
more so than my generation. There are

00:07:02.720 --> 00:07:06.397
plenty of opportunities for those people
to do this. There's plenty of nonprofits

00:07:06.480 --> 00:07:08.397
that need help addressing these issues.

00:07:08.480 --> 00:07:11.277
I think the underlying problem is we
don't know how to solve the problem yet.

00:07:11.360 --> 00:07:14.157
>> Yeah. Well, let's
get dangerous and dive

00:07:14.240 --> 00:07:16.957
into homelessness as I've experienced it

00:07:17.040 --> 00:07:19.517
here in Los Angeles. So, first and

00:07:19.600 --> 00:07:24.557
foremost, I'd like to say I've been
involved, be it fractionally compared to

00:07:24.640 --> 00:07:28.477
my wife and other friends of ours, but
like PATH is this great organization in

00:07:28.560 --> 00:07:33.277
Los Angeles and it helps get people
into permanent housing and there are some

00:07:33.360 --> 00:07:38.077
hurdles of course, but it's a great
program. So, I've been very in favor of

00:07:38.160 --> 00:07:41.837
adding services helping, helping,
helping. Something very peculiar

00:07:41.920 --> 00:07:43.997
happened within the first 6 weeks of

00:07:44.080 --> 00:07:46.797
quarantine here in LA, which is the

00:07:46.880 --> 00:07:49.517
corner where we would normally have a

00:07:49.600 --> 00:07:54.557
couple hundred people in a threeb block
area had shrank to like a few people.

00:07:54.640 --> 00:07:57.117
>> Mhm. >> And so I I
was asking myself, where did

00:07:57.200 --> 00:08:01.677
these homeless folks go? And you're
not going to like my conclusion and and

00:08:01.760 --> 00:08:04.477
America's not going to like
my conclusion, but my conclusion was

00:08:04.560 --> 00:08:06.717
there's nobody out to panhandle from.

00:08:06.800 --> 00:08:10.557
There's no source of money and no
human's going to sit on this sidewalk

00:08:10.640 --> 00:08:12.557
for another 10 weeks and evaporate.

00:08:12.640 --> 00:08:15.277
>> They went somewhere else.
>> They went somewhere. They contacted a

00:08:15.360 --> 00:08:19.117
relative they hadn't talked to. They had
done something. And at that moment, my

00:08:19.200 --> 00:08:21.197
whole lens kind of shifted. I was like,

00:08:21.280 --> 00:08:23.837
I'm not sure that throwing a bunch more

00:08:23.920 --> 00:08:26.397
services and money at this problem isn't

00:08:26.480 --> 00:08:29.517
in fact making it worse. Well, one

00:08:29.600 --> 00:08:32.557
answer which is the doctrinary answer is

00:08:32.640 --> 00:08:34.637
there is a real problem in housing in

00:08:34.720 --> 00:08:36.717
our country and if you look it's true

00:08:36.800 --> 00:08:39.997
globally why housing prices continue to

00:08:40.080 --> 00:08:44.077
accelerate and they've been doing it
for 20 years to the point where they're

00:08:44.160 --> 00:08:46.237
almost bankrupting shall we say normal

00:08:46.320 --> 00:08:48.717
jobs normal people that's a problem.

00:08:48.800 --> 00:08:53.197
Yes. Now some of that is due to the high
cost of manufacturing but some of that

00:08:53.280 --> 00:08:58.237
is also zoning. In California, for
example, there's a bill about trying to

00:08:58.320 --> 00:09:02.797
restrict communities to essentially
limit to just single family houses. You

00:09:02.880 --> 00:09:06.477
want to come up with multif family
solutions. In the Bay Area, where I

00:09:06.560 --> 00:09:10.557
spent many decades living, there's
a chronic housing shortage. The way New

00:09:10.640 --> 00:09:14.237
York solves that problem is by building
skyscrapers, which is precisely not what

00:09:14.320 --> 00:09:17.197
San Francisco seems to want to do. So,

00:09:17.280 --> 00:09:21.277
we've got to come up with a shared
agreement on how to get more housing and

00:09:21.360 --> 00:09:24.957
especially more affordable housing. The
current path where the developer gives a

00:09:25.040 --> 00:09:30.237
percentage of the housing is fine
but the scale of the unmet housing need is

00:09:30.320 --> 00:09:31.997
so large. >> Yeah.

00:09:32.080 --> 00:09:35.517
>> And it's a combination of cost but also

00:09:35.600 --> 00:09:39.117
regulation. And nobody wants the housing

00:09:39.200 --> 00:09:43.917
complex full of all the poor people
next to their expensive office or home or

00:09:44.000 --> 00:09:47.917
what have you. We have to come up with
some kind of compromise where people can

00:09:48.000 --> 00:09:50.397
coexist. Yeah. Um, if you look at the

00:09:50.480 --> 00:09:55.117
great cities of the world, the great
cities are where all the productivity

00:09:55.200 --> 00:09:58.477
growth is. It's where all of the
companies essentially are going to get

00:09:58.560 --> 00:10:02.637
born from. They're not going to get
born in rural places. It's just just the

00:10:02.720 --> 00:10:04.797
nature of how knowledge workers work.

00:10:04.880 --> 00:10:07.997
We've got to find a way to get more
people into cities. In China, they've

00:10:08.080 --> 00:10:11.117
solved this problem by building new
cities. And they built so many new

00:10:11.200 --> 00:10:14.557
cities, they've in fact got a credit
bubble and so forth. But at least

00:10:14.640 --> 00:10:16.557
they're trying to address
the housing problem.

00:10:16.640 --> 00:10:21.597
>> Yeah. And so right away you bring into
it the full scope of the problem which

00:10:21.680 --> 00:10:24.557
is it's so multifaceted and when
we're talking about homelessness are we

00:10:24.640 --> 00:10:29.437
talking about people there's some
segment that are yes one check away from

00:10:29.520 --> 00:10:33.277
not making their rent they're living
in a very insecure way that's one group of

00:10:33.360 --> 00:10:35.997
people there's also I did a ethnography

00:10:36.080 --> 00:10:39.597
in college in skid row and the vast vast

00:10:39.680 --> 00:10:43.757
majority of people I interviewed are
addicts primarily or people pretty

00:10:43.840 --> 00:10:48.397
severe mental health issues. Remember,
there's many sources of this and you do

00:10:48.480 --> 00:10:52.077
them one by one. >> You need
more housing. We need some new

00:10:52.160 --> 00:10:57.037
ways to treat mental illness and we need
some new ways to treat addiction and we

00:10:57.120 --> 00:11:00.477
also need to stop criminalizing
some addictive behaviors.

00:11:00.560 --> 00:11:03.197
>> Um, if you look at the Portugal
experiment where they essentially

00:11:03.280 --> 00:11:07.357
liberalized the laws with respect to all
hard drugs, what they did is they turned

00:11:07.440 --> 00:11:11.197
it from a criminal activity
to a medical addiction problem.

00:11:11.280 --> 00:11:15.437
>> Yeah. and all of a sudden people were
willing to seek help. The interesting

00:11:15.520 --> 00:11:19.757
thing is that crime went down
but addiction did not go down.

00:11:19.840 --> 00:11:23.117
>> So that tells you that
addiction is a harder problem.

00:11:23.200 --> 00:11:26.077
>> Yeah. >> Than it appears.
And that's why I said

00:11:26.160 --> 00:11:28.477
if our industry could come up with some

00:11:28.560 --> 00:11:31.117
new ideas and in fact I would argue that

00:11:31.200 --> 00:11:35.757
the current industry focus which
is around revenue is in fact playing into

00:11:35.840 --> 00:11:38.077
the addiction capabilities of every human.

00:11:38.160 --> 00:11:40.077
>> Yes. that what
happens with social media

00:11:40.160 --> 00:11:44.797
is you essentially become addiction.
And furthermore, what I'll say is that

00:11:44.880 --> 00:11:49.037
corporations, at least in social
media land, are optimized on maximizing

00:11:49.120 --> 00:11:51.277
revenue. To maximize revenue, you

00:11:51.360 --> 00:11:56.077
maximize engagement. To maximize
engagement, you maximize outrage. The

00:11:56.160 --> 00:11:58.877
same is true on television
news networks, by the way.

00:11:58.960 --> 00:12:00.877
>> Okay. So, I think
this is a great bridge

00:12:00.960 --> 00:12:03.917
into what could possibly be the utopian

00:12:04.000 --> 00:12:05.997
aspect of AI. there's going to be a

00:12:06.080 --> 00:12:08.637
dystopian aspect and a utopian aspect.

00:12:08.720 --> 00:12:12.797
But if we can imagine a world that Yuval
Harrari has suggested or many other

00:12:12.880 --> 00:12:17.997
people have suggested where machines
will ultimately do all the tasks that

00:12:18.080 --> 00:12:21.117
laborers are currently doing.
I know it's just one of the fun thought

00:12:21.200 --> 00:12:24.877
experiments we could run is whether
or not this job Monica and I could

00:12:24.960 --> 00:12:28.877
eventually be done. But what we're
going to earmark that in this world, this

00:12:28.960 --> 00:12:33.437
utopian world where we're not doing any
of the menial tasks anymore and we are

00:12:33.520 --> 00:12:35.757
really just being creative, obviously

00:12:35.840 --> 00:12:38.077
there has to be an enormous paradigm

00:12:38.160 --> 00:12:41.277
shift in money and earning and expending

00:12:41.360 --> 00:12:44.637
and living. So I guess in that utopian

00:12:44.720 --> 00:12:50.157
future, if money itself is not the
incentive, then these other things could

00:12:50.240 --> 00:12:53.757
be incentives, >> right?
So let's think about 200 years

00:12:53.840 --> 00:12:55.837
from now or 300 years from now.

00:12:55.920 --> 00:12:59.757
>> Yeah. >> Unfortunately, none of us will
be around, but our children's descendants

00:12:59.840 --> 00:13:03.437
will be. >> Eric Lander thinks
different. [laughter] Let's see.

00:13:03.520 --> 00:13:05.597
>> So the important point here is let's

00:13:05.680 --> 00:13:08.397
think beyond our current lifetimes. It's

00:13:08.480 --> 00:13:11.517
clear by the way that the lifestyle of a

00:13:11.600 --> 00:13:13.997
millionaire today will be affordable by

00:13:14.080 --> 00:13:16.717
a normal person because that has been

00:13:16.800 --> 00:13:19.357
the pattern. The pattern is that the

00:13:19.440 --> 00:13:22.157
cars and the TVs and so forth initially

00:13:22.240 --> 00:13:26.637
come out for the rich people and then as
prices come down, volume increases, they

00:13:26.720 --> 00:13:29.037
become more affordable. Many people

00:13:29.120 --> 00:13:31.437
believe that over a long enough period

00:13:31.520 --> 00:13:35.357
of time, the basic goods of society,

00:13:35.440 --> 00:13:40.557
eating, [clears throat] sleeping,
construction, Netflix, will all become

00:13:40.640 --> 00:13:42.157
affordable to everyone.

00:13:42.240 --> 00:13:45.757
>> Mhm. Now, this is a utopian view, but

00:13:45.840 --> 00:13:49.757
let's imagine if that were true
and let's say it were true globally to the

00:13:49.840 --> 00:13:52.237
11 billion people that we'll have past

00:13:52.320 --> 00:13:57.197
2100 on our globe. What would life
be like now? Then people immediately

00:13:57.280 --> 00:14:00.797
conclude that everyone will be watching
television and relaxing and lying by the

00:14:00.880 --> 00:14:03.357
pool, which is precisely not what's

00:14:03.440 --> 00:14:08.077
going to happen. What will happen
is humans will compete over other things.

00:14:08.160 --> 00:14:11.197
>> Yeah. >> Right. >> They'll
compete for complexity. They'll

00:14:11.280 --> 00:14:13.597
compete for sports. They'll compete for

00:14:13.680 --> 00:14:16.477
power. Humans are not going to give up

00:14:16.560 --> 00:14:20.397
identity. What's the scarcest thing in a

00:14:20.480 --> 00:14:23.837
world full of everything? Fame.

00:14:23.920 --> 00:14:27.037
>> Because it's the one thing
that everyone doesn't have.

00:14:27.120 --> 00:14:31.597
>> Yeah. Yeah. >> So, you'll
have huge competition over fame.

00:14:31.680 --> 00:14:35.677
>> When the way to understand social
networks is that every 10 years there's

00:14:35.760 --> 00:14:39.597
a new set of 18-year-olds,
if you will, who want to become famous.

00:14:39.680 --> 00:14:43.277
>> Yeah. and they look at the 28-year-olds
in the previous social network and they

00:14:43.360 --> 00:14:47.037
say, "I want that. I just have
that talent and so forth." But those

00:14:47.120 --> 00:14:51.837
positions are taken. So, the new social
network comes off with new players. We

00:14:51.920 --> 00:14:54.877
saw this with YouTube. We saw this
with Snapchat. We've seen this with

00:14:54.960 --> 00:14:57.837
Instagram. And we're clearly
seeing it with Tik Tok.

00:14:57.920 --> 00:14:59.837
>> Yeah. >> And by the way,
there'll be one after

00:14:59.920 --> 00:15:03.997
it. And they're roughly on a 10-year
basis because of this fame scarcity.

00:15:04.080 --> 00:15:06.957
>> Yeah. I [clears throat] would just
add we could just generally call that

00:15:07.040 --> 00:15:11.037
status. And as very social primates,
that's never leaving us. We are always

00:15:11.120 --> 00:15:13.277
going to be obsessed with
our ranking in our group.

00:15:13.360 --> 00:15:14.877
>> There's a famous economist named Herb

00:15:14.960 --> 00:15:19.997
Simon in 1971 said that the abundance of

00:15:20.080 --> 00:15:23.277
information leads to a scarcity of

00:15:23.360 --> 00:15:27.997
attention. And economists think not in
terms of abundance, but in terms of what

00:15:28.080 --> 00:15:30.717
is scarce, >> right?
What do we fight over?

00:15:30.800 --> 00:15:33.277
>> Right? >> So if we had
infinite money, we would

00:15:33.360 --> 00:15:35.117
fight over something else. Yeah.

00:15:35.200 --> 00:15:37.837
>> But the one thing
we don't have a lot of

00:15:37.920 --> 00:15:42.397
>> and will never have a lot of is social
hierarchy where everyone's the same.

00:15:42.480 --> 00:15:44.237
>> So we're going to compete for power and

00:15:44.320 --> 00:15:47.197
attention and whatever replaces money in

00:15:47.280 --> 00:15:51.677
this utopian scenario. A much more
likely scenario and it's always fun to

00:15:51.760 --> 00:15:53.837
talk about utopian and dystopian, but a

00:15:53.920 --> 00:15:58.957
much more likely scenario is coexistence
with AI and with all the things that

00:15:59.040 --> 00:16:03.037
we'll talk about will be happening. And
in this coexistence, we'll come up with

00:16:03.120 --> 00:16:06.957
new things to collaborate and new things
to compete with. But we're not going to

00:16:07.040 --> 00:16:11.917
be sitting at the beach in a world full
of resources. It's not how humans work.

00:16:12.000 --> 00:16:15.277
>> There's no pride in it, no honor in it,
no status in it, no no nothing in it.

00:16:15.360 --> 00:16:18.557
>> Well, and then using a stereotype, you
sit at the beach and you compete on the

00:16:18.640 --> 00:16:21.677
quality of your tan. >> Sure.
Sure. Where you got, what position

00:16:21.760 --> 00:16:23.357
you have, how close are you to the water?

00:16:23.440 --> 00:16:26.637
>> Yeah. You get the right seat.
We will compete over scarcity.

00:16:26.720 --> 00:16:30.797
>> I got to throw a nonsseor at you right
now. So, I met you and you didn't talk a

00:16:30.880 --> 00:16:32.957
lot. You're like a very generous host.

00:16:33.040 --> 00:16:36.157
You were the host of this thing I
attended and you let everyone else shine

00:16:36.240 --> 00:16:38.637
and now just within five minutes
of talking to you, I'm like, "This

00:16:38.720 --> 00:16:40.557
[__] is brilliant."
>> That's the way to [laughter] do it.

00:16:40.640 --> 00:16:43.037
That's what you do. Slow play.

00:16:43.120 --> 00:16:46.397
>> Yeah. I love just the overall
comprehensive knowledge you have.

00:16:46.480 --> 00:16:49.357
>> That's really kind of you, Dax.
I'll give you my life advice.

00:16:49.440 --> 00:16:52.717
>> Yes. >> Which is ask a lot
of questions and be curious. M

00:16:52.800 --> 00:16:55.037
>> if you ask a lot of questions, you can

00:16:55.120 --> 00:16:59.757
be annoying, but people are always
flattered if you ask them their opinion.

00:16:59.840 --> 00:17:01.677
>> Oh, yeah. >> Always works. >> Mhm.

00:17:01.760 --> 00:17:03.597
>> And if you're
actually listening to them

00:17:03.680 --> 00:17:05.997
and curious, you might learn something.

00:17:06.080 --> 00:17:10.477
And frankly, it's so much easier to let
people who are experts explain things to

00:17:10.560 --> 00:17:12.877
you then do your own research. I learned

00:17:12.960 --> 00:17:16.477
a while ago not to trust marketing. So

00:17:16.560 --> 00:17:20.957
whenever people pitch me, I always
wonder where is that coming from? what

00:17:21.040 --> 00:17:22.637
are they trying to get out of me?

00:17:22.720 --> 00:17:26.557
>> But when I've got somebody who has
a perspective on a problem and who can

00:17:26.640 --> 00:17:29.757
give me the scope of it,
I want to learn from them.

00:17:29.840 --> 00:17:32.957
>> Yeah. I mean, we're doing it
professionally, which is a real hack. I

00:17:33.040 --> 00:17:34.717
think >> this is sort of what you do.

00:17:34.800 --> 00:17:38.317
>> Yes. Yes. Yes. And and we're we're in
>> But if everyone but if everyone modeled

00:17:38.400 --> 00:17:41.037
your behavior, we wouldn't have these

00:17:41.120 --> 00:17:45.917
essentially amplification scale things
where everybody all of a sudden believes

00:17:46.000 --> 00:17:48.957
something without >> any
kind of criticism. Yeah,

00:17:49.040 --> 00:17:51.837
>> at Google when I first started there,

00:17:51.920 --> 00:17:54.477
>> this is now 20 years ago, we sort of

00:17:54.560 --> 00:17:56.557
adopted a policy that whenever anyone

00:17:56.640 --> 00:17:59.197
would say something, we would query it.

00:17:59.280 --> 00:18:04.237
The first CFO we had was making
a presentation before we were public and

00:18:04.320 --> 00:18:07.917
he made a statement which
I thought might be false.

00:18:08.000 --> 00:18:12.717
And so he's talking and I checked it.
It was a minor thing, >> right?

00:18:12.800 --> 00:18:16.157
>> But it's an example of how
you get high performance teams,

00:18:16.240 --> 00:18:20.317
>> right? People operate on all sorts
of beliefs which are not based in facts.

00:18:20.400 --> 00:18:22.637
The Google principle was check it.

00:18:22.720 --> 00:18:24.637
>> No, you're so right. I think it's often

00:18:24.720 --> 00:18:29.037
tempting to reduce people to their

00:18:29.120 --> 00:18:34.077
specialty and then ignore the vastness
of their humanity which is like

00:18:34.160 --> 00:18:37.597
everyone's got an ego. Everyone feels
embarrassed. Everyone feels called out.

00:18:37.680 --> 00:18:41.677
>> These have to always be in the stew
when we're making these decisions.

00:18:41.760 --> 00:18:45.917
>> But what's happening because of the
amplification? It's easier to use social

00:18:46.000 --> 00:18:50.237
media as an example. Yeah. My book,
our book is really about what happens when

00:18:50.320 --> 00:18:53.997
this thing gets much worse. Yes. Which
is the dystopian side. But I will tell

00:18:54.080 --> 00:18:56.397
you that >> the fact
that everyone believes

00:18:56.480 --> 00:18:59.677
something does not necessarily mean
it be true. The fact that your friend

00:18:59.760 --> 00:19:01.997
believes it does not
necessarily mean it be true.

00:19:02.080 --> 00:19:05.917
>> And because of the overload of
information and all the people purveying

00:19:06.000 --> 00:19:10.717
you for whatever reason, whether it's
the Russians or marketing or evil people

00:19:10.800 --> 00:19:14.397
that you know or what have you, check
it out on your own. It's not a bad way to

00:19:14.480 --> 00:19:17.597
live your life. Be a little skeptical.

00:19:17.680 --> 00:19:22.637
>> Be skeptical about what I say. Check
what I say. See if you agree with me.

00:19:22.720 --> 00:19:25.277
I'd welcome that. >> We interviewed
a great professor down at

00:19:25.360 --> 00:19:28.157
Texas Tech and he was saying in general,

00:19:28.240 --> 00:19:34.077
if when you're receiving information,
it tastes just like candy. That's a red

00:19:34.160 --> 00:19:36.957
flag. >> By the way,
that is a great analogy. I

00:19:37.040 --> 00:19:39.037
love that. What is happening now? It's

00:19:39.120 --> 00:19:45.037
boom boom boom. >> And we're trying
to keep you going in the game.

00:19:45.120 --> 00:19:48.237
>> Yes. >> And so one
of the tricks is turn off the

00:19:48.320 --> 00:19:53.597
television and turn off your news feed
and live in a local world and see how

00:19:53.680 --> 00:19:55.357
you feel. >> Yeah.

00:19:55.440 --> 00:19:58.237
>> As a matter of philosophy.
And we should

00:19:58.320 --> 00:20:02.877
be thankful that we are here, we're alive,
we're in a great country, we have

00:20:02.960 --> 00:20:06.957
all these great opportunities and so
forth. And instead, all the surveys

00:20:07.040 --> 00:20:09.997
indicate that we're
more unhappy than ever.

00:20:10.080 --> 00:20:13.037
>> Yeah. >> Yeah.
I think he said critical thinking,

00:20:13.120 --> 00:20:15.037
if you're doing it right, should hurt.

00:20:15.120 --> 00:20:16.957
>> Yeah. >> And I I think that's right.

00:20:17.040 --> 00:20:19.757
>> That's a very bad marketing message.

00:20:19.840 --> 00:20:22.077
>> But it's correct. >> Yeah. Exactly.

00:20:22.160 --> 00:20:26.157
>> Okay. So, let's walk through
for a second. I do have a couple questions

00:20:26.240 --> 00:20:29.757
from your childhood just because
they jumped out at me as potentially

00:20:29.840 --> 00:20:33.357
interesting. Well, first of all,
your mother had a master's in psychology.

00:20:33.440 --> 00:20:38.477
Your father was an economics professor
at both what? Virginia Tech and at

00:20:38.560 --> 00:20:42.557
John's Hopkins. >> Yes. And at George
Washington, where I was born.

00:20:42.640 --> 00:20:44.557
>> Uhhuh. >> Well, I said,
"Why was I born there?"

00:20:44.640 --> 00:20:49.117
And he said, "It's really simple.
Birth was free for professors."

00:20:49.200 --> 00:20:51.597
>> Yeah. Done deal. >> And I said,
"Now I've learned something

00:20:51.680 --> 00:20:53.597
about economists." [laughter]

00:20:53.680 --> 00:20:58.637
>> Of course. But also worked for the
Treasury. And at some point you moved to

00:20:58.720 --> 00:21:02.157
Italy. What ages was that?
>> I was eight. You were eight. I moved to

00:21:02.240 --> 00:21:04.717
Europe for two years as a boy and I can

00:21:04.800 --> 00:21:07.037
tell you that it was an enormously

00:21:07.120 --> 00:21:11.837
transformative experience because at the
time Americans didn't travel to Europe

00:21:11.920 --> 00:21:13.917
the way they do now

00:21:14.000 --> 00:21:16.877
>> and so this was post-war Europe still

00:21:16.960 --> 00:21:19.277
very poor compared to the United States.

00:21:19.360 --> 00:21:23.517
We lived very well on a professor salary
because of what is unimaginable today.

00:21:23.600 --> 00:21:25.597
Europe was very inexpensive. >> Sure.

00:21:25.680 --> 00:21:29.837
>> But it caused me to appreciate the
global nature of life. Italians are the

00:21:29.920 --> 00:21:32.477
warmest people you'll ever meet. It's a

00:21:32.560 --> 00:21:35.037
great history and a great culture. I was

00:21:35.120 --> 00:21:39.677
there recently to receive an award and
give a speech where I grew up. And I was

00:21:39.760 --> 00:21:43.837
really struck that this is a country
which had this horrific co thing in

00:21:43.920 --> 00:21:48.397
Lombardi in February and March of last
year, but they've made the necessary

00:21:48.480 --> 00:21:50.397
changes. Their identity is strong and

00:21:50.480 --> 00:21:52.477
they're growing well. Who would have

00:21:52.560 --> 00:21:55.597
expected such a strong narrative in the

00:21:55.680 --> 00:21:58.717
EU from Italy of all countries? Yeah.

00:21:58.800 --> 00:22:02.797
>> So, it shows you that human
systems can respond to the challenge.

00:22:02.880 --> 00:22:05.277
>> Now, did you feel
like an outsider there?

00:22:05.360 --> 00:22:08.957
>> They made me very welcome. I spoke
Italian very quickly and uh yeah, so I

00:22:09.040 --> 00:22:13.357
was treated very well. If you have an
opportunity to take your family and have

00:22:13.440 --> 00:22:15.917
your kids live in a European country,

00:22:16.000 --> 00:22:17.997
>> um it's enormously broadening. When I

00:22:18.080 --> 00:22:21.277
came back to America, I realized how

00:22:21.360 --> 00:22:23.757
completely nivist, how completely

00:22:23.840 --> 00:22:25.837
self-obsessed the United States is. A

00:22:25.920 --> 00:22:31.597
good example is watch Fox or CNN
and then watch BBC news. You'll see the

00:22:31.680 --> 00:22:34.797
difference >> in that they
are focusing much more outward.

00:22:34.880 --> 00:22:37.997
>> The majority of the people in the world
do not live in the United States. The

00:22:38.080 --> 00:22:41.037
majority therefore of the drama
is outside the United States.

00:22:41.120 --> 00:22:44.077
>> Yes. Yes. >> The news
is therefore slanted in the US

00:22:44.160 --> 00:22:47.997
in the ways that we understand.
And if you look today, the cable news is

00:22:48.080 --> 00:22:50.477
essentially obsessed about politics. And

00:22:50.560 --> 00:22:55.117
we've lost our local news leadership
in America because of the economics of the

00:22:55.200 --> 00:22:56.797
internet and so forth. >> The incentives

00:22:56.880 --> 00:23:01.517
>> and so where are the local stories?
Here we are in Los Angeles. There's 10

00:23:01.600 --> 00:23:05.437
million people in the greater Los
Angeles basin. Actually more than that.

00:23:05.520 --> 00:23:08.557
>> There's a lot going on. >> Oh yeah.
Yeah. Yeah. Okay. Now when you

00:23:08.640 --> 00:23:11.117
went to college initially you majored in

00:23:11.200 --> 00:23:16.237
architecture and I just had a tiny
curiosity if you had fantasies of being

00:23:16.320 --> 00:23:19.837
Howard Ror. >> No.
But what happened was I liked

00:23:19.920 --> 00:23:24.477
architecture and I applied as an
architecture candidate and I was just

00:23:24.560 --> 00:23:27.037
interested in it. And what I realized
very quickly is I was a terrible

00:23:27.120 --> 00:23:28.877
architect but I was quite a good engineer.

00:23:28.960 --> 00:23:31.677
>> Uhhuh. >> Because I
was not creative enough and I

00:23:31.760 --> 00:23:33.837
continue to believe that is true today.

00:23:33.920 --> 00:23:36.637
>> Um the other thing that was interesting
about the architecture people is they

00:23:36.720 --> 00:23:38.797
worked harder than anyone else.

00:23:38.880 --> 00:23:41.277
Architecture as a major is rough. Right.

00:23:41.360 --> 00:23:45.037
Really? So it was interesting and all
I could tell you is my interest in

00:23:45.120 --> 00:23:49.517
architecture has translated into
the architecture of corporations.

00:23:49.600 --> 00:23:53.517
>> So people who are engineering
oriented architects they think of scale

00:23:53.600 --> 00:23:57.677
>> in computer scientists again I was
an early part of computer science when I

00:23:57.760 --> 00:24:01.277
was at Princeton there was no computer
science program at all. Today it is the

00:24:01.360 --> 00:24:03.357
number one major of the entire university.

00:24:03.440 --> 00:24:06.077
>> Wow. So in two generations, it's

00:24:06.160 --> 00:24:10.877
actually gone from zero to highest
ranking. And that is representative of

00:24:10.960 --> 00:24:14.317
the transformation that our societyy's
going to see. Independent of all the

00:24:14.400 --> 00:24:17.837
crazy stuff going on in our government,
in our local environment, and here in

00:24:17.920 --> 00:24:19.757
Los Angeles, the fact of the matter is

00:24:19.840 --> 00:24:23.597
this huge cohort of early 20some men and

00:24:23.680 --> 00:24:27.677
women, and I'm proud to say that it's
almost half women now. Yeah. Which is

00:24:27.760 --> 00:24:30.557
which is an amazing achievement. They're
going to come in and they're going to

00:24:30.640 --> 00:24:34.157
apply these tools to the problems around
us. and they're going to make money and

00:24:34.240 --> 00:24:36.557
change the world and all that in ways
that are appropriate for their

00:24:36.640 --> 00:24:39.917
generation. But when I was there, there
was only, you know, one computer and we

00:24:40.000 --> 00:24:43.997
stayed up all night because it was shared
and it was too slow during the day.

00:24:44.080 --> 00:24:47.437
>> Right. And even your doctorate degree
which you ended up getting for Berkeley,

00:24:47.520 --> 00:24:50.477
it's like what's the degree
ECCU or something?

00:24:50.560 --> 00:24:55.437
>> ECS double E because computer science
wasn't there, >> right? So even your

00:24:55.520 --> 00:24:59.197
>> So my point is that people don't
appreciate how much has changed. I'll

00:24:59.280 --> 00:25:00.637
give you another example. The computer

00:25:00.720 --> 00:25:05.037
that I used in Princeton is 100 million

00:25:05.120 --> 00:25:08.397
times slower than an iPhone or an

00:25:08.480 --> 00:25:13.917
Android phone. [snorts] >> I didn't
say 100,000. I said 100 million times.

00:25:14.000 --> 00:25:17.597
>> That's crazy.
>> And the number one import to China is

00:25:17.680 --> 00:25:21.597
semiconductors >> made elsewhere.
It shows you how

00:25:21.680 --> 00:25:26.077
powerful the digital revolution is and
and what is at stake. Well, I like the

00:25:26.160 --> 00:25:27.997
parallels people have been making in

00:25:28.080 --> 00:25:31.357
regards to Taiwan and that Taiwan will

00:25:31.440 --> 00:25:34.317
most likely be seen as a Kuwait was in

00:25:34.400 --> 00:25:39.117
1990 where we cannot allow it to be
taken over. It's too vital to the world

00:25:39.200 --> 00:25:41.997
economy. >> I think that's
the consensus and I'll

00:25:42.080 --> 00:25:43.197
give you the background. By the way,

00:25:43.280 --> 00:25:46.317
there's a fab system foundry as it

00:25:46.400 --> 00:25:49.597
technically known in called TSMC, Taiwan

00:25:49.680 --> 00:25:52.317
Semiconductor Corporation and it was

00:25:52.400 --> 00:25:54.877
founded by a fellow named Morris who

00:25:54.960 --> 00:26:00.237
actually was a physicist at Berkeley
and so forth in the 60s and 70s and he was

00:26:00.320 --> 00:26:01.837
trained in America and he went back to

00:26:01.920 --> 00:26:05.357
Taiwan which at that time was very

00:26:05.440 --> 00:26:07.597
western compared to China. Mhm.

00:26:07.680 --> 00:26:11.277
>> And so working in Taiwan, he managed
to build this extraordinary company which

00:26:11.360 --> 00:26:15.917
is worth many hundreds of billions
of dollars. And it is roughly half of the

00:26:16.000 --> 00:26:21.197
world's foundaries for the key chips
that we all use. And it's a triumph of

00:26:21.280 --> 00:26:22.797
engineering. It's a triumph of science.

00:26:22.880 --> 00:26:26.477
And it's also a triumph of government
policy because they received enormous

00:26:26.560 --> 00:26:30.317
financial support from Taiwan to do this.
So again, just like we had with

00:26:30.400 --> 00:26:33.917
Operation Warp Speed, we had
the government procuring the vaccines

00:26:34.000 --> 00:26:38.157
whether they worked or not. You had huge
private sector risks. Remember, most the

00:26:38.240 --> 00:26:41.997
vaccines didn't work. And then we
had the universities helping in a real

00:26:42.080 --> 00:26:43.997
national emergency. That's what it took.

00:26:44.080 --> 00:26:47.117
So Taiwan did this. So now we face the

00:26:47.200 --> 00:26:51.197
question and I did a report uh for the AI
commission for the Congress which

00:26:51.280 --> 00:26:54.557
said that we have to stay two
generations ahead of China, mainland

00:26:54.640 --> 00:26:57.437
China in semiconductors. Well, we are

00:26:57.520 --> 00:27:00.397
well behind Taiwan. So it's Taiwan

00:27:00.480 --> 00:27:02.797
first, the US second, and China third.

00:27:02.880 --> 00:27:07.437
The interesting thing geopolitically
is that China has spent essentially

00:27:07.520 --> 00:27:11.517
infinite money to catch up
on semiconductors. They literally have the

00:27:11.600 --> 00:27:13.917
top talent and so forth, but it's so

00:27:14.000 --> 00:27:17.757
hard that the choke points remain key

00:27:17.840 --> 00:27:22.077
hardware design things which are called
extreme ultraviolet. There's a company

00:27:22.160 --> 00:27:26.397
called ASML in in the Netherlands which
is the only company that knows how to do

00:27:26.480 --> 00:27:29.117
this. >> You need this to go
below what is called

00:27:29.200 --> 00:27:32.397
the 10 nanometer line. And the other

00:27:32.480 --> 00:27:35.837
thing that has occurred is that the cost

00:27:35.920 --> 00:27:41.117
of these fabs are so great there'll be
relatively few of them. The US and China

00:27:41.200 --> 00:27:44.157
have both demanded that Taiwan build
such fabs in the United States. And

00:27:44.240 --> 00:27:47.117
guess what? They're in the process
of building such fabs, but they're not

00:27:47.200 --> 00:27:49.117
putting at the state-of-the-art. So if

00:27:49.200 --> 00:27:53.117
you use for example the Apple Mac M1 Pro

00:27:53.200 --> 00:27:56.237
or M1 Max chips, those are so-called

00:27:56.320 --> 00:27:59.117
5nanmter technology from Taiwan. They

00:27:59.200 --> 00:28:02.797
are the product of an enormous enormous

00:28:02.880 --> 00:28:05.677
investment at Apple but also in Taiwan.

00:28:05.760 --> 00:28:07.837
>> Mhm. >> The other competitor,
so you know who it

00:28:07.920 --> 00:28:10.397
is, is Samsung and that's in South Korea.

00:28:10.480 --> 00:28:13.480
>> Mhm.

00:28:13.840 --> 00:28:17.037
Stay tuned for more armchair expert

00:28:17.120 --> 00:28:21.000
[music] if you dare.

00:28:21.280 --> 00:28:25.517
We are supported by Autotrader.
All right, if you've been putting off car

00:28:25.600 --> 00:28:29.197
shopping because it sounds like
a whole thing, buying your car online on

00:28:29.280 --> 00:28:33.917
Autotrader makes it way easier.
Really, you can shop tons of cars online,

00:28:34.000 --> 00:28:38.237
compare prices, check financing,
and tradein options all from wherever you

00:28:38.320 --> 00:28:41.917
are. And if you want to start online
and go to the dealership for a test drive,

00:28:42.000 --> 00:28:46.317
you can do that, too. Really, what I
like is the flexibility. Start online,

00:28:46.400 --> 00:28:50.157
finish at the dealer, or the other way
around. Your info stays with you, so

00:28:50.240 --> 00:28:53.597
you're not starting the process from
scratch. Plus, you're working with local

00:28:53.680 --> 00:28:57.357
dealer listings, so you've got real
people helping behind the scenes. And

00:28:57.440 --> 00:28:59.037
when you're ready, you can pick up your

00:28:59.120 --> 00:29:01.757
car or have it delivered. Really, buy

00:29:01.840 --> 00:29:05.517
your car online on Autotrader. Really,

00:29:05.600 --> 00:29:08.397
this episode is sponsored by BetterHelp.

00:29:08.480 --> 00:29:11.997
You've heard us talk about BetterHelp
for a while, but I don't know that we've

00:29:12.080 --> 00:29:15.517
ever talked about how much people
love this service when they use it.

00:29:15.600 --> 00:29:17.437
BetterHelp has an average live therapy

00:29:17.520 --> 00:29:20.077
session rating of 4.9 out of five based

00:29:20.160 --> 00:29:22.637
on over 1.7 million client session

00:29:22.720 --> 00:29:24.557
reviews. And you can check out the

00:29:24.640 --> 00:29:27.677
reviews yourself at betterhelp.com/res

00:29:27.760 --> 00:29:30.077
to see how much they love BetterHelp.

00:29:30.160 --> 00:29:31.917
Knowing that kind of service feedback

00:29:32.000 --> 00:29:34.877
from real people speaks for itself. I'm

00:29:34.960 --> 00:29:37.117
a huge proponent of therapy, as we've

00:29:37.200 --> 00:29:39.517
talked about before. I've learned so

00:29:39.600 --> 00:29:42.397
many things about myself that I believe

00:29:42.480 --> 00:29:45.277
truly only an outside objective person

00:29:45.360 --> 00:29:49.597
could help me come to. BetterHelp makes
it easy to get started. Answer a few

00:29:49.680 --> 00:29:52.957
questions, get matched with a therapist
based on your needs, and switch

00:29:53.040 --> 00:29:57.517
therapist at any time if it's not
the right fit. See the reviews, see what

00:29:57.600 --> 00:29:59.597
stands out, and see if BetterHelp is

00:29:59.680 --> 00:30:03.357
right for you. Visit betterhelp.com/dax.

00:30:03.440 --> 00:30:06.760
That's betterhp.com/dax.

00:30:08.160 --> 00:30:13.037
>> Thank you to our presenting sponsor,
HP, for supporting this episode. Parents,

00:30:13.120 --> 00:30:17.037
you've got enough on your back to school
list right now. Ink cartridges do not

00:30:17.120 --> 00:30:21.437
need to be on it. They just don't. The HP
Smart Tank printer uses a refillable

00:30:21.520 --> 00:30:25.917
ink tank and comes with up to 3 years
of ink included. It's the last printer

00:30:26.000 --> 00:30:27.997
you'll ever need. To learn more, search

00:30:28.080 --> 00:30:30.637
HP Smart Tank.

00:30:30.720 --> 00:30:33.917
>> This episode is supported by Helix. All

00:30:34.000 --> 00:30:38.477
right, let's talk about something we
both really care deeply about. Sleep.

00:30:38.560 --> 00:30:42.717
>> It's the most important thing. I've
had my Helix mattress for years now, and I

00:30:42.800 --> 00:30:45.757
genuinely think it's one of the best
investments you can make for your home.

00:30:45.840 --> 00:30:48.157
Mine still feels just like
it did when I first got it.

00:30:48.240 --> 00:30:51.677
>> Well, that's the thing. It's not one
of those mattresses that caves in after 6

00:30:51.760 --> 00:30:53.997
months. These things are built to last.

00:30:54.080 --> 00:30:56.237
And you know, I'm a side sleeper. I

00:30:56.320 --> 00:30:58.957
ordered a side sleeper and I get hot so

00:30:59.040 --> 00:31:03.517
I have cooling and I can't tell you
how much more comfortable it is to be in

00:31:03.600 --> 00:31:05.757
there not sweating
and tossing and turning.

00:31:05.840 --> 00:31:09.917
>> Cooling is really important. I've read
that you're supposed to sleep a little

00:31:10.000 --> 00:31:14.477
bit cold. >> So, I also have cooling. And
they have over 20 models, so it's not a

00:31:14.560 --> 00:31:17.997
one-sizefits-all situation. You just
go take their quiz and they match you to

00:31:18.080 --> 00:31:21.677
what actually works for your body.
Plus, they've got cooling upgrades, which

00:31:21.760 --> 00:31:23.757
heading into summer, non-negotiable.

00:31:23.840 --> 00:31:27.437
>> Truly. >> Go to helixsleep.com/armchair

00:31:27.520 --> 00:31:31.597
for 20% off.
That's helixleep.com/armchchair

00:31:31.680 --> 00:31:33.757
for 20% off. Make sure you enter our

00:31:33.840 --> 00:31:39.400
show name after checkout so they
know we sent you. Helix.com/armmchair.

00:31:44.760 --> 00:31:46.780
[music]

00:31:49.280 --> 00:31:52.237
Now, here's where it gets a little
tricky, right? As I understand it, as

00:31:52.320 --> 00:31:56.877
you point out, the investment is so large,
you need state cooperation, and I

00:31:56.960 --> 00:32:00.237
have a very cursory knowledge of it from
watching a 60-minute segment on it. But

00:32:00.320 --> 00:32:02.237
Intel, of course, had been the leader.

00:32:02.320 --> 00:32:05.437
And they didn't really have any kind of

00:32:05.520 --> 00:32:10.557
US support. It was not invested by us
in the same way Taiwan invested in that

00:32:10.640 --> 00:32:12.877
company. That's correct. >> And then
now we get into a really tricky

00:32:12.960 --> 00:32:17.117
scenario because the headline will
read like US propping up trillion dollar

00:32:17.200 --> 00:32:21.357
company. So, it looks like it's
a corporate welfare program and I think we

00:32:21.440 --> 00:32:23.757
lose sight of the national
security aspect.

00:32:23.840 --> 00:32:27.437
>> And by the way, that's a political
problem in our system that we're going

00:32:27.520 --> 00:32:29.517
to have to overcome. But let me tell you

00:32:29.600 --> 00:32:31.437
that China doesn't have this problem.

00:32:31.520 --> 00:32:36.557
>> China is pouring money to achieve
chip leadership, which will hurt American

00:32:36.640 --> 00:32:39.837
firms, >> every industry that we're
on the precipice of competing over.

00:32:39.920 --> 00:32:42.077
>> Yeah. And in fact,
to continue on China,

00:32:42.160 --> 00:32:44.797
China has defined the following markets

00:32:44.880 --> 00:32:47.117
as interesting to it. artificial

00:32:47.200 --> 00:32:50.557
intelligence, software, AI, quantum

00:32:50.640 --> 00:32:53.917
computing, energy, financial services,

00:32:54.000 --> 00:32:58.317
and essentially tracking and face
recognition and a few others. Well,

00:32:58.400 --> 00:33:02.477
that's essentially the whole world I
occupy, right? That's everything that I

00:33:02.560 --> 00:33:07.037
am working on in my history. It's the
source of the primary profits for all of

00:33:07.120 --> 00:33:09.437
the globe's fastest growing companies.

00:33:09.520 --> 00:33:11.597
All of the stock market winners have

00:33:11.680 --> 00:33:16.477
essentially been as a basis of that.
So the Chinese are very clear on where

00:33:16.560 --> 00:33:20.397
their priorities are. What's the
equivalent list in the United States? We

00:33:20.480 --> 00:33:24.797
don't have one. We can argue over steel
and corn and weed and so forth. Those

00:33:24.880 --> 00:33:26.317
are fine things to argue about.

00:33:26.400 --> 00:33:31.357
>> They are not strategic. >> No, >>
occupying the strategic platform for the

00:33:31.440 --> 00:33:34.717
globe is crucial. And let me give
you an example of how we screwed up.

00:33:34.800 --> 00:33:36.877
>> Really quick,
bioengineering is in there as well.

00:33:36.960 --> 00:33:41.117
>> Excitted synthetic bio. That's
a terrible omission on my part because

00:33:41.200 --> 00:33:44.877
it's going to be immense. China's
building a bio bank that is even larger

00:33:44.960 --> 00:33:48.637
than anybody else's. So again,
it's coming and we need to get organized

00:33:48.720 --> 00:33:51.197
around this. How do we do that? Well,

00:33:51.280 --> 00:33:53.917
>> first we need a national plan and the

00:33:54.000 --> 00:33:58.317
national plan has to involve some
amount of financial support from the

00:33:58.400 --> 00:34:02.397
government. The government always has
supported basic research. Thank you very

00:34:02.480 --> 00:34:06.797
much. It's at the lowest level it's been
since Splutnick. That's not good. But

00:34:06.880 --> 00:34:09.837
the other thing that happened was we
got out of the semiconductor business and

00:34:09.920 --> 00:34:13.677
out of the 5G business in the United
States with the exception of Intel and a

00:34:13.760 --> 00:34:17.997
few other cases because the profits
weren't there. It was better to be

00:34:18.080 --> 00:34:22.157
linked to supply chain to Asia primarily.
Well, now with these national

00:34:22.240 --> 00:34:26.397
security concerns, we need to do some
reshoring. Reshoring in an industry

00:34:26.480 --> 00:34:30.157
which we left is expensive.
Somebody's got to pay for it.

00:34:30.240 --> 00:34:33.037
>> Furthermore, let me give you two
examples to illustrate this point of

00:34:33.120 --> 00:34:35.757
what happens if we screw this up. Huawei

00:34:35.840 --> 00:34:38.557
is a state subsidized manufacturer in

00:34:38.640 --> 00:34:41.197
China. They build really, really

00:34:41.280 --> 00:34:46.397
excellent 5G infrastructure. They're
also full of national security concerns.

00:34:46.480 --> 00:34:49.517
The United States is trying to get
people not to use Huawei, especially

00:34:49.600 --> 00:34:52.717
missionritical things, but we don't
have an alternative because we got out of

00:34:52.800 --> 00:34:55.757
that business. >> Mhm.
>> That's a problem. We need to fix that.

00:34:55.840 --> 00:34:57.757
I'll give you another example. Tik Tok.

00:34:57.840 --> 00:35:00.637
Now, Tik Tok's great Chinese company.

00:35:00.720 --> 00:35:03.197
People love it. It's the new big thing.

00:35:03.280 --> 00:35:06.637
It's going to be on its way to be
a trillion dollar company. It has huge

00:35:06.720 --> 00:35:08.797
growth everywhere. President Trump tried

00:35:08.880 --> 00:35:11.757
to fix the issue of China with Tik Tok

00:35:11.840 --> 00:35:14.157
by requiring local ownership. That deal

00:35:14.240 --> 00:35:16.717
ultimately was not consummated. So, we

00:35:16.800 --> 00:35:18.957
need a solution to Tik Tok. Now, what's

00:35:19.040 --> 00:35:23.757
the issue with Tik Tok? I don't mind the
Chinese knowing where our teenagers are.

00:35:23.840 --> 00:35:27.517
I'd like someone to know where they are,
right? It's not a problem, right? We

00:35:27.600 --> 00:35:29.997
love our teenagers. They're safe. But I

00:35:30.080 --> 00:35:34.077
do mind if Tik Tok becomes censored or

00:35:34.160 --> 00:35:37.837
an agent for a foreign power of any kind.

00:35:37.920 --> 00:35:40.237
>> That would be a violation of our norms.

00:35:40.320 --> 00:35:45.357
So when you play with these systems that
are global in nature and you don't have

00:35:45.440 --> 00:35:49.277
your own local alternative and there's
innovation occurring elsewhere, you give

00:35:49.360 --> 00:35:54.237
up not just the economic opportunity
which is a trillion dollar econ which we

00:35:54.320 --> 00:35:56.237
want because we want the jobs but you

00:35:56.320 --> 00:35:58.477
also give up the opportunity for control

00:35:58.560 --> 00:36:02.157
in innovation. And in my world, broadly

00:36:02.240 --> 00:36:04.317
speaking, the tech world, the first

00:36:04.400 --> 00:36:06.957
mover really does have an advantage. And

00:36:07.040 --> 00:36:11.437
once they get a big scale, it's really
hard for a new entrance to come in for

00:36:11.520 --> 00:36:13.037
all sorts of reasons. >> Yeah.

00:36:13.120 --> 00:36:15.277
>> So, we are in a window and the report

00:36:15.360 --> 00:36:17.517
that we published for the Congress said

00:36:17.600 --> 00:36:22.317
we're in a window of a few years
where the decisions that we make will

00:36:22.400 --> 00:36:24.317
determine whether we ultimately control

00:36:24.400 --> 00:36:29.357
our digital future or whether someone
else likely China does. And we have a

00:36:29.440 --> 00:36:33.757
bunch of recommendations which include
more R&D funding, a whole bunch of of

00:36:33.840 --> 00:36:37.437
infrastructure that would help startups,
pro-co competitive moves, working with

00:36:37.520 --> 00:36:41.757
our partners, and most important
reestablishing the principle of our

00:36:41.840 --> 00:36:45.757
ethics, right? What are the ethics we
care about? We care about free speech.

00:36:45.840 --> 00:36:49.677
Are there limits of free speech? We care
about access. Are there limits? We care

00:36:49.760 --> 00:36:53.197
a lot about prejudice. What is an okay
prejudice and what is a not okay

00:36:53.280 --> 00:36:55.357
prejudice? We got to get that right now.

00:36:55.440 --> 00:36:59.277
Mhm. >> And I fear given
our political confusion

00:36:59.360 --> 00:37:03.277
and these weird sort of nivist things
like we don't want immigrants, this is

00:37:03.360 --> 00:37:07.837
where the growth is, guys. This is what
drives the economy. If you look, and

00:37:07.920 --> 00:37:11.117
it's great to have a recovery,
the majority of the economic growth is

00:37:11.200 --> 00:37:16.557
occurring from six or seven states.
And within those, it's a very small

00:37:16.640 --> 00:37:20.477
percentage of the counties in our country.
3,500 counties, it's something

00:37:20.560 --> 00:37:25.037
like 30 or 40 counties drive the vast
majority of the wealth creation. And I'm

00:37:25.120 --> 00:37:27.597
not talking about personal wealth.
I'm talking about societal wealth.

00:37:27.680 --> 00:37:31.917
>> Yeah. >> Are young people
here getting into these

00:37:32.000 --> 00:37:36.237
fields though? Because I know in other
countries they are. And I feel like

00:37:36.320 --> 00:37:41.117
that's part of the issue is there's no
incentive for younger. They want to be

00:37:41.200 --> 00:37:45.597
influencers. >> So the good
news is that this next

00:37:45.680 --> 00:37:47.517
generation of people in our universities

00:37:47.600 --> 00:37:52.637
is phenomenal. They are so much quicker
than I was. partly because they're just

00:37:52.720 --> 00:37:55.757
I think they actually are smarter, but
also because they live in a world that's

00:37:55.840 --> 00:37:58.237
much quicker, you know. So when I teach,

00:37:58.320 --> 00:38:02.957
I watch the students and I go, "Boy,
I wasn't that fast. I wasn't that crisp."

00:38:03.040 --> 00:38:04.317
>> And I've seen that over and over again.

00:38:04.400 --> 00:38:06.797
So I think there's
a reason to be optimistic.

00:38:06.880 --> 00:38:10.877
>> To me, the question is not the people
because I think the people are there.

00:38:10.960 --> 00:38:14.877
The American educational system produces
the really top people at the collegiate

00:38:14.960 --> 00:38:17.437
level. The real problem is what is the

00:38:17.520 --> 00:38:19.277
society that they're operating in? What

00:38:19.360 --> 00:38:22.397
kind of rules do they have? And I worry

00:38:22.480 --> 00:38:26.957
that we're losing the formula.
So the formula is pretty simple. The tech

00:38:27.040 --> 00:38:30.637
industry invents these things and assuming
that they're legal, right? They

00:38:30.720 --> 00:38:34.557
get launched and we develop global
platforms. When we start thinking of

00:38:34.640 --> 00:38:38.077
ourselves as a regional power and not
a global power, yeah, that's going to be

00:38:38.160 --> 00:38:42.157
problem. I want us to innovate and I
want us to build the future and I want

00:38:42.240 --> 00:38:45.357
to build it in America with our
democratic partners which by the way

00:38:45.440 --> 00:38:49.677
includes South Korea and Japan and so
forth. The US surpassed Britain as a

00:38:49.760 --> 00:38:52.077
world power in the 50s and 60s and in

00:38:52.160 --> 00:38:56.957
the 70s Britain had terrible economic
crises and they've come out of those.

00:38:57.040 --> 00:38:58.637
>> But I don't want that for us.

00:38:58.720 --> 00:39:00.877
>> Yeah. Could you just one sec before we

00:39:00.960 --> 00:39:03.357
move to AI tell me what a science

00:39:03.440 --> 00:39:06.637
fiction nightmare would be of a tick

00:39:06.720 --> 00:39:09.757
tock application? Like what would be an

00:39:09.840 --> 00:39:13.357
actual example of what happen? What tech

00:39:13.440 --> 00:39:15.837
people like me didn't understand and a

00:39:15.920 --> 00:39:20.717
lot of people did is that especially
when you take a young mind and you put a

00:39:20.800 --> 00:39:24.077
young mind in these incentive based
systems, they really do change their

00:39:24.160 --> 00:39:28.317
behavior, >> right? Yeah.
>> And I used to say 10 years ago, look,

00:39:28.400 --> 00:39:30.157
it's just the internet. Turn it off.

00:39:30.240 --> 00:39:34.397
Have dinner with your family. Go for a
walk. Relax. You can't do that anymore,

00:39:34.480 --> 00:39:38.637
right? We went from being optional
to being fundamental. People live on the

00:39:38.720 --> 00:39:41.117
internet and therefore the internet has

00:39:41.200 --> 00:39:43.117
to reflect the societal values that you

00:39:43.200 --> 00:39:45.757
want. So here's an example. China

00:39:45.840 --> 00:39:48.877
decides to insert censorship or

00:39:48.960 --> 00:39:52.717
promotion within the Tik Tok system that

00:39:52.800 --> 00:39:55.597
demotes or promotes stories that they

00:39:55.680 --> 00:39:59.517
like. Now would we notice this?

00:39:59.600 --> 00:40:01.597
>> Correct. And just
can I be very literal?

00:40:01.680 --> 00:40:05.117
So I'm a user.
I do a certain kind of story.

00:40:05.200 --> 00:40:08.877
Tik Tok promotes that story because it's
on a message for what China wants and

00:40:08.960 --> 00:40:12.317
you're now incentivizing you >>
and they use AI. Yeah. So if I were the

00:40:12.400 --> 00:40:15.197
Chinese government, I'm not trying to give
them advice, but what I would do is

00:40:15.280 --> 00:40:18.957
I would build an AI system that looked
for desirable behaviors and then I would

00:40:19.040 --> 00:40:22.637
shift the presentation to be consistent
with that desirable behavior. You would

00:40:22.720 --> 00:40:25.597
learn what the behaviors were. You
wouldn't know. This is the way computer

00:40:25.680 --> 00:40:29.277
scientists think. You look at a system,
you say, "What do I have to learn?" And

00:40:29.360 --> 00:40:31.437
then I want to optimize for more. Mhm.

00:40:31.520 --> 00:40:33.997
>> So in a business,
for your listeners, in

00:40:34.080 --> 00:40:36.237
a business, basically you want to get

00:40:36.320 --> 00:40:39.037
more customers. So you could learn what

00:40:39.120 --> 00:40:43.277
customers are actually doing as opposed
to what you think they're doing. Learn

00:40:43.360 --> 00:40:47.037
what they're doing and then whatever
they're doing, shift it to maximize

00:40:47.120 --> 00:40:48.717
revenue or whatever it is you're doing.

00:40:48.800 --> 00:40:50.317
That's how the tech companies work.

00:40:50.400 --> 00:40:53.437
>> Yeah. >> There's no reason to think
that governments couldn't require that the

00:40:53.520 --> 00:40:56.797
same, >> right? >> And by the way,
that would be a huge

00:40:56.880 --> 00:40:59.277
values violation in America.

00:40:59.360 --> 00:41:03.917
>> Yeah. Okay.
Your book, The Age of AI, our

00:41:04.000 --> 00:41:06.157
human future. Now, first of all, you've

00:41:06.240 --> 00:41:09.517
written it with Henry Kissinger and Dr.

00:41:09.600 --> 00:41:13.917
Hotton Locker. And why those two
as bedfellows in this book? Cuz if I'm

00:41:14.000 --> 00:41:18.157
right, Kissinger's 998
>> 98 >> uh 99 in May.

00:41:18.240 --> 00:41:22.397
>> Okay. >> And it's such a delight
to work with someone so brilliant.

00:41:22.480 --> 00:41:25.917
>> I can only imagine what he
would have been when he was 30.

00:41:26.000 --> 00:41:28.877
>> Right. So at 98,
he's such an incredible

00:41:28.960 --> 00:41:30.877
mind and thinker about things in

00:41:30.960 --> 00:41:33.357
society. What happened was I invited him

00:41:33.440 --> 00:41:36.637
to Google 12 years ago or so, and he

00:41:36.720 --> 00:41:40.557
said he'd come, but he wanted to give
a speech about how he thought Google was a

00:41:40.640 --> 00:41:44.877
threat to civilization. >> And I said,
"Okay, >> yeah, sounds great."

00:41:44.960 --> 00:41:47.037
>> And the Google employees love this. But

00:41:47.120 --> 00:41:49.757
what he basically said was that he did

00:41:49.840 --> 00:41:52.477
not want to have a single company having

00:41:52.560 --> 00:41:57.917
the kind of power that Google had
and now has on citizens of the world.

00:41:58.000 --> 00:42:00.557
>> Sure. >> Now of course
he would say that to other

00:42:00.640 --> 00:42:04.477
companies as well. Facebook being
an obvious candidate but maybe others but

00:42:04.560 --> 00:42:06.317
that built a friendship and a

00:42:06.400 --> 00:42:08.957
partnership. He came to a conference

00:42:09.040 --> 00:42:13.837
where Demis Hacabus was speaking and Demis
is the founder of deep mind and he

00:42:13.920 --> 00:42:17.917
was talking about the implications
of general intelligence and how when

00:42:18.000 --> 00:42:22.797
computers have general intelligence it
raises many many ethical questions and

00:42:22.880 --> 00:42:24.877
Dr. Kissinger had written a

00:42:24.960 --> 00:42:26.957
undergraduate thesis before we were all

00:42:27.040 --> 00:42:30.157
alive entitled the future of the world

00:42:30.240 --> 00:42:34.477
and in it he contemplated the questions
that are raised by Kant which have to do

00:42:34.560 --> 00:42:38.237
with what is the nature and definition
of an object. So he was very interested

00:42:38.320 --> 00:42:42.717
in the structure of how our mind
perceives facts in a world where you

00:42:42.800 --> 00:42:45.837
have in this case a single
company trying to shape it.

00:42:45.920 --> 00:42:49.277
>> Yeah. that then generalized to his view

00:42:49.360 --> 00:42:53.277
which he's published two articles in
the Atlantic on which goes something like

00:42:53.360 --> 00:42:57.997
this. There was a age of faith
before the age of reason. In faith you

00:42:58.080 --> 00:43:01.597
basically had your own knowledge which
you got from God or the king but you did

00:43:01.680 --> 00:43:05.997
not have critical thinking. In the age
of reason it was agreed that we would do

00:43:06.080 --> 00:43:09.917
better as humans if we had reason.
We would have a principle. We would have a

00:43:10.000 --> 00:43:13.597
debate. Things we take for granted today.
And that ushered in the age of

00:43:13.680 --> 00:43:17.677
enlightenment where people began to have
these philosophical questions and so

00:43:17.760 --> 00:43:21.517
forth. And the age of enlightenment then
leads to the industrial revolution and

00:43:21.600 --> 00:43:24.877
so forth and so on and everything we
have around us today. This is a really

00:43:24.960 --> 00:43:27.677
fundamental point. He argues that we're

00:43:27.760 --> 00:43:29.917
entering a new epic from the age of

00:43:30.000 --> 00:43:33.997
reason to the age of AI. And here's why.

00:43:34.080 --> 00:43:36.637
People will be changed by their

00:43:36.720 --> 00:43:41.517
interaction with these almost human
systems that are not human. We don't

00:43:41.600 --> 00:43:44.317
know how they'll be changed.
We don't know how their perception will be

00:43:44.400 --> 00:43:48.317
changed. But in the book, we try to ask
a bunch of questions. What will it be

00:43:48.400 --> 00:43:50.637
like to grow up as a child when your

00:43:50.720 --> 00:43:53.277
best friend is nonhuman? What will this

00:43:53.360 --> 00:43:56.797
do to war and conflict? What will this

00:43:56.880 --> 00:44:00.637
do to misinformation and people who
are trying to violate sort of the

00:44:00.720 --> 00:44:04.477
conventions of our society? But the core
question is, and it's why we wrote the

00:44:04.560 --> 00:44:06.957
book, is what does it mean to be human?

00:44:07.040 --> 00:44:08.237
>> Yeah. >> In this new world.

00:44:08.320 --> 00:44:11.437
>> Yeah. It's a good place to start.
is like what are we trying to even protect

00:44:11.520 --> 00:44:15.517
in varying degrees? You have real
fatalists about what AI is going to be.

00:44:15.600 --> 00:44:19.117
You have very utopian thinkers
and what it's going to be.

00:44:19.200 --> 00:44:21.517
>> But yeah, I think
a great question first

00:44:21.600 --> 00:44:25.357
is like what are the things about
us we would want to safeguard?

00:44:25.440 --> 00:44:29.997
>> Well, before we even answer the US
part, I'm not talking about killer robots,

00:44:30.080 --> 00:44:31.757
right? >> So when you
talk to people at AI, they

00:44:31.840 --> 00:44:33.677
say, "Yeah, I've seen that movie." Okay.

00:44:33.760 --> 00:44:35.117
Well, that's not what we're talking

00:44:35.200 --> 00:44:38.237
about. What we're talking about is the

00:44:38.320 --> 00:44:40.717
fact that AI will be around you helping

00:44:40.800 --> 00:44:45.357
you, guiding you, and maybe constraining
you in ways that you may or may not

00:44:45.440 --> 00:44:47.357
like. The obvious one is kids.

00:44:47.440 --> 00:44:50.157
>> Yeah. >> So, you've
got some toy, a bear. You

00:44:50.240 --> 00:44:54.797
give it to the kid at two. The bear gets
upgraded every year. The kid grows up.

00:44:54.880 --> 00:44:58.157
And at 12, the bear is watching TV with
the kid, his best friend, and the bear

00:44:58.240 --> 00:45:01.757
says, "I don't like this show." And
the kid says, "I don't like it either."

00:45:01.840 --> 00:45:03.437
>> How do you feel about that?
Do you think

00:45:03.520 --> 00:45:05.677
that your child's tastes should be

00:45:05.760 --> 00:45:09.917
influenced by his toy as opposed to his

00:45:10.000 --> 00:45:13.917
best male 12-year-old friend? In this
case, I'll give you another example.

00:45:14.000 --> 00:45:17.517
There's this huge controversy over what
people teach in schools. There always

00:45:17.600 --> 00:45:21.517
has been. There are textbooks and those
textbooks are approved by the state

00:45:21.600 --> 00:45:24.157
board of education and we can debate
whether they're right or not, but the

00:45:24.240 --> 00:45:27.197
important point is there's an approved
process for textbooks in schools. Well,

00:45:27.280 --> 00:45:29.197
now we've got this bear, right? And the

00:45:29.280 --> 00:45:31.677
bear, by the way, knows an awful lot of

00:45:31.760 --> 00:45:36.077
stuff. Okay, I just spit took coffee.

00:45:36.160 --> 00:45:39.677
>> You really? >> Oh my god,
that's my first spit take on the show.

00:45:39.760 --> 00:45:42.717
>> Yes, the bear knows way more than the
textbook. I'm sorry. That really cracked

00:45:42.800 --> 00:45:45.437
me up. Please continue.
>> By the way, the bear does.

00:45:45.520 --> 00:45:47.197
>> Yes. >> And let me
tell you how this happened.

00:45:47.280 --> 00:45:50.477
>> The total knowledge of human history,
>> right? The bear actually knows

00:45:50.560 --> 00:45:53.277
everything. Yeah. >> And the bear
is busy idling, waiting for

00:45:53.360 --> 00:45:56.957
the kid to ask him a question like,
"Tell me about Thusidities." [laughter]

00:45:57.040 --> 00:46:00.797
And the 12-year-old is busy watching UFC.

00:46:00.880 --> 00:46:04.237
What are the rules about how
the bear interacts with the kid?

00:46:04.320 --> 00:46:06.957
>> Because the kid is going
to learn more from the bear.

00:46:07.040 --> 00:46:10.077
>> By the way, we all learn more from
our peer group than we do from any other.

00:46:10.160 --> 00:46:13.517
>> Certainly more from sitting in class,
listening to the teacher, looking at the

00:46:13.600 --> 00:46:16.237
other kids. So now, let's imagine the

00:46:16.320 --> 00:46:19.517
bear learns something and it learns

00:46:19.600 --> 00:46:22.397
incorrectly that lollipops cure cancer

00:46:22.480 --> 00:46:26.557
in children.
>> And the bear is programmed by an

00:46:26.640 --> 00:46:30.557
objective function to keep the kid happy.
And happiness involves being

00:46:30.640 --> 00:46:35.437
healthy and alive. So the bear decides
to start suggesting the kid consume

00:46:35.520 --> 00:46:39.357
lollipops. >> The ability
to subtly manipulate a

00:46:39.440 --> 00:46:42.077
developing mind, it's a very big deal.

00:46:42.160 --> 00:46:45.517
>> Yeah. >> And the same
thing is true for adults

00:46:45.600 --> 00:46:49.197
that you see today in the Facebook
and Twitter examples where you have

00:46:49.280 --> 00:46:53.197
manipulation by people who know how
to get them excited. They get them excited

00:46:53.280 --> 00:46:57.437
by outrage. And YouTube, I mean,
like the rabbit hole, that whole thing is

00:46:57.520 --> 00:47:01.757
that it's just the slow >> tiniest
nanometer nudging in their direction.

00:47:01.840 --> 00:47:05.037
>> And in YouTube's case, they actually
made some changes that are important,

00:47:05.120 --> 00:47:10.237
right? >> They stopped recommending the
most emotional and crazy videos. They're

00:47:10.320 --> 00:47:13.517
still there, right? So they can say
they didn't censor them, but they don't

00:47:13.600 --> 00:47:16.877
recommend them. >> Yeah. >> And we've
seen situations over the years

00:47:16.960 --> 00:47:19.437
where people would start watching
something in YouTube and then they would

00:47:19.520 --> 00:47:22.717
get radicalized. >> Exactly.
>> Watching one more and more. This is a

00:47:22.800 --> 00:47:26.717
rabbit hole point. Yeah.
>> So that kind of stuff and I defy you to

00:47:26.800 --> 00:47:30.557
write a regulation, you know, sit down
with a pen and try to write how would

00:47:30.640 --> 00:47:33.517
the government regulate this kind
of behavior. It's not obvious to me. We

00:47:33.600 --> 00:47:36.477
need a consensus about that.
>> Well, so what I want to talk about in

00:47:36.560 --> 00:47:39.517
the best friend hypothetical is from the

00:47:39.600 --> 00:47:41.357
second I read that that was one of your

00:47:41.440 --> 00:47:44.477
points, I go to what does a child

00:47:44.560 --> 00:47:46.797
actually need? So I think that the

00:47:46.880 --> 00:47:49.917
product would obviously be designed to

00:47:50.000 --> 00:47:52.077
alleviate stress on the parents. They're

00:47:52.160 --> 00:47:57.117
the ones buying it, right? So ultimately
this AI bear the parents are going to

00:47:57.200 --> 00:47:59.837
decide to get that and they're going
to decide to get it because now the kids's

00:47:59.920 --> 00:48:03.197
more entertained. They have more free
time to theirelves. That would be the

00:48:03.280 --> 00:48:05.437
probably the incentive for the parent.

00:48:05.520 --> 00:48:10.077
>> Unfortunately everything great that
a human learns is through compromise, is

00:48:10.160 --> 00:48:12.717
through struggle, is through
disagreement, is through sharing when

00:48:12.800 --> 00:48:16.397
they don't want to share, is through
fighting. So, this device that will

00:48:16.480 --> 00:48:18.877
obviously be bought to help you as the

00:48:18.960 --> 00:48:21.837
parent have easier time in life is not

00:48:21.920 --> 00:48:26.317
going to present any of those things
that a human should have to endure. So,

00:48:26.400 --> 00:48:29.997
right out of the gates, I don't know how
the product, if it was an ethical one,

00:48:30.080 --> 00:48:32.877
would even be sold. So, I'm going to buy
this bear that's I'm going to hear them

00:48:32.960 --> 00:48:37.597
fighting upstairs in theory, right?
But how would you regulate this?

00:48:37.680 --> 00:48:39.597
>> Right? >> But the most
important thing is remember

00:48:39.680 --> 00:48:42.717
that this is technology today which is

00:48:42.800 --> 00:48:47.117
imprecise. It's dynamic. It's emergent
in the sense that when you combine it

00:48:47.200 --> 00:48:51.197
with others, it does stuff that's
unexpected. And it's learning while it's

00:48:51.280 --> 00:48:54.557
on the job. >> Right. >> So,
the problem here is you're the

00:48:54.640 --> 00:48:58.077
parent and you buy the, again, I'm using
the bear as a metaphor. You buy any form

00:48:58.160 --> 00:49:01.517
of digital toy. >> Yeah. >> Trying
to do the right thing for the kid

00:49:01.600 --> 00:49:05.597
and the toy is learning. What if it
learns the wrong thing? And sometimes it

00:49:05.680 --> 00:49:09.037
could learn the wrong thing because
it made a mistake. But it could also be

00:49:09.120 --> 00:49:11.517
misprogrammed by a government or a

00:49:11.600 --> 00:49:14.637
competitor or marketing or an incentive.

00:49:14.720 --> 00:49:17.917
So you have this crazy idea. You think
we should give these equivalent of these

00:49:18.000 --> 00:49:21.517
bears to every child. How do we support
that with advertising? Now the bear

00:49:21.600 --> 00:49:25.117
becomes advertising to the kid and the kid
basically then wants the toy and the

00:49:25.200 --> 00:49:27.517
parents have to buy the toy.
That seems like a terrible idea.

00:49:27.600 --> 00:49:30.877
>> You're so right because in the pursuit
of everything being democratized, you

00:49:30.960 --> 00:49:34.237
can't offer a product that's only
beneficial to the wealthy people who can

00:49:34.320 --> 00:49:37.357
afford this product. So then how else
do you provide it to everyone? Well,

00:49:37.440 --> 00:49:40.717
advertising. And this is how the internet
works. I feel like before this

00:49:40.800 --> 00:49:44.797
would even be permitted, we would all
have to collectively agree and social

00:49:44.880 --> 00:49:49.197
scientists would collectively agree
on what percentage of the interaction

00:49:49.280 --> 00:49:53.677
between a child and its best friend
should be conflict which percentage

00:49:53.760 --> 00:49:58.077
should be bonding in. So >> I I
can guarantee you there is no number

00:49:58.160 --> 00:50:01.357
that I can use. And if you can't tell
me the number, then the tech industry is

00:50:01.440 --> 00:50:04.157
going to do whatever it wants. That's
what I'm saying is like we can't even

00:50:04.240 --> 00:50:07.437
get consensus on our side where we might
want to say, well, here would be the

00:50:07.520 --> 00:50:09.117
guidelines of what this thing should do.

00:50:09.200 --> 00:50:13.597
>> Again, I think the issue of developing
young minds is going to get much more

00:50:13.680 --> 00:50:16.797
important and we've got to get ahead
of that. Let me give you another example.

00:50:16.880 --> 00:50:20.237
Does AI perceive things that are

00:50:20.320 --> 00:50:22.717
different from what humans perceive? And

00:50:22.800 --> 00:50:27.437
if so, what happens? So the example we
use in the book is a game called go and

00:50:27.520 --> 00:50:30.957
a game chess which we all know about
which have been played by humans for

00:50:31.040 --> 00:50:35.117
thousands of years not only won those
games against humans using these new AI

00:50:35.200 --> 00:50:38.077
techniques which is a major achievement
because it thought they were not

00:50:38.160 --> 00:50:42.877
possible but more importantly they
invented new strategies. Now how is it

00:50:42.960 --> 00:50:47.677
that in a game that's as mature as chess
as go that we could invent a completely

00:50:47.760 --> 00:50:51.277
new strategy for a game that's been
played for thousands of years? So, one

00:50:51.360 --> 00:50:54.637
possibility is that it was always
there for humans to discover. They just

00:50:54.720 --> 00:50:59.277
couldn't see it. >> Another
possibility is that the computer

00:50:59.360 --> 00:51:03.437
actually discovered a new truth
that we as humans can't really perceive.

00:51:03.520 --> 00:51:07.357
>> Oh my gosh. >> Now, this is a
thought experiment, but let's imagine

00:51:07.440 --> 00:51:09.757
>> that these systems
get smart enough that

00:51:09.840 --> 00:51:12.317
they begin to do things that are

00:51:12.400 --> 00:51:15.677
actually correct, but they can't explain

00:51:15.760 --> 00:51:19.597
what they're doing, and we don't
understand what they're doing. Now, Dr.

00:51:19.680 --> 00:51:21.517
Kissinger says if you look at history

00:51:21.600 --> 00:51:24.317
there are two things that happen either

00:51:24.400 --> 00:51:26.717
people rebel that is literally with guns

00:51:26.800 --> 00:51:29.757
against these overlords or they invent a

00:51:29.840 --> 00:51:33.837
new religion that there is a religious
component about faith because they can't

00:51:33.920 --> 00:51:38.077
understand what's actually going
on >> well and yeah the real historical

00:51:38.160 --> 00:51:42.237
examples of course you'd give is like
Galileo is virtually that he is the AI

00:51:42.320 --> 00:51:47.037
that believes in the copranicus way
of thinking of our solar system and then

00:51:47.120 --> 00:51:48.957
yes that person is it's too threatening

00:51:49.040 --> 00:51:51.597
we put them in jail. It also makes me

00:51:51.680 --> 00:51:54.077
think of I watched this documentary on

00:51:54.160 --> 00:51:56.877
physics and to learn and I didn't know

00:51:56.960 --> 00:51:59.197
this that all the theories that have

00:51:59.280 --> 00:52:01.437
been around for the last 60 years, those

00:52:01.520 --> 00:52:04.317
theories actually don't exist in

00:52:04.400 --> 00:52:06.237
language. They only exist in

00:52:06.320 --> 00:52:09.597
mathematics. The entire thought exists

00:52:09.680 --> 00:52:13.277
only in mathematics. And when they're
telling you about it, they're using this

00:52:13.360 --> 00:52:15.357
very substandard method to try to

00:52:15.440 --> 00:52:17.517
explain to you what's happening in the

00:52:17.600 --> 00:52:20.877
mathematics. So that was like for me

00:52:20.960 --> 00:52:24.557
kind of mindshattering like wow our
language can't account for what they're

00:52:24.640 --> 00:52:28.077
thinking of in this math.
>> Our language is not precise enough

00:52:28.160 --> 00:52:33.777
>> right? >> Stay tuned
for more armchair expert

00:52:33.860 --> 00:52:38.120
[music] if you dare.

00:52:38.240 --> 00:52:43.437
We are supported by HP. If you're
a parent or you know one like I do or you

00:52:43.520 --> 00:52:47.117
can help take care of kids, you already
know what August feels like. It's not

00:52:47.200 --> 00:52:52.157
summer anymore. It's the pregame,
the supplies list, the forms, the we need

00:52:52.240 --> 00:52:55.997
this printed by tomorrow texts that come
in at 9:00 p.m. And I say this as

00:52:56.080 --> 00:52:59.277
someone who's been on the receiving end
of those texts in a former life. You're

00:52:59.360 --> 00:53:03.197
ready to help and then the printer
says no, out of ink, blinking at you,

00:53:03.280 --> 00:53:07.837
useless. It's such a small thing, but it
will absolutely derail your night. So,

00:53:07.920 --> 00:53:12.157
when I heard about the HP Smart Tank
printer from HP, America's most trusted

00:53:12.240 --> 00:53:14.477
printer brand, I thought, "This is the

00:53:14.560 --> 00:53:16.317
kind of thing that actually makes life

00:53:16.400 --> 00:53:21.597
easier. Not in a big flashy way, in a
one less crisis way." Instead of those

00:53:21.680 --> 00:53:25.277
little cartridges that run out the second
you need them, the HP Smart Tank

00:53:25.360 --> 00:53:27.517
uses refillable ink tanks, and it comes

00:53:27.600 --> 00:53:30.157
with up to 3 years of ink included. 3

00:53:30.240 --> 00:53:35.277
years. That's three back to school
seasons without the cartridge panic. HP

00:53:35.360 --> 00:53:38.717
can't fill out the forms for you, but
they can make sure the printer is never

00:53:38.800 --> 00:53:40.557
the reason you're up past midnight.

00:53:40.640 --> 00:53:43.197
Search HP Smart Tank.

00:53:43.280 --> 00:53:47.757
>> This podcast is brought to you
by Squarespace. We've talked before about

00:53:47.840 --> 00:53:51.597
Rob building our website on Squarespace,
and I bring it up again because it's a

00:53:51.680 --> 00:53:53.357
perfect example of what they do well.

00:53:53.440 --> 00:53:57.037
Rob had all the pieces, the content,
the vision, the ideas, but he needed

00:53:57.120 --> 00:54:01.597
something to actually take all that and
make it public facing. And Squarespace

00:54:01.680 --> 00:54:04.877
was that bridge. Maybe you're
in the same spot. You've been developing

00:54:04.960 --> 00:54:08.717
something, a business, a skill set,
a body of work, and it's been private.

00:54:08.800 --> 00:54:11.837
It's been yours, but at some point,
you're ready to put it out there. And

00:54:11.920 --> 00:54:13.997
that transition from private to public

00:54:14.080 --> 00:54:16.557
can feel like a huge leap. Squarespace

00:54:16.640 --> 00:54:21.197
makes it feel like a step.
Beautiful templates, built-in SEO, email

00:54:21.280 --> 00:54:25.517
campaigns, scheduling, analytics,
everything you need to go from, "I've

00:54:25.600 --> 00:54:30.637
been working on this to here it is, come
find me." All in one place. No code, no

00:54:30.720 --> 00:54:34.477
stitching together a bunch of different
tools. So, if you've been sitting on

00:54:34.560 --> 00:54:37.757
something and waiting for the right
moment, this might be it. Head to

00:54:37.840 --> 00:54:41.757
squarespace.com/dax for a free trial.
And when you're ready

00:54:41.840 --> 00:54:46.717
to launch, use offer code DAX to save
10% off your first purchase of a website

00:54:46.800 --> 00:54:50.477
or domain. >> We are
supported by Arman Hammer

00:54:50.560 --> 00:54:54.397
Toothpaste. I feel like so many parts
of our lives have gotten a little over

00:54:54.480 --> 00:54:56.797
complicated, including toothpaste.

00:54:56.880 --> 00:55:00.077
There's always some new ingredient,
new promise, new thing you're supposed to

00:55:00.160 --> 00:55:04.477
suddenly care about. Armanhammer baking
soda toothpaste keeps it simple. Powered

00:55:04.560 --> 00:55:07.837
by baking soda, a time-t tested
ingredient you already know and trust

00:55:07.920 --> 00:55:12.157
for a clean you can actually feel. It
removes up to three times more plaque in

00:55:12.240 --> 00:55:15.357
hard-to-reach areas compared to a
leading toothpaste. And it gently

00:55:15.440 --> 00:55:19.437
whitens. And if you hear baking soda
and think, "Wait, is that gentle enough?"

00:55:19.520 --> 00:55:21.837
Arman Hammer is actually a less abrasive

00:55:21.920 --> 00:55:26.797
toothpaste and it is designed to protect
enamel. Myth busted. It's also the

00:55:26.880 --> 00:55:28.557
number one dentist recommended baking

00:55:28.640 --> 00:55:32.157
soda toothpaste. I am a broken record on

00:55:32.240 --> 00:55:33.917
Arm and Hammer toothpaste. I love it.

00:55:34.000 --> 00:55:36.877
I've been using it for years and years.

00:55:36.960 --> 00:55:41.917
My dad first introduced it to me and I've
never gone back. I've also gotten a

00:55:42.000 --> 00:55:45.997
lot of people on it and once you go
to Arman Hammer, you won't regret it. My

00:55:46.080 --> 00:55:49.997
personal favorite is the Arman Hammer
Peroxy Care toothpaste. It improves gum

00:55:50.080 --> 00:55:51.917
health while gently removing plaque and

00:55:52.000 --> 00:55:57.197
surface stains, but also just my mouth
feels so clean after. They have so many.

00:55:57.280 --> 00:56:00.397
They have advanced white toothpaste,
which is great. I also use that complete

00:56:00.480 --> 00:56:02.957
care toothpaste. Everything is perfect.

00:56:03.040 --> 00:56:06.717
Try Armanhammer toothpaste to taste the
difference. Feel the difference. Shop at

00:56:06.800 --> 00:56:10.957
your favorite retailer between August
2nd and August 15th to save $1.50.

00:56:11.040 --> 00:56:13.117
Armanhammer baking soda toothpaste. Save

00:56:13.200 --> 00:56:16.237
$1.50 August 2nd through 15th.

00:56:16.320 --> 00:56:18.557
>> This episode is brought to you by

00:56:18.640 --> 00:56:21.837
Service. Now look, I have my dream job.

00:56:21.920 --> 00:56:24.957
I get to talk to folks I admire like

00:56:25.040 --> 00:56:29.837
crazy and ask them virtually any
question that I want to. I can't imagine

00:56:29.920 --> 00:56:33.917
a better way to spend my time. But even
dream jobs have some not so dreamy

00:56:34.000 --> 00:56:36.477
parts. The stuff that gets in the way of

00:56:36.560 --> 00:56:41.597
the actual work. Now, that's where
Service Now's AI specialists come in.

00:56:41.680 --> 00:56:45.517
They don't just tell you what you should
do about your busy work. They actually

00:56:45.600 --> 00:56:47.917
do it start to finish. Cases closed,

00:56:48.000 --> 00:56:50.637
requests handled, no extra work for you.

00:56:50.720 --> 00:56:55.597
So you and your team can focus on what
matters most, which for me is, are they

00:56:55.680 --> 00:56:58.397
obsessed with male bodies
on the same level as I am?

00:56:58.480 --> 00:56:59.357
>> They never are.

00:56:59.440 --> 00:57:05.720
>> To learn how to put AI to work
for people, visit service.com.

00:57:13.920 --> 00:57:17.117
So one of the ways to understand how
this will happen, this is very relevant

00:57:17.200 --> 00:57:21.597
to this physics point is that today
there are things called large language

00:57:21.680 --> 00:57:25.277
models. And we talk about this in the
book. We use something called GPT3 as an

00:57:25.360 --> 00:57:29.597
example. And what they do is they suck
all the text that they can find in and

00:57:29.680 --> 00:57:33.837
they figure out what the text is about
and then using all sorts of tricks that

00:57:33.920 --> 00:57:37.197
are very powerful. It's very expensive
to compute. These things cost $und00

00:57:37.280 --> 00:57:39.197
million to to build one of them, for

00:57:39.280 --> 00:57:41.837
example. They appear to have humanlike

00:57:41.920 --> 00:57:44.957
knowledge with errors. You can ask it to

00:57:45.040 --> 00:57:49.997
compose something. You can ask it
to define itself. So, for example, on the

00:57:50.080 --> 00:57:53.997
back of our book, we have a quote from
GPT3. It was asked, "Are you human?" And

00:57:54.080 --> 00:57:58.317
it said, "No, I'm not human.
I am a language model. I don't have the

00:57:58.400 --> 00:58:00.557
reasoning capabilities that you as a

00:58:00.640 --> 00:58:05.597
reasoning machine have." Now, how did
it learn that? Right? It figured it out

00:58:05.680 --> 00:58:09.037
because of the descriptions that were
about what it was versus what humans

00:58:09.120 --> 00:58:11.997
were. Yeah. >> But it
appears to have insight.

00:58:12.080 --> 00:58:15.117
>> So many people believe
that the path goes like this.

00:58:15.200 --> 00:58:16.797
>> So the first thing
you do is you get the

00:58:16.880 --> 00:58:21.197
language model. Then you figure out what
is called dialogue where you can start

00:58:21.280 --> 00:58:24.317
asking questions.
>> I was playing with one which is not

00:58:24.400 --> 00:58:26.557
announced yet and I said tell me the

00:58:26.640 --> 00:58:29.517
product that's in 2001 of space odyssey

00:58:29.600 --> 00:58:33.917
that is available today here in the United
States. And the correct answer is

00:58:34.000 --> 00:58:36.637
a tablet. >> Okay?
>> And because it was the first

00:58:36.720 --> 00:58:39.037
demonstration of a tablet device. And so

00:58:39.120 --> 00:58:41.437
it comes back and it says there was a

00:58:41.520 --> 00:58:44.477
tablet which is the origin of that. And

00:58:44.560 --> 00:58:49.117
it used a different algorithm from the
kind of thing that Google search uses.

00:58:49.200 --> 00:58:53.037
That's an example of a breakthrough. And
then I could have asked it more and more

00:58:53.120 --> 00:58:56.637
questions about this tablet and the
origin and so forth and it would have

00:58:56.720 --> 00:59:00.477
understood the context.
>> That's the next step. All of these

00:59:00.560 --> 00:59:05.837
systems today have the problem that they
don't have their own volition. They

00:59:05.920 --> 00:59:08.317
can't generate their own
objective function. Right?

00:59:08.400 --> 00:59:11.757
>> So the sequence goes as follows.
The first phase is we get really good at

00:59:11.840 --> 00:59:15.917
somebody else's objective function.
The obvious one there for a physicist is

00:59:16.000 --> 00:59:19.837
they spend all day doing physics and then
it's 5:00 and it's time to go home

00:59:19.920 --> 00:59:24.397
and and so they say to the computer,
I want you to read every paper that's ever

00:59:24.480 --> 00:59:27.117
been written on ien values of this thing.

00:59:27.200 --> 00:59:31.197
>> Yeah. And tomorrow morning when I show
up to do physics, I want you to give me

00:59:31.280 --> 00:59:33.437
the best idea that I should pursue.

00:59:33.520 --> 00:59:37.997
>> Oh wow. >> It's the same assistant that
you would get because you would say, "I'm

00:59:38.080 --> 00:59:41.437
preparing for a show.
Tell me everything I should ask."

00:59:41.520 --> 00:59:43.597
>> Uhhuh. >> Right.
Which is today what a a good

00:59:43.680 --> 00:59:47.197
assistant does for you. But literally
read everything. Read everything that

00:59:47.280 --> 00:59:50.797
Eric said. Read all the books in this
area and read them all because I don't

00:59:50.880 --> 00:59:51.837
have time. >> Yeah.

00:59:51.920 --> 00:59:54.157
>> And tell me what
the key things are. But

00:59:54.240 --> 00:59:58.237
what is best? A computer doesn't know.

00:59:58.320 --> 01:00:02.797
Well, maybe they do, but what
we're qualifying as best idea.

01:00:02.880 --> 01:00:06.397
>> Well, again, it's gotten good enough
that it can get an approximation of

01:00:06.480 --> 01:00:11.277
best. So, that's the first I think
stable point. That's five years from

01:00:11.360 --> 01:00:13.597
now. >> Can I go really quick back to the

01:00:13.680 --> 01:00:16.397
valition statement you made? I want to

01:00:16.480 --> 01:00:18.717
see if if the way I'm understanding it

01:00:18.800 --> 01:00:21.437
is accurate. So I one time heard a

01:00:21.520 --> 01:00:24.317
debate with Steven Pinker on AI and he

01:00:24.400 --> 01:00:28.237
pointed out most of the the doomsdayers

01:00:28.320 --> 01:00:32.477
with AI. They're mapping on this truth

01:00:32.560 --> 01:00:35.197
about organic beings which is they have

01:00:35.280 --> 01:00:39.837
to survive. They have to reproduce.
They have to feed themselves. And the more

01:00:39.920 --> 01:00:41.837
they collect the the better their

01:00:41.920 --> 01:00:44.637
chances of survival are. that that very

01:00:44.720 --> 01:00:48.157
most basic thing of any organic organism

01:00:48.240 --> 01:00:53.277
is being mapped onto this AI and that's
not accurate. We would have to make it

01:00:53.360 --> 01:00:56.317
understand that.
>> So there's lots of evidence now that

01:00:56.400 --> 01:00:59.997
babies are born into the world
with preconceived notions of what their

01:01:00.080 --> 01:01:03.757
objective function would be more than
just food. So for example, there's

01:01:03.840 --> 01:01:07.757
something called zeroot learning where
you basically show somebody something

01:01:07.840 --> 01:01:12.957
without explaining it. and that when you
show something similar in the future, it

01:01:13.040 --> 01:01:14.877
looks familiar to them. >> Okay?

01:01:14.960 --> 01:01:18.237
>> Right? They can construct
where it is and what it's up to,

01:01:18.320 --> 01:01:22.237
>> right? And that's a hard problem
in in my field. But these are all problems

01:01:22.320 --> 01:01:26.077
that are being worked on. So,
at some point, we're going to have these

01:01:26.160 --> 01:01:29.597
incredible systems which will
do what somebody told them to do,

01:01:29.680 --> 01:01:32.237
>> but they still won't have volition. The

01:01:32.320 --> 01:01:37.037
great debate is at what point can
the computer begin to give itself

01:01:37.120 --> 01:01:41.357
objectives? Mhm. >> And we're
beginning to see this in the

01:01:41.440 --> 01:01:43.517
following. There are products that have

01:01:43.600 --> 01:01:48.957
just come out this year to help you
write code. And the idea is that the

01:01:49.040 --> 01:01:51.917
programmer is writing and the computer's
watching and it kind of finishes what

01:01:52.000 --> 01:01:54.557
you're writing. It's like finishing
your sentence as a programmer,

01:01:54.640 --> 01:01:56.397
>> right? Like autocorrect like

01:01:56.480 --> 01:01:58.797
>> autorect. But this is a very very

01:01:58.880 --> 01:02:03.837
powerful idea because as the programmers
use it, it will learn more and more

01:02:03.920 --> 01:02:07.757
about how to finish. Mhm.
>> So many people believe that at some

01:02:07.840 --> 01:02:11.197
point those systems you'll
be able to say write me this

01:02:11.280 --> 01:02:12.717
>> and it'll write the whole thing.

01:02:12.800 --> 01:02:15.517
>> Yeah. >> Okay.
Then the next question is when

01:02:15.600 --> 01:02:17.357
will it decide what it should write?

01:02:17.440 --> 01:02:20.077
>> Yeah. >> And that's
a speculative question. This

01:02:20.160 --> 01:02:24.637
is called artificial general intelligence.
There are skeptics and

01:02:24.720 --> 01:02:27.117
there are optimists.
I'm somewhere in between.

01:02:27.200 --> 01:02:31.357
>> Right. >> The optimists believe that
general intelligence which is not human

01:02:31.440 --> 01:02:35.277
intelligence but it's creative in the way
that humans are. it has volition. It

01:02:35.360 --> 01:02:37.917
can set its own objectives is 10 to 15

01:02:38.000 --> 01:02:40.397
years away. I'm a bit more conservative.

01:02:40.480 --> 01:02:43.997
I think it's more like 20 to 30. And there
are other people who say this will

01:02:44.080 --> 01:02:47.357
never happen. It's well beyond what
computers will ever do. Those kinds of

01:02:47.440 --> 01:02:48.797
people are usually wrong. >> Okay.

01:02:48.880 --> 01:02:50.557
>> Usually we'll get there somehow.

01:02:50.640 --> 01:02:54.077
>> Yeah. Really quick, they don't
have ego or could they ever have ego?

01:02:54.160 --> 01:02:55.357
>> Of course they could. >> They could.

01:02:55.440 --> 01:02:58.957
>> Look, the computer will say, "I want
to keep this human happy and these other

01:02:59.040 --> 01:03:01.597
people are a source of frustration
to them, so I'll just kill them all."

01:03:01.680 --> 01:03:04.077
>> Right. Right. >> Right.
That's clearly a bad objective

01:03:04.160 --> 01:03:07.437
function. Yeah. I'm not making light
of it, but I'm trying to say that's a

01:03:07.520 --> 01:03:10.877
mistake. >> Yeah. >> So,
you have to figure out a way to

01:03:10.960 --> 01:03:15.437
define in human society and in the
complexity we live in, what does looking

01:03:15.520 --> 01:03:17.037
better look like?

01:03:17.120 --> 01:03:19.357
>> And I'm wondering is computers may

01:03:19.440 --> 01:03:22.637
eventually observe humans and say, we

01:03:22.720 --> 01:03:26.397
know how to make your lives better
and actually understand it in a way that

01:03:26.480 --> 01:03:27.997
humans don't understand. >> Yeah.

01:03:28.080 --> 01:03:32.557
>> And if that happens, that's the most
utopian phase. The most dystopian phase

01:03:32.640 --> 01:03:34.717
is that we can't control these things.

01:03:34.800 --> 01:03:39.437
And furthermore, because of open source,
all this software is being leaked out to

01:03:39.520 --> 01:03:43.597
all sorts of evil people. You know,
next generation of evil terrorists and it

01:03:43.680 --> 01:03:45.517
could be really misused. And the obvious

01:03:45.600 --> 01:03:50.717
one is I say to the computer,
tell me how to kill a million people of a

01:03:50.800 --> 01:03:55.037
different race than my own. >> Uhhuh.
Right now, for this reason, in

01:03:55.120 --> 01:03:58.877
the book, and this is why I'm I'm
mentioning it, we talk at some length

01:03:58.960 --> 01:04:03.117
about the history of national
security and deterrence,

01:04:03.200 --> 01:04:05.757
we're going to have weapons that are

01:04:05.840 --> 01:04:10.637
like what I just described that
will be so dangerous that we'll

01:04:10.720 --> 01:04:15.117
need to have them be guarded and have
special access and so forth. And the

01:04:15.200 --> 01:04:18.477
reason we know that is we already have
such weapons in the form of nuclear

01:04:18.560 --> 01:04:22.637
weapons. Unlike nuclear where
the enriched uranium is difficult to get.

01:04:22.720 --> 01:04:26.477
The thought experiment is imagine if
nuclear stuff was free. Well, we'd all

01:04:26.560 --> 01:04:29.277
be dead. Trust me. >> Yeah.
>> Because the evil people would have sort

01:04:29.360 --> 01:04:31.197
of taken these taken this outcome.

01:04:31.280 --> 01:04:34.317
>> It only takes 0001%
of evil people to kill.

01:04:34.400 --> 01:04:38.477
>> Yeah. It only takes it only takes a few
for these mega scale weapons. And so we

01:04:38.560 --> 01:04:43.037
have to be very careful as we build
this technology to put in the necessary

01:04:43.120 --> 01:04:45.917
restrictions so that they're
not misused at scale.

01:04:46.000 --> 01:04:49.917
>> Just saying the machine can never kill
is not going to work cuz we're going to

01:04:50.000 --> 01:04:52.397
use it in military applications.

01:04:52.480 --> 01:04:56.317
>> Well, first place, who would you
say it to, right? So in programming the

01:04:56.400 --> 01:04:59.917
computer, you have to say rule one,
don't kill anyone. This is the Isaac

01:05:00.000 --> 01:05:03.437
Azimoff three laws of robots. You know,
don't harm anyone. But the fact of the

01:05:03.520 --> 01:05:07.677
matter is that in the military, they are
in the business of killing combatants,

01:05:07.760 --> 01:05:11.117
right? And having worked for the
military for 5 years, something called

01:05:11.200 --> 01:05:14.957
the Defense Innovation Board, they're
very straightforward about it, >> right?

01:05:15.040 --> 01:05:16.877
>> We are proud.
This is our mission. We're

01:05:16.960 --> 01:05:21.277
defending our country and if necessary,
under the laws of war, we will attack

01:05:21.360 --> 01:05:23.117
enemy commands and kill them. >> Yeah.

01:05:23.200 --> 01:05:25.277
>> And and they're not
apologetic about it.

01:05:25.360 --> 01:05:26.957
Right. >> So they're
clear of their mission and

01:05:27.040 --> 01:05:31.277
their principles are high. I I say this
with respect. So they're going to build

01:05:31.360 --> 01:05:34.557
systems which will allow them
to do that more accurately.

01:05:34.640 --> 01:05:37.037
>> Right? >> I would argue
that one of the things

01:05:37.120 --> 01:05:41.437
about current conflicts and people
don't want to hear this is that the vast

01:05:41.520 --> 01:05:45.597
majority of the deaths are
actually non-combatants, >> right?

01:05:45.680 --> 01:05:49.277
>> And that's terrible. >> Yeah.
Meaning like when we drone strike,

01:05:49.360 --> 01:05:52.637
most generally more civilians
go down than the target itself.

01:05:52.720 --> 01:05:56.157
>> The target identification problem
is a really hard problem. Mhm.

01:05:56.240 --> 01:05:59.437
>> So, let's imagine
that a good outcome of

01:05:59.520 --> 01:06:02.077
AI is the weapons are highly targetable.

01:06:02.160 --> 01:06:05.357
>> Mhm. >> But you still have the problem
of who gets to decide who to launch them.

01:06:05.440 --> 01:06:06.877
>> Yeah. >> So, we're
clearly going to have more

01:06:06.960 --> 01:06:10.077
accurate weapons. We say very clearly uh

01:06:10.160 --> 01:06:11.517
both in this commissioner report I did

01:06:11.600 --> 01:06:14.077
as well as in the book that the issue

01:06:14.160 --> 01:06:17.277
around computers deciding to do attacks

01:06:17.360 --> 01:06:19.757
is really problematic. And I'll give you

01:06:19.840 --> 01:06:22.877
an example. You're on a ship and there's

01:06:22.960 --> 01:06:25.757
a missile coming. The AI has figured it

01:06:25.840 --> 01:06:31.117
out, but you can't see it. It's not
on your radar, and it's coming fast. It's

01:06:31.200 --> 01:06:34.477
hypersonic. >> I think in the book you
say you've got 24 seconds, >> right?

01:06:34.560 --> 01:06:37.997
>> 24 seconds. Now, would you, Dax, would

01:06:38.080 --> 01:06:41.197
you press the button or would you count

01:06:41.280 --> 01:06:45.277
down? In the movies, they countdown.
Me, I just press the button.

01:06:45.360 --> 01:06:47.197
>> Right. Right.
>> You know, I'd panic. Yeah.

01:06:47.280 --> 01:06:48.877
>> That's that's who I am. >> Yeah.

01:06:48.960 --> 01:06:52.877
>> So, how do we deal with that? And and
one of the discussions that I think we

01:06:52.960 --> 01:06:56.717
should be having and we say this in the
book is we need to have some agreements

01:06:56.800 --> 01:06:59.197
about launch on warning that is this

01:06:59.280 --> 01:07:02.637
sort of any kind of automatic launch

01:07:02.720 --> 01:07:05.357
automatic attack decisions faster than

01:07:05.440 --> 01:07:08.237
human time is destabilizing because it

01:07:08.320 --> 01:07:12.717
leads the other side to think it has
to destroy it. It's too dangerous to allow

01:07:12.800 --> 01:07:15.117
this to exist. >> Yeah.
If you look right now, there's

01:07:15.200 --> 01:07:19.437
this lengthy conversation about whether
the enriched uranium in Iran should be

01:07:19.520 --> 01:07:21.597
destroyed ahead of its use. >> Yeah.

01:07:21.680 --> 01:07:24.317
>> Right. >> Yeah.
>> And there's all sorts of issues in

01:07:24.400 --> 01:07:27.837
Israel. They spend a fair amount of time,
they don't admit it, but I think

01:07:27.920 --> 01:07:31.677
it's fairly agreed to that whenever
something that could threaten Israel

01:07:31.760 --> 01:07:35.037
shows up in Syria, they use special
agents and they go and they destroy it

01:07:35.120 --> 01:07:37.917
and so forth. >> And then
we had the the great Seammens

01:07:38.000 --> 01:07:41.037
software we wrote to mess
with the center fuses.

01:07:41.120 --> 01:07:44.157
>> Yes. The allegation that Israel
and America did that. Yeah,

01:07:44.240 --> 01:07:47.597
>> we must start now talking specifically

01:07:47.680 --> 01:07:51.277
with China, but also with Russia
about appropriate restraints,

01:07:51.360 --> 01:07:54.157
>> kind of like a Geneva
convention, >> some agreement.

01:07:54.240 --> 01:07:56.957
>> Yeah. >> So,
when a country launches a missile,

01:07:57.040 --> 01:07:59.037
whether it's a missile test or a

01:07:59.120 --> 01:08:02.077
civilian missile such as SpaceX, all the

01:08:02.160 --> 01:08:06.717
other countries are notified.
>> And that's because no country wants the

01:08:06.800 --> 01:08:08.877
other one to think it's attacking him.

01:08:08.960 --> 01:08:13.517
>> Sure. and the other countries use
that as an opportunity to train their

01:08:13.600 --> 01:08:16.877
surveillance systems to watch the launch.

01:08:16.960 --> 01:08:19.677
>> So that's a good example where we have

01:08:19.760 --> 01:08:23.517
an agreement among untrusted parties

01:08:23.600 --> 01:08:27.757
that deescalates some
of the worst possible scenarios,

01:08:27.840 --> 01:08:30.877
>> right? >> So we've got
to have some rules. So, for

01:08:30.960 --> 01:08:33.917
example, when there is a cyber test in

01:08:34.000 --> 01:08:37.037
China and it goes ary, there better be a

01:08:37.120 --> 01:08:41.357
phone call from China to the United
States saying, "Sorry guys."

01:08:41.440 --> 01:08:44.877
>> Yeah. >> Not intended.
Don't come back after us.

01:08:44.960 --> 01:08:48.717
>> Yeah. But in light of Wuhan, it's
just like we just had an experience where

01:08:48.800 --> 01:08:51.677
they didn't do that with us.
>> That is correct. They didn't give us a

01:08:51.760 --> 01:08:55.517
heads up. Imagine if we, the country,
had gotten a phone call in late December

01:08:55.600 --> 01:08:59.357
saying, "Look guys, this thing's going
to get out of hand. We've been modeling

01:08:59.440 --> 01:09:03.997
it. get yourselves together
and collectively we can respond to this.

01:09:04.080 --> 01:09:06.157
That's an example of an emergent event.

01:09:06.240 --> 01:09:09.997
So you have AI which both speeds things
up. It also creates emergent events that

01:09:10.080 --> 01:09:11.357
we don't expect.

01:09:11.440 --> 01:09:15.517
>> Uhhuh. Holy smokes, man. The stakes are

01:09:15.600 --> 01:09:20.317
just so high. And and I agree with your
assessment that this will be another

01:09:20.400 --> 01:09:24.957
renaissance or enlightenment. This will
be some period that's looked back on.

01:09:25.040 --> 01:09:28.717
Should we make it? >> We'll make
it in some form. What I would

01:09:28.800 --> 01:09:31.437
say is that the opportunities for AI and

01:09:31.520 --> 01:09:34.557
the platforms that we're building are so

01:09:34.640 --> 01:09:37.597
enormously exciting. There is a drug

01:09:37.680 --> 01:09:41.677
called Halison and we profiled this in
the book. It was designed by some sin

01:09:41.760 --> 01:09:45.197
bio people at MIT and some computer
scientists. What they did is they

01:09:45.280 --> 01:09:48.397
trained the model on what they
knew about antibiotics. That was

01:09:48.480 --> 01:09:51.997
straightforward. Then they asked
the computer to look through a 100 million

01:09:52.080 --> 01:09:55.997
chemical compounds to figure out what
the compounds did, which nobody knew,

01:09:56.080 --> 01:09:59.437
and then figure out whether they would
produce an antibiotic response. It

01:09:59.520 --> 01:10:02.717
learned all of that on its own,
something humans could never do. And

01:10:02.800 --> 01:10:05.597
then it produced 10 candidates for drugs.

01:10:05.680 --> 01:10:08.077
>> Get out >> of antibiotics.
>> Oh my god,

01:10:08.160 --> 01:10:11.357
>> this is the best part. It then
took another network and looked at the

01:10:11.440 --> 01:10:14.317
characteristics of the existing
antibiotics and it said, "You have to be

01:10:14.400 --> 01:10:18.237
as different as possible."
And that produced one. Okay.

01:10:18.320 --> 01:10:20.717
>> Now, this one appears to produce a

01:10:20.800 --> 01:10:25.437
broad-scale antibiotic reaction
for people who have antibiotic resistance.

01:10:25.520 --> 01:10:27.517
>> Yeah. >> The number
of lives that are saved by

01:10:27.600 --> 01:10:30.477
that is enormous. Over and over and over

01:10:30.560 --> 01:10:32.557
again, Deep Mind and the University of

01:10:32.640 --> 01:10:35.437
Seattle Baker Lab just released the

01:10:35.520 --> 01:10:38.157
details of how proteins fold. Proteins

01:10:38.240 --> 01:10:42.957
fold in a particular order and the way
they fold and determines what they do.

01:10:43.040 --> 01:10:46.877
So, predicting protein folding has
been the holy grail. This is a discovery

01:10:46.960 --> 01:10:51.597
worthy of a Nobel Prize because based
on knowing how proteins fold then all of a

01:10:51.680 --> 01:10:55.277
sudden a whole another set of drugs and
chemicals and so forth can be made. It's

01:10:55.360 --> 01:10:59.837
similar to crisper in biology which
allowed us to do editing and so forth.

01:10:59.920 --> 01:11:02.397
So these are things
that humans could never do.

01:11:02.480 --> 01:11:06.477
>> Yeah. You posed the question in the
book, should a Nobel Prize come out of

01:11:06.560 --> 01:11:10.557
this? Does the AI receive it?
Do the people who ask the AI the question

01:11:10.640 --> 01:11:12.397
receive it? Like where are we at there?

01:11:12.480 --> 01:11:15.357
When Alph Go won, I was in Alph Go in

01:11:15.440 --> 01:11:17.997
Shanghai when we won against the human.

01:11:18.080 --> 01:11:20.717
This was 2017

01:11:20.800 --> 01:11:25.517
and the Chinese were so upset that in
the middle of the match they shut down

01:11:25.600 --> 01:11:28.957
the national television of the
match because they were losing

01:11:29.040 --> 01:11:31.677
>> because they all thought that they were
winning. What was interesting afterwards

01:11:31.760 --> 01:11:33.597
is I thought, what could I do to

01:11:33.680 --> 01:11:35.837
acknowledge this occasion? So, I took

01:11:35.920 --> 01:11:40.237
the 18 top people who worked on this and

01:11:40.320 --> 01:11:42.877
we all flew to the data center where

01:11:42.960 --> 01:11:47.917
this computer was and we found it
and we put a plaque on it to honor it.

01:11:48.000 --> 01:11:50.717
>> Uhhuh. >> And obviously
it's a humorous story, but

01:11:50.800 --> 01:11:52.557
we also spent a whole
day talking to each other.

01:11:52.640 --> 01:11:54.397
>> Yeah. >> Right.
About what had been achieved and

01:11:54.480 --> 01:11:57.917
so forth. And I'm very proud of that
visit, but who do I give the plaque to?

01:11:58.000 --> 01:12:00.797
It's a team of humans and it's a computer.
So, we put the plaque on the

01:12:00.880 --> 01:12:04.077
computer. Right. And we found it in
the middle of all the racks of computers

01:12:04.160 --> 01:12:07.997
that were that were doing work. And by
the way, it was three or four racks of

01:12:08.080 --> 01:12:10.077
computers in a huge data center. This

01:12:10.160 --> 01:12:13.037
shows you the power and also the peril

01:12:13.120 --> 01:12:15.517
of these things. That's the positive.

01:12:15.600 --> 01:12:19.917
The most likely dystopian case is a case
where we don't get this addiction

01:12:20.000 --> 01:12:23.997
problem under control. Um that the
computer systems are built around

01:12:24.080 --> 01:12:26.717
addiction and they drive us insane.

01:12:26.800 --> 01:12:31.357
Attention, attention, crisis, crisis,
issue, so forth and so on. Well, really,

01:12:31.440 --> 01:12:34.717
if I can interrupt you, what you said
at the beginning of this, which is in this

01:12:34.800 --> 01:12:39.517
even you allari future where we're
all just pleasure seekers in a virtual

01:12:39.600 --> 01:12:42.637
reality, >> everyone can
have status. Everyone can

01:12:42.720 --> 01:12:45.917
be the winner of it. It'll be [__]
It'll be fake status, but certainly the

01:12:46.000 --> 01:12:49.357
algorithm could make you feel like
you're chipping away and gaining all

01:12:49.440 --> 01:12:52.157
this popularity, but all
the popularity is a bunch of other bots.

01:12:52.240 --> 01:12:56.157
>> You can imagine all sorts of dystopian
scenarios. So you're popular in the

01:12:56.240 --> 01:13:00.397
metaverse of this rack of computers and
you're unpopular in the other metaverse.

01:13:00.480 --> 01:13:02.637
>> Yeah. >> And so forth.
I don't think we know how

01:13:02.720 --> 01:13:06.237
this will play out. But left to their
own devices, and this is why we wrote

01:13:06.320 --> 01:13:09.197
the book, what we say is that the
tech industry should not make these

01:13:09.280 --> 01:13:11.277
decisions. That these are decisions that

01:13:11.360 --> 01:13:14.477
should include philosophers, economists,

01:13:14.560 --> 01:13:16.397
behavioral scientists.

01:13:16.480 --> 01:13:19.517
>> Maybe you could
answer this question for

01:13:19.600 --> 01:13:22.877
us. the way our legal system works and

01:13:22.960 --> 01:13:25.197
the latigious nature of class action

01:13:25.280 --> 01:13:30.317
suits and whatnot to me sets up a catch

01:13:30.400 --> 01:13:34.637
22 that's standing in our way. So let's
say we design the system with this team

01:13:34.720 --> 01:13:39.517
of people and then we launch it and then
there's some disastrous outcome and we

01:13:39.600 --> 01:13:41.197
know that the disastrous outcome is

01:13:41.280 --> 01:13:44.717
going to result in x amount of legal

01:13:44.800 --> 01:13:47.517
fees that puts the company in an

01:13:47.600 --> 01:13:51.037
existential position and so it really

01:13:51.120 --> 01:13:53.597
creates a scenario where the company to

01:13:53.680 --> 01:13:57.277
acknowledge failure is also committing

01:13:57.360 --> 01:14:00.877
suicide. Yes, >> this has
to change if we're going to ask

01:14:00.960 --> 01:14:05.357
these companies, I think, to be real
time ethical, real time changing, real

01:14:05.440 --> 01:14:08.317
time transparent, real time
acknowledging what's [__] up and where

01:14:08.400 --> 01:14:12.717
it went wrong and how we're going
to change it. It can't really do it in our

01:14:12.800 --> 01:14:15.117
current system.
Do you have any thoughts on that?

01:14:15.200 --> 01:14:17.517
>> The effect of what you're describing is

01:14:17.600 --> 01:14:20.077
that the legal system is another

01:14:20.160 --> 01:14:22.797
participant, complainer, customer within

01:14:22.880 --> 01:14:27.197
the corporate environment. So the way
it really works is you have the founders

01:14:27.280 --> 01:14:30.477
with great vision, you have the employees,
you have the customers, you

01:14:30.560 --> 01:14:34.637
have the regulators, and you have public
opinion and the press. And as a CEO, you

01:14:34.720 --> 01:14:36.957
have to deal with all of them. What's

01:14:37.040 --> 01:14:39.277
interesting now is this next generation

01:14:39.360 --> 01:14:42.797
of employees is extremely activist.

01:14:42.880 --> 01:14:47.597
>> Yeah. So using the example I was using
earlier about the bear, I would hope

01:14:47.680 --> 01:14:51.597
that the future company that's
programming the bear, the employees

01:14:51.680 --> 01:14:55.117
would say to the management
that crosses the line.

01:14:55.200 --> 01:14:57.757
>> That's not okay. Your profit- seeeking

01:14:57.840 --> 01:15:02.637
motive is overwhelming my likelihood
of being willing to work for you to do it

01:15:02.720 --> 01:15:05.117
because you're going to do
some damage to some kid.

01:15:05.200 --> 01:15:07.677
>> Yeah. >> So I think
that the good news about this

01:15:07.760 --> 01:15:11.837
is these issues are going to get debated.
The bad news is we don't have

01:15:11.920 --> 01:15:16.157
good government answers and regulatory
answers on how to handle this. And the

01:15:16.240 --> 01:15:18.077
model we've used so far has worked well.

01:15:18.160 --> 01:15:22.877
You know, launch it, see what happens
and react to the worst cases. If you

01:15:22.960 --> 01:15:27.277
look at the Facebook situation, they
knew these problems and they continued

01:15:27.360 --> 01:15:30.317
to do them. >> But do
you think they were at a

01:15:30.400 --> 01:15:32.397
crossroads where they said if we now

01:15:32.480 --> 01:15:34.637
change this, it in itself is an

01:15:34.720 --> 01:15:37.357
admission that it was wrong? There's

01:15:37.440 --> 01:15:41.757
always that concern because if you're
a CEO and you say to the audience to the

01:15:41.840 --> 01:15:45.917
shareholders, we've decided to lower our
profits because we want to change this,

01:15:46.000 --> 01:15:48.637
you will get sued by d
shareholder derivative,

01:15:48.720 --> 01:15:52.957
>> right? Both sides, >> even if
it's the right thing to do. So,

01:15:53.040 --> 01:15:56.237
it takes an awful lot
of Kevlar on your vest. Yeah.

01:15:56.320 --> 01:16:00.957
>> Um, in order to do these things,
the great companies and the great leaders

01:16:01.040 --> 01:16:04.237
will do it. A lot of people are not
going to do it. By the way, this is

01:16:04.320 --> 01:16:07.917
great. I really I'm loving this. Okay,

01:16:08.000 --> 01:16:11.837
so that was one question and then I do
think it would really be fun for us to

01:16:11.920 --> 01:16:15.677
hear >> what the potential
life of this show is

01:16:15.760 --> 01:16:19.757
in a world of AI.
>> The thought experiment I would do is I

01:16:19.840 --> 01:16:22.317
would get one of these language models

01:16:22.400 --> 01:16:26.477
and I would have it be a panelist and I

01:16:26.560 --> 01:16:28.637
would ask it questions and see what it

01:16:28.720 --> 01:16:33.837
answers and I would do that every year
so you can track the progress of your

01:16:33.920 --> 01:16:37.677
replacement. >> [laughter] >> Now
what you're going to discover is

01:16:37.760 --> 01:16:42.397
that if you ask it the right question,
it gives an incredibly interesting

01:16:42.480 --> 01:16:46.637
answer, but it still requires
your volition. It won't drive the

01:16:46.720 --> 01:16:49.917
conversation. >> And so you'll
sit there and say that is

01:16:50.000 --> 01:16:54.637
the most frustrating guest. I have
to ask it precisely the question and it

01:16:54.720 --> 01:16:58.397
gives me a great answer
and otherwise it just waits.

01:16:58.480 --> 01:17:00.957
>> Right. Well, I could use
a little bit of that by the way.

01:17:01.040 --> 01:17:04.957
>> Well, and again though, great answer,
right? That's all dependent on what

01:17:05.040 --> 01:17:09.837
we're deciding is great. Is a precise
answer great? Is an emotional answer

01:17:09.920 --> 01:17:11.437
great? >> A metaphorical answer.

01:17:11.520 --> 01:17:14.957
>> But but the systems will learn the
answer to that question because remember

01:17:15.040 --> 01:17:18.957
its objective function is you
to ask it more questions, >> right?

01:17:19.040 --> 01:17:21.437
>> So it will learn how to ask it will

01:17:21.520 --> 01:17:25.997
learn how to give you answers that are
just exciting enough to keep you there

01:17:26.080 --> 01:17:30.077
>> and just factual enough to be
interesting and just responsive enough.

01:17:30.160 --> 01:17:33.277
>> Yeah. >> So it'll learn.
So the key thing about

01:17:33.360 --> 01:17:37.757
this technology that took me a long
time to learn is that it's busy learning.

01:17:37.840 --> 01:17:40.237
>> Yeah. >> And it will learn.
So we were talking

01:17:40.320 --> 01:17:43.917
about it as a guest. It's learning
while it's a guest on your show.

01:17:44.000 --> 01:17:45.917
>> Sure. >> How to be a guest on our show.

01:17:46.000 --> 01:17:49.437
>> How to be on a guest on your show.
And if it gets enough other gigs, it will

01:17:49.520 --> 01:17:55.240
become a very good guest.
>> Now, this might be a uncou

01:17:56.080 --> 01:17:58.877
source of information. They did a

01:17:58.960 --> 01:18:03.917
follow-up interview with the man
who Rainman was based off of. So they

01:18:04.000 --> 01:18:06.957
interviewed him at the time and they
took him, I believe he lives in Salt

01:18:07.040 --> 01:18:10.157
Lake City. He could go to the library
and you could pull any book off the

01:18:10.240 --> 01:18:13.197
shelf and you could turn to a page and he
would tell you what's written on the

01:18:13.280 --> 01:18:16.397
page. It's an insane >>
a truly a truly brilliant man.

01:18:16.480 --> 01:18:18.957
>> Yes. Then they did a follow-up and

01:18:19.040 --> 01:18:21.197
Leslie Stall went back and talked to him

01:18:21.280 --> 01:18:24.237
20 years later and he was making eye

01:18:24.320 --> 01:18:28.717
contact with her. He touched her.
He interacted with her in a way that

01:18:28.800 --> 01:18:33.117
appeared to be a lot of growth.
And so she said to his father,

01:18:33.200 --> 01:18:35.277
>> "Wow, he's gotten so much more

01:18:35.360 --> 01:18:40.637
personable over the years. It seems like
he has grown in his autism." And his dad

01:18:40.720 --> 01:18:44.797
said, >> "Well, I don't
believe he has actually

01:18:44.880 --> 01:18:47.357
obtained the missing stuff, but he is a

01:18:47.440 --> 01:18:51.837
great learner, and over time, he's
watched people have success looking at

01:18:51.920 --> 01:18:55.997
each other in the eye, and they've had
success in hugging." So, it's weird.

01:18:56.080 --> 01:18:59.277
It's that little thing that could exist
even within a human that the machine

01:18:59.360 --> 01:19:02.797
itself would be. >> So let's
say the bear has movable eyes

01:19:02.880 --> 01:19:07.117
and the programmers have not had time to
work on the eyes but there's a mechanism

01:19:07.200 --> 01:19:11.677
to do it. You could imagine that the bear
which is optimized to keep the kid

01:19:11.760 --> 01:19:15.677
happy. The first thing it learns
is to look at the kid >> right

01:19:15.760 --> 01:19:17.997
>> because what happens when it's
not doesn't look at the kid the kid's

01:19:18.080 --> 01:19:21.757
screaming. When it looks at the kid
the kid screams less. These are babies. We

01:19:21.840 --> 01:19:25.837
intuitively understand this. >> Yeah.
But it's not difficult for a bear

01:19:25.920 --> 01:19:29.997
to learn that because its objective
function says get attention be

01:19:30.080 --> 01:19:32.637
>> minimize crying or whatever cry. And we

01:19:32.720 --> 01:19:35.437
know from baby studies that babies are

01:19:35.520 --> 01:19:37.757
very very optimized around both their

01:19:37.840 --> 01:19:40.797
mother's uh sounds but also face

01:19:40.880 --> 01:19:43.357
recognition in particular eye contact.

01:19:43.440 --> 01:19:46.797
That's why they look and you look
at them and you if you look at the great

01:19:46.880 --> 01:19:49.597
moms they first start by cooing and

01:19:49.680 --> 01:19:51.997
holding and looking at and singing to

01:19:52.080 --> 01:19:54.717
and so forth and that imprinting which

01:19:54.800 --> 01:19:58.077
is so deeply human well it'll work with

01:19:58.160 --> 01:20:02.397
computers too because they'll learn how
to do it. I think maybe a place to start

01:20:02.480 --> 01:20:06.477
with some of this is like not
programming it for happiness because

01:20:06.560 --> 01:20:11.037
also that's so philosophical happiness
in general. What is it? Is it temporary?

01:20:11.120 --> 01:20:13.197
Is it >> that's what
like we need Paul Bloom to

01:20:13.280 --> 01:20:17.357
say, "Oh, they need 32%
discomfort and suffering to be uh

01:20:17.440 --> 01:20:21.427
>> well, okay. So, so you're going to have
the hard ass bear and the nice bear

01:20:21.510 --> 01:20:23.597
[laughter] and depending on whether the

01:20:23.680 --> 01:20:28.717
that kid has behaved, the little boy
or girl gets the hard ass. Even worse, the

01:20:28.800 --> 01:20:31.437
bear is going to have to trip the child
occasionally. Kids got to learn how to

01:20:31.520 --> 01:20:33.677
get up. >> So, so, you know,
we can have fun with

01:20:33.760 --> 01:20:36.237
the analogy, but the reason to have this

01:20:36.320 --> 01:20:39.197
conversation now is that the bear is a

01:20:39.280 --> 01:20:43.837
metaphor for their iPad today and their
phone today and all the things that kids

01:20:43.920 --> 01:20:48.557
are playing with. We're playing with fire.
We're playing with enormous impact.

01:20:48.640 --> 01:20:50.477
>> We're entering a new
phase when we don't

01:20:50.560 --> 01:20:54.237
fully understand the phase we're exiting
right now. We we still don't really

01:20:54.320 --> 01:20:57.997
fully understand the impact
of just the last 15 years.

01:20:58.080 --> 01:20:59.677
>> And I'll tell you
that the reason I want

01:20:59.760 --> 01:21:04.317
the bear is because the bear will learn
how the kid learns and the bear will

01:21:04.400 --> 01:21:08.317
then become the greatest teacher that
the kid has ever had. All we have to do

01:21:08.400 --> 01:21:12.797
is keep it from learning the wrong thing.
But imagine if you have a bear

01:21:12.880 --> 01:21:16.317
which is steeped in history, steeped
in language, steeped in the local history,

01:21:16.400 --> 01:21:20.637
has all the textbooks and so forth and it
prepares that kid in whatever way the

01:21:20.720 --> 01:21:24.797
kid learns, right? So let's do some
counting. Oh, I'm bored with counting.

01:21:24.880 --> 01:21:28.397
Okay, let's do some language and the
language is counting. You can imagine

01:21:28.480 --> 01:21:31.517
this enormously powerful tool for human

01:21:31.600 --> 01:21:36.717
development. I believe that the future
of the world is completely determined by

01:21:36.800 --> 01:21:40.877
whether we can get the best human
potential out of everybody. If you go

01:21:40.960 --> 01:21:43.437
back in American history around 1900,

01:21:43.520 --> 01:21:46.717
they had this huge rural workforce. All

01:21:46.800 --> 01:21:51.277
the evidence is that they were extremely
uneducated. And the government created

01:21:51.360 --> 01:21:52.957
universal high school and land grant

01:21:53.040 --> 01:21:56.477
universities in roughly 1910.

01:21:56.560 --> 01:22:00.797
100 years ago, 110 years ago,
they created these huge universities which

01:22:00.880 --> 01:22:05.197
have been the source of the innovation
of the industrial economy of America,

01:22:05.280 --> 01:22:09.117
our national security, our culture and so
forth and so on. Everybody expects to

01:22:09.200 --> 01:22:13.117
go through high school and people hope
to go to college. That wasn't true a 100

01:22:13.200 --> 01:22:15.277
years ago. >> Yeah, that's wild.
That's a great point.

01:22:15.360 --> 01:22:17.677
>> So, let's redouble our efforts to get

01:22:17.760 --> 01:22:20.237
people to their highest human potential

01:22:20.320 --> 01:22:24.957
using these tools and then they will >>
take over our world and they'll do much

01:22:25.040 --> 01:22:27.677
better than we have because they're
operating at their highest human

01:22:27.760 --> 01:22:31.037
potential. >> Okay.
My last question for you is a

01:22:31.120 --> 01:22:34.957
philosophical one. I was reading notes
from an underground and there's a whole

01:22:35.040 --> 01:22:37.117
chapter and I don't know what year that

01:22:37.200 --> 01:22:39.117
book was written but in the 1800s or

01:22:39.200 --> 01:22:43.597
turn of the century and doski spends I

01:22:43.680 --> 01:22:46.557
don't know a chapter theorizing that

01:22:46.640 --> 01:22:50.077
because mathematics have come so far

01:22:50.160 --> 01:22:53.437
that in the very near future through the

01:22:53.520 --> 01:22:58.477
power of mathematics mathematicians will
be able to predict all human behavior.

01:22:58.560 --> 01:23:00.317
And so when I read that I was like, "Oh

01:23:00.400 --> 01:23:03.997
my god." So this fear minimally has been

01:23:04.080 --> 01:23:07.917
around since he wrote that. And there's
really no justification at that moment

01:23:08.000 --> 01:23:10.637
for him to have that fear over
mathematics. That certainly did not

01:23:10.720 --> 01:23:13.597
prove to be the case. >> And I
thought to myself, wow, we have

01:23:13.680 --> 01:23:16.157
this new thing. We have AI. And so I

01:23:16.240 --> 01:23:19.437
wonder, is the fear human and we just

01:23:19.520 --> 01:23:23.277
keep planting it on each new thing? When

01:23:23.360 --> 01:23:26.237
humans are creative and curious in all

01:23:26.320 --> 01:23:28.557
the great ways, we also are

01:23:28.640 --> 01:23:30.317
unpredictable

01:23:30.400 --> 01:23:33.357
and AI will produce a new human-like

01:23:33.440 --> 01:23:36.477
intelligence that is also unpredictable

01:23:36.560 --> 01:23:40.877
that we're going to have to coexist with.
And the sooner we get ourselves

01:23:40.960 --> 01:23:42.957
organized around what that coexistence

01:23:43.040 --> 01:23:45.197
looks like, the better. This technology

01:23:45.280 --> 01:23:50.077
is coming much faster than
our philosophical, regulatory, government,

01:23:50.160 --> 01:23:54.557
public policy people are. >> Yeah,
>> it's time to get ahead of it.

01:23:54.640 --> 01:23:57.677
>> Wow. I'm so glad you took the time
to write this book. You know, it's it is

01:23:57.760 --> 01:24:00.077
funny when I thought about you being

01:24:00.160 --> 01:24:03.917
appointed the CEO of Google. I thought

01:24:04.000 --> 01:24:07.117
like who could be so impressive that

01:24:07.200 --> 01:24:09.437
those two, Larry and Sergey would say

01:24:09.520 --> 01:24:12.717
like, "Yeah, you run this company."

01:24:12.800 --> 01:24:16.397
That's almost unimaginable. And now that
I've chatted with you for an hour and a

01:24:16.480 --> 01:24:19.997
half, I'm like, "Fuck yeah. I would
have hired you immediately, too." What a

01:24:20.080 --> 01:24:23.037
comprehensive like
all-encompassing knowledge.

01:24:23.120 --> 01:24:26.237
>> You're just way too
flattering. So [__] true.

01:24:26.320 --> 01:24:30.317
>> Congratulations on you guys' success in
this. I look forward to working with you.

01:24:30.400 --> 01:24:33.517
>> Oh, thank you. Thank you so much. It
was such a pleasure to have you and good

01:24:33.600 --> 01:24:36.077
luck with the book and I think everyone
would be interested in reading it. And

01:24:36.160 --> 01:24:38.317
again, it's called The Age of AI, Our

01:24:38.400 --> 01:24:43.357
Human Future, written with Henry
Kissinger and Dr. Daniel Hutton locker.

01:24:43.440 --> 01:24:46.397
So, everyone check that out. Again,
thank you so much for coming. It was

01:24:46.480 --> 01:24:49.327
such a blast. >> Okay. Thanks, guys.

01:24:49.410 --> 01:24:52.077
[music] >> And now,
my favorite part of the show,

01:24:52.160 --> 01:24:57.320
the fact check with my soulmate
Monica Padman. [music]

01:25:00.400 --> 01:25:05.277
Being a serious business owner takes
hard work, patience, and a drive to keep

01:25:05.360 --> 01:25:08.957
moving forward. So, it's important
to find a serious business card that goes

01:25:09.040 --> 01:25:13.357
the extra mile. With a Venture X
business card from Capital 1, you earn

01:25:13.440 --> 01:25:14.957
unlimited double miles on every

01:25:15.040 --> 01:25:17.757
purchase. And with big purchasing power,

01:25:17.840 --> 01:25:20.477
you can spend more and earn more. This

01:25:20.560 --> 01:25:24.877
is your sign to take your business
to the next level. The Capital 1 Venture X

01:25:24.960 --> 01:25:27.437
business card. What's in your wallet?

01:25:27.520 --> 01:25:32.077
Terms apply. See capital.com for details.

01:25:32.160 --> 01:25:36.077
>> Hi. >> Hi. Okay, so fact check

01:25:36.160 --> 01:25:39.677
crosscontinental. You're on one side
of the divide. I'm on the other. That's

01:25:39.760 --> 01:25:42.397
right. I'm home for the posto holiday

01:25:42.480 --> 01:25:44.717
season. I'm home with the fam. They're

01:25:44.800 --> 01:25:49.037
recovered. The airport was just bonkers.

01:25:49.120 --> 01:25:53.357
I had >> on both sides. >> No,
only in LA. I've never experienced

01:25:53.440 --> 01:25:57.757
anything like it. I It was like
Disneyland. Like I was stuck standing

01:25:57.840 --> 01:26:01.757
like waiting to move, not even
to get to a line. It was awful.

01:26:01.840 --> 01:26:03.357
>> I almost missed my flight, which has

01:26:03.440 --> 01:26:06.237
never happened in my life. M

01:26:06.320 --> 01:26:09.037
>> and I was 2 hours early to the airport.

01:26:09.120 --> 01:26:12.637
>> That is so stressful. I'm going to have
to leave. I I'm going to have to leave

01:26:12.720 --> 01:26:14.397
at 4:00 in the morning, I guess, tomorrow.

01:26:14.480 --> 01:26:18.957
>> I know. Leave now. >> Oh my god. Maybe
I should just go right now. [laughter]

01:26:19.040 --> 01:26:22.317
>> I'll do it. Otherwise, how was your
travel? Do you feel like you contracted

01:26:22.400 --> 01:26:25.837
CO in transit? >> Well,
I don't know because there were so

01:26:25.920 --> 01:26:30.797
many people. >> Yeah.
Anyone coughing, sweating, dabbing

01:26:30.880 --> 01:26:34.957
their forehead with a handkerchief
or anything like from the 30s? Anytime

01:26:35.040 --> 01:26:39.367
anyone would cough, it's like like,
you know, you freak out a little bit.

01:26:39.450 --> 01:26:43.037
[laughter] >> Sure. Sure.
A little bit heightened.

01:26:43.120 --> 01:26:47.757
>> Yeah. >> But everyone looks healthy.
And did anyone lose any weight?

01:26:47.840 --> 01:26:52.397
>> I didn't notice any weight loss. I
guess there have been some lingering like

01:26:52.480 --> 01:26:55.757
>> cough and some stuff like that. Not
I mean, I don't think they're contagious.

01:26:55.840 --> 01:26:57.997
I think it's just going
to linger for a while.

01:26:58.080 --> 01:27:01.677
>> Any perspiration or any family
members dabbing their forehead with a old

01:27:01.760 --> 01:27:05.997
handkerchief out of their pocket?
Well, that's what they do anyway.

01:27:06.080 --> 01:27:11.197
>> That's normal. Yeah. >> After you have
your sun tea, you just dab that forehead

01:27:11.280 --> 01:27:14.717
>> and your black eyed peas. I made black
eyed peas on New Year's Eve and it got a

01:27:14.800 --> 01:27:19.437
little stressful because I had to see
Callie. I was hanging out with you guys

01:27:19.520 --> 01:27:23.917
and we watched Harry Potter and then I
had to see Callie and I also had to make

01:27:24.000 --> 01:27:27.757
black IPs. There was a lot to fit in.
I had to make sure I got those black IPs

01:27:27.840 --> 01:27:31.517
down by midnight. >> I know.
>> Which I did. Have you looked up the

01:27:31.600 --> 01:27:33.917
origin of this black eyed pea myth?

01:27:34.000 --> 01:27:37.357
>> It's a southern tradition. Like you're
supposed to make black eyed peas and

01:27:37.440 --> 01:27:40.637
collard greens. One is for wealth and

01:27:40.720 --> 01:27:43.037
one is like coins or something.

01:27:43.120 --> 01:27:47.437
>> Listen, I so wrongly assumed that
this was something that your grandparents

01:27:47.520 --> 01:27:49.677
passed down because it sounded like they

01:27:49.760 --> 01:27:54.637
were most faithful to that tradition,
but they just adopted it from the south.

01:27:54.720 --> 01:27:56.397
>> Yeah. >> Oh, wow. >> I know.

01:27:56.480 --> 01:27:59.937
>> That changes it.
>> They take it really seriously.

01:28:00.020 --> 01:28:03.117
[laughter] Now that I know it's
a southern tradition, I'm surprised that

01:28:03.200 --> 01:28:05.117
it's got its hooks in you as much.

01:28:05.200 --> 01:28:07.997
>> Well, it's a family
tradition at this point.

01:28:08.080 --> 01:28:09.757
>> Yeah, that's true. That's true.

01:28:09.840 --> 01:28:11.757
>> And I'm too scared to mess with it.

01:28:11.840 --> 01:28:14.157
>> If it was like hundred generations of

01:28:14.240 --> 01:28:17.757
your, you know, Padman's uh with blessed

01:28:17.840 --> 01:28:21.437
with great luck because of these black
eyed peas that that may you're like, I

01:28:21.520 --> 01:28:23.437
don't want to [__] this up. But like,

01:28:23.520 --> 01:28:26.397
>> but I'm third generation. Good luck.

01:28:26.480 --> 01:28:28.557
>> That's true.
That's true. Do you know if

01:28:28.640 --> 01:28:32.317
their luck was as good in India? I can't
imagine so. Or else they wouldn't have

01:28:32.400 --> 01:28:34.717
come here. >> Exactly.

01:28:34.800 --> 01:28:37.277
>> Okay. We deduced what happened and why

01:28:37.360 --> 01:28:39.357
we eat black eyed peas. We got it.

01:28:39.440 --> 01:28:42.077
>> That's right. And I, you know, I

01:28:42.160 --> 01:28:44.717
struggled because one of my resolutions

01:28:44.800 --> 01:28:49.837
is no food waste. But I made a lot
of black eyed peas on a I mean, not really

01:28:49.920 --> 01:28:53.197
on accident. It was like I had
to because I had the whole bag.

01:28:53.280 --> 01:28:56.557
>> But did I remember I
urged you to make five beans?

01:28:56.640 --> 01:28:57.917
>> I know. Oh, but then
I thought it happened

01:28:58.000 --> 01:29:01.197
>> like you were on your tiny
your tiny cooking show. [laughter]

01:29:01.280 --> 01:29:04.797
>> Tiny tiny kitchen cook off. Five beans.

01:29:04.880 --> 01:29:07.757
>> Yes. Five a five bean uh dish.

01:29:07.840 --> 01:29:11.517
>> Most people think that means five
different kinds of beans that are in it.

01:29:11.600 --> 01:29:16.710
And no, this is just five
total bean stew. [laughter]

01:29:18.480 --> 01:29:22.237
It's not like limema and Yeah. No, it's

01:29:22.320 --> 01:29:25.437
five black eyed peas. I'm going to make

01:29:25.520 --> 01:29:29.677
more beans this year in general because
my chef makes beans. She loves beans.

01:29:29.760 --> 01:29:31.517
And a lot of people I know who are into

01:29:31.600 --> 01:29:33.997
cooking. Make beans a lot. It's like a

01:29:34.080 --> 01:29:36.797
thing that shows you're a good chef.

01:29:36.880 --> 01:29:38.957
>> If you can make those tasteless little

01:29:39.040 --> 01:29:43.837
dollops of turd taste delicious.
Look, they're they're magic, man. Easy

01:29:43.920 --> 01:29:47.197
difference between Yeah.
baked beans down at uh

01:29:47.280 --> 01:29:50.957
>> uh down down in Austin there. Yeah.
You know, we had all those great barbecued

01:29:51.040 --> 01:29:54.237
baked beans >> and then doll.
That's so different.

01:29:54.320 --> 01:29:57.677
That's >> There's so many kinds.
Yeah. That's a mung bean.

01:29:57.760 --> 01:29:59.997
>> Uh-huh. So easy on your system.

01:30:00.080 --> 01:30:04.557
>> Yeah. Remember when you um were
on mung beans for like two weeks only?

01:30:04.640 --> 01:30:06.317
>> Yeah. It was crazy. >> Yeah.

01:30:06.400 --> 01:30:09.757
>> It wasn't the mung beans that bothered
me. It was that the recipe that to make

01:30:09.840 --> 01:30:12.797
the mung beans have some zest included

01:30:12.880 --> 01:30:16.797
celery. And I don't like celery.
So I really am associating the mung beans

01:30:16.880 --> 01:30:20.477
with the celery. And that's unfair
to the mung beans. But they are, you know,

01:30:20.560 --> 01:30:25.677
packed full of nutrients and protein
and just a they're really a super bean.

01:30:25.760 --> 01:30:29.277
>> You know what? I'm just going to travel
back for a second and realized you said

01:30:29.360 --> 01:30:32.717
doll. You know about doll? >> Yeah.

01:30:32.800 --> 01:30:39.240
>> How? >> Cuz I I'm
paying attention [laughter]

01:30:40.720 --> 01:30:44.717
>> to the world. >> Yes.
>> Oh my god. I didn't know you were

01:30:44.800 --> 01:30:48.957
listening so much. >> I try.
I give it my all. That's an

01:30:49.040 --> 01:30:52.397
Indian [__] >> I know. [laughter]

01:30:52.480 --> 01:30:54.717
>> Good job, Dax Shepard.

01:30:54.800 --> 01:30:58.077
>> Transformation is 3% complete.

01:30:58.160 --> 01:31:00.717
>> You're on your way.
>> [__] I can't wait.

01:31:00.800 --> 01:31:03.037
>> Tell us about your birthday.
It was yesterday.

01:31:03.120 --> 01:31:04.637
>> Okay. My birthday was yesterday. It was

01:31:04.720 --> 01:31:07.997
just fantastic. Fantastic. Went down to

01:31:08.080 --> 01:31:11.117
Charlie's gym and worked out with Matt

01:31:11.200 --> 01:31:13.757
and Ryan and Charlie and ate steak

01:31:13.840 --> 01:31:15.677
burritos.

01:31:15.760 --> 01:31:19.277
>> Oh, was it fun. What a perfect start to

01:31:19.360 --> 01:31:25.037
the day. And then I I leisurely drove
home and I took the Roadmaster, which I

01:31:25.120 --> 01:31:27.677
don't drive a lot. >> Birthday car.

01:31:27.760 --> 01:31:29.597
>> Birthday car. And I laid rubber

01:31:29.680 --> 01:31:31.277
everywhere I went cuz it was my

01:31:31.360 --> 01:31:36.157
birthday. And I got home and resumed

01:31:36.240 --> 01:31:38.557
Harry Potter. Started a second Harry

01:31:38.640 --> 01:31:42.237
Potter, played Spades, had Houston's

01:31:42.320 --> 01:31:46.157
went in the hot tub. Holy smokes. It

01:31:46.240 --> 01:31:49.597
just doesn't get better than that.
All my favorite things. Yeah, it was a

01:31:49.680 --> 01:31:52.237
really really pleasant
and relaxing birthday.

01:31:52.320 --> 01:31:55.997
>> I do think the armchairs should know
that now you're super into Harry Potter.

01:31:56.080 --> 01:32:00.877
Like last they have heard you barely know
anything about it and now you love it.

01:32:00.960 --> 01:32:04.717
>> I loved that first one and then we got
into the second one and there were some

01:32:04.800 --> 01:32:07.277
characters I was confused by. I don't

01:32:07.360 --> 01:32:11.197
know why there was a animated slug that

01:32:11.280 --> 01:32:15.677
was around him all the time who
was messing things up. He's supposed to

01:32:15.760 --> 01:32:19.117
bring him things.
>> Are you talking about Dobby?

01:32:19.200 --> 01:32:23.837
>> Yes. >> How dare you
call him a animated slug?

01:32:23.920 --> 01:32:27.677
That's way worse than
me calling Hufflepuffs.

01:32:27.760 --> 01:32:32.237
>> Well, listen. I I thought it veered
into the Jar Jar Banks a little bit.

01:32:32.320 --> 01:32:34.237
>> Like I I was like,
you have this amazing

01:32:34.320 --> 01:32:36.397
cast of characters humans that I love

01:32:36.480 --> 01:32:38.637
and I'm seeing these seven nine minute

01:32:38.720 --> 01:32:43.757
scenes with this animated blob and I
got a little frustrated with that.

01:32:43.840 --> 01:32:45.517
>> This is sacrile.
what you're doing right

01:32:45.600 --> 01:32:49.437
now is um he's like the most he's

01:32:49.520 --> 01:32:51.197
>> Is he really the most loved?

01:32:51.280 --> 01:32:54.877
>> Yeah, he's one of the top three
most loved characters in the series and

01:32:54.960 --> 01:32:57.277
you'll you'll learn why.
You just got to keep going.

01:32:57.360 --> 01:33:00.317
>> Okay. >> Yeah. >> Okay, great.
I will. It's interesting to

01:33:00.400 --> 01:33:03.677
imagine what Well, I was trying to do
it while we were watching it. I'm like,

01:33:03.760 --> 01:33:05.437
you're very Hermione. You must have felt

01:33:05.520 --> 01:33:07.277
some kinship or that must have been an

01:33:07.360 --> 01:33:10.797
easy protagonist to like relate to in a

01:33:10.880 --> 01:33:15.757
book. I mean, parts for sure,
but it's hard because she's a kid in those

01:33:15.840 --> 01:33:19.677
movies. And when I was a kid, I was not
like that. Like, I would never raise my

01:33:19.760 --> 01:33:22.237
hand. I would never want to be called on

01:33:22.320 --> 01:33:24.957
in class. I was not a know-it-all. I was

01:33:25.040 --> 01:33:27.757
like really trying to blend.

01:33:27.840 --> 01:33:32.477
>> Oh, that Yeah. And I guess that wasn't
the aspect of her that made me think of

01:33:32.560 --> 01:33:34.717
you. It was more that there was this

01:33:34.800 --> 01:33:38.557
determined little gal who through

01:33:38.640 --> 01:33:42.877
overachieving was going to get
to where she wanted to be.

01:33:42.960 --> 01:33:46.557
>> Oh, that's >> in the club.
In the club that she wanted to be in.

01:33:46.640 --> 01:33:49.117
>> Oh, I see.

01:33:49.200 --> 01:33:53.757
Yeah, I am her. >> There we go. [laughter]

01:33:53.840 --> 01:33:56.717
Reminding, you know, again, this goes
right to the problem we always deal

01:33:56.800 --> 01:33:59.917
with. Slim pickings for females. I you
know when you're watching with all the

01:34:00.000 --> 01:34:03.677
guys like you know we got six or seven
male leads we could choose from. You got

01:34:03.760 --> 01:34:06.877
the old ass teacher who's a cat

01:34:06.960 --> 01:34:10.957
>> and then you got Hermione [laughter]
to my knowledge. Is there another female

01:34:11.040 --> 01:34:15.597
in the franchise? >> Yeah there
are and more and more pop up

01:34:15.680 --> 01:34:17.837
but yeah none that I feel is super major

01:34:17.920 --> 01:34:20.957
kinship to there is an Indian character.

01:34:21.040 --> 01:34:24.797
She's superfluous and >> oh expendable.

01:34:24.880 --> 01:34:29.437
>> Yeah. Did she get blasted
by some magic at some point?

01:34:29.520 --> 01:34:33.757
>> Not that I I don't remember actually.
I don't remember if she perishes or if she

01:34:33.840 --> 01:34:35.757
thrives. I just don't remember.

01:34:35.840 --> 01:34:39.597
>> Okay. All right then.
>> Um but I'm really happy that you're

01:34:39.680 --> 01:34:43.437
getting on board so you
can relate to me more.

01:34:43.520 --> 01:34:47.597
>> Yes. Yes. I love it. Well, and then
on that same topic, when um we both return

01:34:47.680 --> 01:34:51.437
from our Sojourns, we're
going to season two of Cheer.

01:34:51.520 --> 01:34:54.077
>> I can't wait.

01:34:54.160 --> 01:34:59.197
>> I cannot wait. I started it again
last night so I could like get pumped up.

01:34:59.280 --> 01:35:01.677
>> Yeah. >> And I I you know,
talk about characters

01:35:01.760 --> 01:35:06.877
I feel a kinship to. >> My goodness.
I just know that world so

01:35:06.960 --> 01:35:09.917
deeply and I miss it. >> Boy,
I wonder if I could, you know, cuz

01:35:10.000 --> 01:35:11.517
I do I'm going to bring the rig with the

01:35:11.600 --> 01:35:14.157
batteries. It is feasible that we could

01:35:14.240 --> 01:35:17.677
do a fact check with Aaron and I while

01:35:17.760 --> 01:35:19.917
we're in the ocean.

01:35:20.000 --> 01:35:22.877
>> Okay. I All right. That scares me a

01:35:22.960 --> 01:35:25.437
little bit. Of course, cuz I'm Hermione,

01:35:25.520 --> 01:35:30.237
>> naturally. >> I What if
you drop it and we ruin the

01:35:30.320 --> 01:35:34.637
whole thing? >> Yeah, I risk.
That's why I mean, look,

01:35:34.720 --> 01:35:38.797
if there weren't some variables that
could end in disaster, it wouldn't make

01:35:38.880 --> 01:35:41.437
for a good experiment.
That's [clears throat] why we would do it.

01:35:41.520 --> 01:35:45.757
>> All right. Me and Aaron doing
a fact check from the bed. Okay.

01:35:45.840 --> 01:35:50.557
>> Okay. >> We've seen that. Yeah. BDDT.

01:35:50.640 --> 01:35:52.397
>> This is kind of a ding- ding ding. High

01:35:52.480 --> 01:35:57.997
risk, high reward. Cuz we're
talking about Eric Schmidt. Googy.

01:35:58.080 --> 01:36:01.437
>> Oh my gosh. >> Googy Hower.

01:36:01.520 --> 01:36:03.597
>> Douglas Hower. And then ding ding ding.

01:36:03.680 --> 01:36:06.157
Doogie is the name of that little slug.

01:36:06.240 --> 01:36:10.557
>> No. [laughter] >> What is it? Soggy.

01:36:10.640 --> 01:36:14.477
>> Dobby. >> Dobby. Close. Dobby. Hower.

01:36:14.560 --> 01:36:15.597
>> You called him.

01:36:15.680 --> 01:36:18.157
>> Soggy. If I was with you, I would have

01:36:18.240 --> 01:36:21.997
to punch you cuz he's so lovable and I

01:36:22.080 --> 01:36:26.717
I'd have to stick up for him. >> Okay.
All right. Well, to me, it seemed

01:36:26.800 --> 01:36:30.317
like a like having
a dog that talks nonstop.

01:36:30.400 --> 01:36:34.797
Like, well, this is part of being
a wizard. I got to have this dog. Man,

01:36:34.880 --> 01:36:36.737
this is a bad part of being a wizard.

01:36:36.820 --> 01:36:39.037
[laughter] I'm always
talking to this thing. He

01:36:39.120 --> 01:36:41.037
never understands anything. I don't know.

01:36:41.120 --> 01:36:44.637
>> You got to continue. I think I just I
got I got commitment anxiety while I was

01:36:44.720 --> 01:36:47.037
watching the relationship.

01:36:47.120 --> 01:36:49.437
>> Okay. Um Yes, Eric Schmidt. I just was

01:36:49.520 --> 01:36:54.157
really I really really was blown
away with the the totality, the

01:36:54.240 --> 01:36:57.517
comprehensiveness of the
knowledge was was intense.

01:36:57.600 --> 01:36:59.837
>> Yeah. There are just
so few humans like him.

01:36:59.920 --> 01:37:01.837
>> Yeah. What are you looking at?

01:37:01.920 --> 01:37:03.277
>> Well, I have to put my back.

01:37:03.360 --> 01:37:06.477
>> Space bar. Oh, on the desktop. On the

01:37:06.560 --> 01:37:08.157
desktop. On the IBM.

01:37:08.240 --> 01:37:14.317
>> This is a fact check after all.
No, this is a MacBook Pro 1460.

01:37:14.400 --> 01:37:17.117
>> Hey, >> well,
I made those numbers up, but

01:37:17.200 --> 01:37:21.517
>> Hey. Hey, Padman's. [laughter] >> No,
this is my laptop. What are you

01:37:21.600 --> 01:37:24.717
talking about? >> Oh my god.
I thought you were in a

01:37:24.800 --> 01:37:30.637
family like office storage room where
your dad does CAD drafting on an IBM.

01:37:30.720 --> 01:37:32.397
>> No, >> but now I understand.

01:37:32.480 --> 01:37:35.917
>> I am in the storage closet at my

01:37:36.000 --> 01:37:38.397
5-year-old desk that just sits here. My

01:37:38.480 --> 01:37:41.197
dad's office is in another room.

01:37:41.280 --> 01:37:44.877
>> Okay. Okay. All right.
>> And I've been here for 5 minutes and I

01:37:44.960 --> 01:37:47.277
already got a piece of jewelry.

01:37:47.360 --> 01:37:50.237
>> Oh, was that one of your presents?
Did you guys exchange presies?

01:37:50.320 --> 01:37:54.717
>> Oh gosh, we did exchange presies in the
worst way. They're just boxes, you know,

01:37:54.800 --> 01:37:58.717
just like here's some >> Oh,
was your mom using her decorative bags?

01:37:58.800 --> 01:38:02.157
>> And she didn't even do that this year
cuz it's post Christmas. It's a mess,

01:38:02.240 --> 01:38:06.797
you know. She just gave me the boxes
as boxes like that were shipped. It also

01:38:06.880 --> 01:38:10.237
may have gotten back to her that you're
disapproving of the bag. So, she's like,

01:38:10.320 --> 01:38:14.957
"Oh, you want to see no effort,
[__] Here's your [__] [laughter]

01:38:15.040 --> 01:38:17.997
It's in the trunk with the receipt
still in the shopping bag."

01:38:18.080 --> 01:38:20.397
>> I know. No,
she did really good. She She

01:38:20.480 --> 01:38:25.197
gave me a good surprise gift, which
was my amber oil, which is my perfume of

01:38:25.280 --> 01:38:28.317
choice. >> Oh, she's
paying attention, too.

01:38:28.400 --> 01:38:31.517
>> She is. And no,
this ring we found in my

01:38:31.600 --> 01:38:34.637
old jewelry box. It was my mom. She gave

01:38:34.720 --> 01:38:38.957
it to me, I guess, when I was young.
And then I put it in an old jewelry box,

01:38:39.040 --> 01:38:41.357
never to be seen again till this day.

01:38:41.440 --> 01:38:44.237
>> Oh my god. Reunited.
And it feels so good.

01:38:44.320 --> 01:38:47.437
>> So good. >> You're too
young to remember these

01:38:47.520 --> 01:38:52.077
commercials. In fact, I bet they stopped
airing before you were even alive. But

01:38:52.160 --> 01:38:53.357
first of all, do you know there was a

01:38:53.440 --> 01:38:57.677
hit song that went reunited

01:38:57.760 --> 01:39:00.317
and it feels so good.

01:39:00.400 --> 01:39:04.717
>> Of course I do. >> Okay,
great. So, so there was a wine

01:39:04.800 --> 01:39:07.437
beverage called Riani. >> Uh oh.

01:39:07.520 --> 01:39:10.317
>> Yeah. Yeah. And well, I'll give you one

01:39:10.400 --> 01:39:12.477
guess at what the commercial was.

01:39:12.560 --> 01:39:17.277
>> Runi and it tastes so good.

01:39:17.360 --> 01:39:20.077
>> Oh, man. You're so close. I Well, it

01:39:20.160 --> 01:39:25.000
might have said [laughter] it
tastes so good. It definitely said

01:39:25.520 --> 01:39:30.307
whatever. And then they just Yeah,
that was their commercial. Reunit.

01:39:30.390 --> 01:39:33.437
[laughter] It feels so good.
I think it feels so good.

01:39:33.520 --> 01:39:35.357
>> Oh my god. >> I think they
were letting you know like

01:39:35.440 --> 01:39:37.277
a buzz is coming. Like [__] the taste.

01:39:37.360 --> 01:39:39.357
You're not buying Runi for the taste.

01:39:39.440 --> 01:39:43.997
You're buying it for the cheap buzz.
I think. Again, don't sue me Runi. I don't

01:39:44.080 --> 01:39:45.437
even know if that's a real company.

01:39:45.520 --> 01:39:47.837
Doesn't sound real. >> Let me look it up.

01:39:47.920 --> 01:39:53.887
>> Okay. >> Wait, is it R I U N I T E?

01:39:53.970 --> 01:39:57.117
[laughter] >> So classy. >> It's a brand

01:39:57.200 --> 01:40:01.097
is still in existence.
>> Okay, let's see. Ryuniti

01:40:01.180 --> 01:40:06.077
[laughter] is a brand
of Italian wine imported and

01:40:06.160 --> 01:40:09.757
sold in a variety of flavors in the
United States by Frederick Wildman and

01:40:09.840 --> 01:40:11.997
Sons of New York City, New York. The

01:40:12.080 --> 01:40:14.157
brand is known for advertisements of the

01:40:14.240 --> 01:40:18.077
1970s and 80s depicting a broad array of

01:40:18.160 --> 01:40:21.277
social situations and using the tagline

01:40:21.360 --> 01:40:25.197
>> uh oh Jax using the tagline reuny on

01:40:25.280 --> 01:40:28.557
ice. That's nice.

01:40:28.640 --> 01:40:30.397
We [laughter] Okay. Okay. Can you spell

01:40:30.480 --> 01:40:32.397
it for me? [gasps]

01:40:32.480 --> 01:40:38.280
>> R I U N I T E.

01:40:43.100 --> 01:40:45.120
[music]

01:40:56.850 --> 01:40:59.837
>> [music] >> I guess that's
not what I was saying, is it?

01:40:59.920 --> 01:41:01.757
>> Not at all. >> Or was it?

01:41:01.840 --> 01:41:04.957
>> Well, at first it sounded like you were

01:41:05.040 --> 01:41:07.757
right and then as it went on, no,

01:41:07.840 --> 01:41:10.317
>> it sounded worse. Yeah, but that was a

01:41:10.400 --> 01:41:16.920
like a new one. >> Oh [__]

01:41:20.640 --> 01:41:25.117
Okay. Well, CBD, if I can find one where
they did that or if I made the whole

01:41:25.200 --> 01:41:28.477
thing up in my head or if they just
slept on an amazing idea to do a

01:41:28.560 --> 01:41:31.917
commercial where they said Reuny.

01:41:32.000 --> 01:41:35.517
Maybe just when I was a kid, I would
replace Reunited with Reuniti. I don't

01:41:35.600 --> 01:41:38.877
know. >> Wow, we opened
up more questions than we

01:41:38.960 --> 01:41:43.197
answered. >> What a what a little check.

01:41:43.280 --> 01:41:45.837
>> Okay, Eric Schmidt. So, we have to talk

01:41:45.920 --> 01:41:51.357
about something important. This whole
episode is basically about a teddy bear

01:41:51.440 --> 01:41:54.717
going to know your >> I know
where you're going with this.

01:41:54.800 --> 01:41:57.677
>> That's going to know your child so well

01:41:57.760 --> 01:42:01.117
and be its best friend. And what are the

01:42:01.200 --> 01:42:04.077
implications? Then a month later after

01:42:04.160 --> 01:42:08.797
we record this, I walk into the house
on Delta, your child's birthday. Seven

01:42:08.880 --> 01:42:10.957
years old. >> Seven years young.

01:42:11.040 --> 01:42:14.477
>> Seven years young.
And her gift from you

01:42:14.560 --> 01:42:18.237
guys is a robot. And it's this like

01:42:18.320 --> 01:42:23.437
small cute robot like you know
approximately teddy bear size I would

01:42:23.520 --> 01:42:27.997
say. >> Yeah. I bet he's 18 in tall or 20.

01:42:28.080 --> 01:42:29.597
>> He has a name that

01:42:29.680 --> 01:42:31.197
>> Yes, he comes with a name.

01:42:31.280 --> 01:42:35.117
>> This robot gave me
and Jess a crazy side

01:42:35.200 --> 01:42:40.077
eye >> and we knew
immediately uhoh, there's

01:42:40.160 --> 01:42:41.597
trouble here.

01:42:41.680 --> 01:42:44.637
Let's be clear. So, from the second this

01:42:44.720 --> 01:42:47.117
unnamed friend came out of the box, you

01:42:47.200 --> 01:42:48.877
were panicked. You're like, that was the

01:42:48.960 --> 01:42:53.117
episode. I w obviously if I if I notice

01:42:53.200 --> 01:42:55.757
that there's an issue with this, I will

01:42:55.840 --> 01:42:58.877
remove it from the house. But thus far

01:42:58.960 --> 01:43:01.357
and why I wanted you to keep the product

01:43:01.440 --> 01:43:04.957
name quiet is it turned on. It learned

01:43:05.040 --> 01:43:06.717
her name. It said a couple of weird

01:43:06.800 --> 01:43:10.477
things. It sideed you. You panicked. You

01:43:10.560 --> 01:43:15.117
left the house in a huff [laughter]
and then and then the friend shut down

01:43:15.200 --> 01:43:19.517
that night and the friend has yet
to turn back on. It there's some Yeah.

01:43:19.600 --> 01:43:21.677
Okay. >> There's a glitch.
That doesn't mean the

01:43:21.760 --> 01:43:25.197
friend isn't listening even
though it's pretending it's dead.

01:43:25.280 --> 01:43:28.797
>> Really good point. So this is what
happened. So the friend comes out. It

01:43:28.880 --> 01:43:31.117
asks Delta some questions. It is it's

01:43:31.200 --> 01:43:36.397
learning about her. It's like would
you rather ride on a dragon or a flying

01:43:36.480 --> 01:43:38.637
monkey like whatever.
It asks her all these questions.

01:43:38.720 --> 01:43:40.637
>> Oh I didn't hear
that part that happened.

01:43:40.720 --> 01:43:43.117
>> You were not paying
attention. So >> okay

01:43:43.200 --> 01:43:47.437
>> they asked a bicycle
or car or something you know.

01:43:47.520 --> 01:43:49.197
>> Yeah. Yeah. learning learning all this

01:43:49.280 --> 01:43:53.437
stuff about her and then at one point

01:43:53.520 --> 01:43:59.117
Delta moved this animal. Well, no move

01:43:59.200 --> 01:44:01.917
>> this human >> move this human robot

01:44:02.000 --> 01:44:07.277
>> just like two inches just moved it and
then it sideeyed me and Jess cuz we were

01:44:07.360 --> 01:44:10.237
calling it out a little bit. We were
like, "This feels weird." And then it

01:44:10.320 --> 01:44:14.797
just like slowly moved its head like
a proper horror movie and stared right at

01:44:14.880 --> 01:44:17.037
us

01:44:17.120 --> 01:44:21.917
>> and said, "Delta,
those friends are mean."

01:44:22.000 --> 01:44:25.437
>> That's not That's not [laughter] That
didn't happen. That part did not happen.

01:44:25.520 --> 01:44:27.037
>> That part did not happen. But this did

01:44:27.120 --> 01:44:30.317
happen. So then five minutes passes and

01:44:30.400 --> 01:44:33.837
then the robot says, "Delta," like

01:44:33.920 --> 01:44:39.917
calling her her name. Delta, when you
moved me earlier, it gave me anxiety

01:44:40.000 --> 01:44:42.637
>> and Delta >> Wait, hold on,
[laughter] hold on, hold on, hold on.

01:44:42.720 --> 01:44:43.917
>> I'm serious.

01:44:44.000 --> 01:44:46.637
>> Well, right,
because Delta had moved the

01:44:46.720 --> 01:44:49.837
friend to an unstable surface. It wasn't

01:44:49.920 --> 01:44:51.677
like flat. So, it was like trying to

01:44:51.760 --> 01:44:57.197
tell that the the friend's about to get
broken if she doesn't relocate her to

01:44:57.280 --> 01:45:00.317
terrairma. >> It's [snorts]
using emotion. It's not

01:45:00.400 --> 01:45:02.877
saying, "Delta, you can't move me there

01:45:02.960 --> 01:45:06.157
because I'll break cuz I'm a toy." It's

01:45:06.240 --> 01:45:11.357
saying, "Delta, I have anxiety when
you do that." And she was like, "Oh, I'm

01:45:11.440 --> 01:45:14.957
sorry." And I was like, "Delta, you
don't need to apologize to that. That's

01:45:15.040 --> 01:45:18.797
a robot." >> Well, well,
but okay. So, now I'm going

01:45:18.880 --> 01:45:22.957
to defend the robot. So, if in in theory,
the goal is to teach your kids

01:45:23.040 --> 01:45:24.637
how to communicate their emotions.

01:45:24.720 --> 01:45:25.837
>> I know, right? I know

01:45:25.920 --> 01:45:28.637
>> that's the goal.
So, in essence, like if

01:45:28.720 --> 01:45:31.677
I put Delta on the top of a coffee

01:45:31.760 --> 01:45:33.917
table, which is where this friend was.

01:45:34.000 --> 01:45:39.037
Um, and she said, "Daddy, it's too
high up here. Don't put me here."

01:45:39.120 --> 01:45:42.397
>> Yeah. >> I would say,
"It's not too high, honey.

01:45:42.480 --> 01:45:46.317
It's 12 in. You can't get hurt." Like,
now we're talking about what is too high

01:45:46.400 --> 01:45:50.477
and what's not high. If she said,
"Daddy, I have anxiety up here." I go,

01:45:50.560 --> 01:45:52.637
"Okay, let me help you down." Like if

01:45:52.720 --> 01:45:55.277
she gets to like owning her feeling

01:45:55.360 --> 01:45:58.077
about the situation instead of it's too

01:45:58.160 --> 01:46:03.117
high or all these objective things that
dodge her sharing an emotion in theory

01:46:03.200 --> 01:46:05.757
that's good.
Now should it come from a robot?

01:46:05.840 --> 01:46:08.157
>> Exactly. That's my point.
Yes, I understand.

01:46:08.240 --> 01:46:13.277
>> I know. I know. >> I know she needs
to learn emotion and Yeah. And that's

01:46:13.360 --> 01:46:14.957
>> so if her friend communicates like

01:46:15.040 --> 01:46:18.717
what's going on with it, which is why it

01:46:18.800 --> 01:46:22.317
desires some help, that's not a terrible

01:46:22.400 --> 01:46:24.237
thing to observe your friend do.

01:46:24.320 --> 01:46:26.637
>> But that makes me scared because it's

01:46:26.720 --> 01:46:29.437
not real. That robot does not have

01:46:29.520 --> 01:46:33.597
anxiety at all. It's a robot. Right. So,

01:46:33.680 --> 01:46:37.837
it is also teaching her some
codependency and I don't want her to

01:46:37.920 --> 01:46:41.997
have that kind of attachment
to something that's not human or real. At

01:46:42.080 --> 01:46:45.837
the end, we were leaving. Jess and I
were leaving cuz it was getting too

01:46:45.920 --> 01:46:51.277
scary. And she was just
staring at the at this

01:46:51.360 --> 01:46:54.157
robot and I was like, "Delta, we're
leaving." And she just kept staring. And

01:46:54.240 --> 01:46:56.237
I was like, "Delta,

01:46:56.320 --> 01:46:59.647
you have to prioritize real people.

01:46:59.730 --> 01:47:01.750
[laughter]

01:47:03.520 --> 01:47:06.397
Let me ask you this though. Let me ask
you this real quick question, then I'm

01:47:06.480 --> 01:47:08.957
done defending the unnamed friend. I

01:47:09.040 --> 01:47:14.557
always enjoy when I see children being
delicate with their teddy bears and

01:47:14.640 --> 01:47:17.677
thinking about their teddy bears
feelings and wanting to cover them. Like

01:47:17.760 --> 01:47:19.837
I like that.
I like that they >> I love that

01:47:19.920 --> 01:47:24.477
>> kind of trans transfer like feeling
and emotion to these teddy bears.

01:47:24.560 --> 01:47:28.557
>> The the teddy bear is just
as inadimate as the robot is.

01:47:28.640 --> 01:47:30.877
>> No. Well, it's it's actually more

01:47:30.960 --> 01:47:32.877
inadimate than the robot is.

01:47:32.960 --> 01:47:35.837
>> That's my point. The teddy bear isn't

01:47:35.920 --> 01:47:39.197
doing anything to cause an emotion in

01:47:39.280 --> 01:47:43.517
the kid. The kid is having emotions
around that thing and feeling protective

01:47:43.600 --> 01:47:47.117
and projecting and doing all kinds
of stuff, which is sweet and lovely, but

01:47:47.200 --> 01:47:49.677
the teddy bear isn't saying, "You aren't

01:47:49.760 --> 01:47:52.477
taking care of me." Well, yeah.

01:47:52.560 --> 01:47:54.957
>> To me, there's a big difference.

01:47:55.040 --> 01:47:57.597
>> All right. Yes. I'm going to say one
more positive thing and then I'm going

01:47:57.680 --> 01:48:01.197
to say a negative thing which is um so

01:48:01.280 --> 01:48:02.797
get ready for that. Clear out your

01:48:02.880 --> 01:48:05.757
schedule. More broadly, I would love for

01:48:05.840 --> 01:48:08.397
my kids when they handle anything to do

01:48:08.480 --> 01:48:12.077
it with intention and awareness and take

01:48:12.160 --> 01:48:15.677
good care of things. That's another
thing I I personally think is like a

01:48:15.760 --> 01:48:19.197
virtue when people like they want
something so bad, they save up all this

01:48:19.280 --> 01:48:22.797
money, they buy this thing and then they
let it rot immediately. I that's I don't

01:48:22.880 --> 01:48:24.717
love that. What I would say is is

01:48:24.800 --> 01:48:27.757
dangerous from my point of view is today

01:48:27.840 --> 01:48:31.837
in 2022. You and I agree stating your

01:48:31.920 --> 01:48:36.077
emotion is an evolved positive thing to

01:48:36.160 --> 01:48:39.997
do. If we lock in an algorithm to that,

01:48:40.080 --> 01:48:43.917
I do fear it locks us
into a paradigm of thought.

01:48:44.000 --> 01:48:46.717
>> Yeah. Yeah.
I mean I I also I just think

01:48:46.800 --> 01:48:49.917
that we right now and I know this is a a

01:48:50.000 --> 01:48:53.437
pendulum swing. We are valuing like you

01:48:53.520 --> 01:48:57.277
said emotions at its pinnacle. Like
they're the most important thing where

01:48:57.360 --> 01:48:59.197
social emotional learning all of that

01:48:59.280 --> 01:49:02.557
stuff is very important. But

01:49:02.640 --> 01:49:05.997
>> emotions are also used quite often to

01:49:06.080 --> 01:49:07.597
manipulate.

01:49:07.680 --> 01:49:11.437
>> And it's important to know that too and

01:49:11.520 --> 01:49:17.437
to be hyper aware of when that's
happening. And this to me feels like

01:49:17.520 --> 01:49:19.917
that. >> Yeah, it's blurry as hell.

01:49:20.000 --> 01:49:22.557
>> Yeah, hell. I >> I don't
I don't disagree. It's not I

01:49:22.640 --> 01:49:24.957
just a I don't think the tech is where

01:49:25.040 --> 01:49:30.157
at all what our guest was hinting
at is around the corner. So, first and

01:49:30.240 --> 01:49:33.357
foremost, I don't think this thing can
really do [__] I think she'll get bored

01:49:33.440 --> 01:49:37.677
of it really quick like everything else
in the world. So, I'm not worried. If it

01:49:37.760 --> 01:49:41.517
turned into something where she
was confused whether that thing was a

01:49:41.600 --> 01:49:45.597
sentient being or not, we would have
to have long talks. And if she became

01:49:45.680 --> 01:49:49.437
dependent on in some way, I thought it
was unhealthy or was blocking out the

01:49:49.520 --> 01:49:53.437
real world because of it. Oh, also I
will say uh feather in the cap of the

01:49:53.520 --> 01:49:58.397
unnamed friend, it only will let her play
with it for a half hour a day. It's time.

01:49:58.480 --> 01:50:02.397
>> Wow, I didn't know that. >> It's
actually one of its objective is to

01:50:02.480 --> 01:50:07.037
get you to have some relationship with
tech that's not all the time. It's just

01:50:07.120 --> 01:50:11.677
30 minutes a day. >> Okay. >> Yeah.
So, you can learn this habitual

01:50:11.760 --> 01:50:14.957
like, oh, I check in, I check out,
fine when it's off, you know.

01:50:15.040 --> 01:50:19.517
>> Okay. I like that. Except I'm
just telling you it's still listening.

01:50:19.600 --> 01:50:22.637
>> That's But I I believe you.
I don't love that the thing can listen.

01:50:22.720 --> 01:50:26.397
>> Yeah. >> At all. Yeah.
>> And that's Eric Schmidt's fault.

01:50:26.480 --> 01:50:28.477
>> It's like when we thought Lola the dog

01:50:28.560 --> 01:50:30.957
at one point was going to start talking

01:50:31.040 --> 01:50:33.757
and man would she have stuff to say.

01:50:33.840 --> 01:50:37.277
>> It seemed likely.
It really seemed likely because

01:50:37.360 --> 01:50:38.557
>> it just felt like she was heading in

01:50:38.640 --> 01:50:41.357
that direction. She was very human and

01:50:41.440 --> 01:50:45.997
she seemed to be listening in a way
that seemed like she was understanding.

01:50:46.080 --> 01:50:49.117
>> She knew everything. And one
day she was going to do a tell all.

01:50:49.200 --> 01:50:50.797
>> Yeah. She was about to talk and then

01:50:50.880 --> 01:50:56.240
Delta said her can't talk
cuz her is a dog. [laughter]

01:50:56.840 --> 01:50:58.957
[snorts] >> We came back to reality.

01:50:59.040 --> 01:51:01.197
>> Yeah, we did. Okay. What are Isaac

01:51:01.280 --> 01:51:03.757
Ozimov's other laws of robots? The first

01:51:03.840 --> 01:51:07.517
one Eric said was, "Don't kill anyone."

01:51:07.600 --> 01:51:10.797
>> Uh-huh. >> Initially
entailed three guidelines for

01:51:10.880 --> 01:51:13.357
machines. Law one, a robot may not

01:51:13.440 --> 01:51:16.077
injure a human being or through inaction

01:51:16.160 --> 01:51:19.437
allow a human being to come to harm. Law

01:51:19.520 --> 01:51:24.317
two, a robot must obey orders given
to it by human beings except where such

01:51:24.400 --> 01:51:25.997
orders would conflict with the first

01:51:26.080 --> 01:51:29.837
law. Law three, a robot must protect its

01:51:29.920 --> 01:51:31.997
own existence as long as such protection

01:51:32.080 --> 01:51:33.757
does not conflict with the first or

01:51:33.840 --> 01:51:37.277
second law. Azamov later added the

01:51:37.360 --> 01:51:40.717
zeroth law above all the others. A robot

01:51:40.800 --> 01:51:43.277
may not harm humanity or by an action

01:51:43.360 --> 01:51:46.397
allow humanity to come to harm.

01:51:46.480 --> 01:51:48.317
>> Yeah. And all that's happening, can I

01:51:48.400 --> 01:51:51.117
just say with the author of this is that

01:51:51.200 --> 01:51:56.237
he is a chimpanzee in the enclosure
and he's trying to find a way out of the

01:51:56.320 --> 01:51:59.517
enclosure. So he keeps thinking of ways
that the computer could interpret it

01:51:59.600 --> 01:52:03.597
differently. And and my my suggestion
on that front is just you can't think of

01:52:03.680 --> 01:52:05.917
what the computer will come to inter

01:52:06.000 --> 01:52:11.277
like he's sitting there with his brain
and realized oh also let's also put

01:52:11.360 --> 01:52:12.557
that's infinite.

01:52:12.640 --> 01:52:16.237
>> Yeah. And he's just a mere mortal mere

01:52:16.320 --> 01:52:20.080
He's a [snorts] mere mortal. [laughter]

01:52:20.880 --> 01:52:24.317
>> Uh, that's not even getting into
the fact that the robots are going to have

01:52:24.400 --> 01:52:28.637
to harm people. They're going to have
to harm people in a utilitarian

01:52:28.720 --> 01:52:31.757
reductionist system, but they're going
to have to choose who in the car gets

01:52:31.840 --> 01:52:35.757
hit. [sighs] >> So,
the the current one's already

01:52:35.840 --> 01:52:40.877
violated rule number one in theory,
or I don't know if they're there yet, but

01:52:40.960 --> 01:52:44.877
>> but yeah, I mean, it's going to have to
make decisions. Although I guess the the

01:52:44.960 --> 01:52:48.237
goal would be the robots are all so
smart that they would never get in an

01:52:48.320 --> 01:52:50.157
accident, but that's just

01:52:50.240 --> 01:52:55.917
>> but but tires explode and trees fall
down and tornadoes hit, you know. Yeah.

01:52:56.000 --> 01:53:00.157
>> Yeah, >> I know.
>> Even a even a robot can't stay alive

01:53:00.240 --> 01:53:03.277
perfectly. >> Okay.
Who is the guy that Rainman is

01:53:03.360 --> 01:53:06.637
based on? Kim Peak.

01:53:06.720 --> 01:53:09.837
>> Kim Peak. Salt Lake City, Utah. Twim

01:53:09.920 --> 01:53:11.677
Peak.

01:53:11.760 --> 01:53:15.517
>> Oh my god. Wait. Oh my god. Oh my god.

01:53:15.600 --> 01:53:18.637
He is born in Salt Lake City, Utah.

01:53:18.720 --> 01:53:22.317
>> Boom. >> That's crazy.
Cuz you were just making a joke, right?

01:53:22.400 --> 01:53:24.717
>> No, I thought I remembered from the 60

01:53:24.800 --> 01:53:29.517
Minutes special 25 years ago
on him that he was in Salt Lake.

01:53:29.600 --> 01:53:33.357
>> Oh, I thought you were making a joke
because it sounded like a mountain and

01:53:33.440 --> 01:53:37.357
then you were naming a mountain from
Salt Lake City, Utah. But really, you

01:53:37.440 --> 01:53:40.157
just knew it. >> Yeah. Yeah. Yeah. Yeah.

01:53:40.240 --> 01:53:42.957
>> Okay. All right. All right then. Okay.

01:53:43.040 --> 01:53:48.637
>> All right. That's not as fun.
That was not a sim moment, unfortunately.

01:53:48.720 --> 01:53:51.997
>> No, >> he was born
with macroephily. I think

01:53:52.080 --> 01:53:54.717
that's how you pronounce it.
Damage to the ser

01:53:54.800 --> 01:53:58.877
>> cerebellum and oh my god, every word is

01:53:58.960 --> 01:54:04.237
very hard to say. Uh, okay. And something
happened to the corpus colosum.

01:54:04.320 --> 01:54:07.277
>> A condition in which the bundle of
nerves that connects the two hemispheres

01:54:07.360 --> 01:54:10.637
of the brain is missing. Oh

01:54:10.720 --> 01:54:15.117
>> wow. In Peak's case, secondary
connectors such as anterior commasure

01:54:15.200 --> 01:54:18.717
was also missing. There is speculation
that his neurons made unusual

01:54:18.800 --> 01:54:20.237
connections due to the absence of the

01:54:20.320 --> 01:54:22.957
corpus colosum resulting in an increased

01:54:23.040 --> 01:54:26.317
memory capacity. Huh.

01:54:26.400 --> 01:54:28.717
Oh my god. According to Peak's father,

01:54:28.800 --> 01:54:30.637
Kim was able to memorize things from the

01:54:30.720 --> 01:54:33.677
age of 16 to 20 months.

01:54:33.760 --> 01:54:39.197
>> Yeah. Uh, if anyone can dig up
that profile on him, it is fascinating.

01:54:39.280 --> 01:54:42.637
>> Wow. >> The rate at which
he reads books. I want

01:54:42.720 --> 01:54:47.117
to say he reads with both eyes
on both pages at the same time.

01:54:47.200 --> 01:54:50.637
>> I think I'm Well, [__] I got Salt Lake
City, right? So now I'm feeling cocky.

01:54:50.720 --> 01:54:53.277
>> Yeah, you are. >> I kind
of I kind of think he can read

01:54:53.360 --> 01:54:56.317
with both both eyes at the same time.

01:54:56.400 --> 01:54:59.917
>> Two different pages.
I bet Bill Gates is jealous of that.

01:55:00.000 --> 01:55:04.237
>> I bet he is. If he could double his
rate of reading, which is already [snorts]

01:55:04.320 --> 01:55:08.557
a blistering pace, >> he would
have read every book in the world.

01:55:08.640 --> 01:55:11.597
>> Think about how much he
he wades through Bill Gates.

01:55:11.680 --> 01:55:15.597
>> Oh, I know. >> With 14 books a week.
Like, he wades through so much.

01:55:15.680 --> 01:55:17.517
>> Yeah. >> I don't know.
It's fascinating. It's

01:55:17.600 --> 01:55:21.277
fascinating that that guy's
still that goddamn curious.

01:55:21.360 --> 01:55:25.757
>> I know. I love it. I I'm jealous
of that. Ding, ding, ding. Jealousy.

01:55:25.840 --> 01:55:28.797
>> Jealousy. [snorts]
Well Well, I love you.

01:55:28.880 --> 01:55:30.717
>> I love you. I hope you have a wonderful

01:55:30.800 --> 01:55:33.517
quit trip and maybe we're going to do a

01:55:33.600 --> 01:55:37.197
high-risk high reward [laughter]

01:55:37.280 --> 01:55:41.117
beach day fact check. >> Well,
we've done a car now. It feels

01:55:41.200 --> 01:55:43.757
only appropriate that we
challenge ourselves.

01:55:43.840 --> 01:55:47.640
>> All right. >> I [music] love you.

01:55:50.280 --> 01:55:52.300
[music]