The Link Between Artificial Intelligence Life Extension

[Music]

i intend to live forever

or die trying

in when i was born i was born

with a genetic challenge and so

members of my family had a thing called

polycystic kidneys and basically what

that is

is a kidney that ends up having

fluid-filled cysts and ultimately

your kidney fails you either go on

dialysis or you get a kidney transplant

my grandfather died at 40 years old from

polycystic kidneys

back in the 50s so

i go to wild cornell medicine

new york presbyterian hospital great

bunch of doctors

and i uh in 2008

get prepared for a kidney transplant

i got a special kind of transplant it

was called an

incompatible transplant

and it was amazing it was amazing

because what they basically could do

is give me a transplant an

organ from somebody with a different

blood type

now typically going back 10 years before

that that was impossible

right you would automatically get the

transplant you would reject the

transplant that would be end of it but

what they do what they did for me

is they create they remove the

antibodies

from my blood through a process called

plasmapheresis and then they

just jammed in a bunch of immunoglobulin

and then they put you on a protocol once

you’re done with the transplant and you

stay on this medicine

and that’s amazing that was amazing to

me

now when you think about the first

kidney transplant happening

in 1954 that was the first

transplant and you think about the

period right after that from 1996 to

2000

where they were able to not only do

transplants

but to incompatible blood type

transplants

and then my transplant in 2008

that’s an amazing period of time where

it went from my grandfather dying

at 40 years old in the 50s to me being

10 years out in a transplant

and having just like an awesome life

feeling great

doing great it’s awesome today

they do almost no compatible

transplants anymore

and the reason is they’ve gotten so

advanced in the last 10 or 11 years

on advanced kidney swaps that it’s no

longer required so just think about how

awesome the science is around that

this talk is really intended especially

for

a lot of the people in the audience and

i can imagine a lot of you are

relatively young this is going to

certainly add

two years to your life if you think

about things in terms of how this

information is presented

and most likely many of you especially

the younger people that are here from

high school

many of you are going to live to well

over

100 years old and so what i want to talk

to you about today is the link

between artificial intelligence and a

longer life

this is what i would refer to as the

ultimate paradigm shift and why is it

the ultimate paradigm shift

if you think about it what is more

powerful

to most people than living a longer

productive life

there is nothing more powerful when i

first got out of college

i went to work at new york presbyterian

hospital

and there two or three years before

the shah of iran had stayed and he had

spent

millions upon millions of dollars just

to renovate a floor

so he could receive his cancer treatment

and it dawned on me

there was nothing more important to him

and in spite of all the money he had

he wouldn’t be able to expand the

longevity

for his life and his wellness

setting that aside it’s going to be a

one

trillion dollar industry by 2030.

so you just imagine how powerful that is

um what what’s driving this thing

to make you be able to live beyond a

hundred the first is genome science we

all know

the challenges associated with the human

genome we all know about the advances

that have occurred

second is big data and big data is

around the data that you collect about

individuals and

third is all of the devices that you can

use today from a wearable standpoint

to track this information in real time

dynamically

what i’m going to talk to you about is

how that intersects with artificial

intelligence

and they’re a big companies that are

focused on

providing this information now right if

you think about alphabet

many of you know the the parent company

of google

novartis illumina these are companies

that are bringing

this technology to the cusp of enormous

unprecedented increase in increases in

wellness

[Music]

and so a lot of stuff is being written

now about how ai can

become the type of breakthrough that

scientists

have developed to add 20 years on

average to your life expectancy

there’s a lot of research associated

with that

and what’s interesting about the

research

is that people who live past 100

get sick much later in life and for a

shorter period of time

so what do we do to create technology to

understand

how to solve for that based on the

realization of people living

longer lives the first is find out their

secrets

and then develop drugs that mimic the

ability for them

to respond better over life

and then deliver that to the rest of us

the second is

sort data now machine learning is the

kind of thing where you take data

you put it in a machine the machine

learns you train it

you look at it you do what’s called

backward propagation you look at the

back

of the results of the data you see where

you went wrong

but what they’re suggesting today is to

use machine learning to sit with

scientists

to work together hand in hand to

innovate

and replicate as long as the machines

are running

third is to look at the similarities in

terms of dna

of those that live longer and then

obviously devise a plan

for how you can identify and block

a molecular pathway

scientists have built a number of ai

models to analyze data

now one of the challenges of data and

the analysis of data

is the hipaa regulations we all know

about that but they actually have built

some very advanced technology to be able

to assess an individual

from the time they’re born and then

assess what are the risks of them dying

young so

for example the university of nottingham

is doing a lot of work in that area

and what they’ve realized is by doing

this kind of math

medical knowledge will double will

double

every 73 days by 2020

as compared to every three and a half

years in the past

the second is as i mentioned earlier

this genomic sequencing

has come down so dramatically in price

that the abilities that afford you the

opportunity

to do something genetically is is is

much faster approaching

that we found from an artificial

intelligence perspective

is this notion of iq why is it a

challenge

they have done a number of studies and

they have been able to correlate

iq with

longevity and so that is now a big part

of what’s coming out and what they refer

to as cognitive

epidemiology uh there is

obviously a the research is clear

there’s a strong link

iq and and it’s and it’s proven

with an iq of 100 just 15 points

[Music]

they did this test and what they did was

the biggest form of backward propagation

you could imagine they started with

students

in 1932 they administered an iq test to

them

and they basically waited for them to

get old and once they realized that they

got old they looked at the people with

the 15-point iq advantage

and they realize that that was a 21

greater chance of that person surviving

now the real question is what does that

mean

right what does that mean is that

basically that

people are smart enough to recognize

that if they exercise and don’t smoke

and wear a seatbelt

that 15 points correlates to a longer

life or is it based on

what i was referring to earlier in terms

of this uh genetic

propensity and these people in the space

of epidemiological research

are really coming up with those answers

and so you know there’s many studies of

people it’s like

common sense right you some group of

people who live in greece or

live in italy the average age is a lot

higher than normal

it’s based on a mediterranean diet we

kind of get that

you know we understand that if you smoke

you have a higher risk of death and

you want to kind of make sure you don’t

smoke and these are pretty

simple things but the thing about iq and

what my company focuses on is partially

on iq

is that it raises enormous ethical

standards and can

really generate significant social

injustice i mean if you think this is

the absolute paradigm shift

if you think that the richest people in

the world are going to want to live

longer

what kind of ethical challenges does

that deliver

right it’s going to be extremely

expensive at first i mean if you think

about

universal health care whatever side of

the political

equation you’re on if you think about

health care think about what it would

mean to somebody to be able to add years

to their life and what that would pay

for that

and so these disparities are going to

continue to be a big challenge so i

refer to it as

inequality and mortality and some of the

dilemmas

are really really challenging dilemmas

right so for example are we committed to

extending life

indefinitely if we can uh

what are we how are we committed to life

saving will everyone have an equal

chance

uh there’s enormous number of questions

that are that arise

uh is the loss of a child and the

passing of an elderly person the same

thing so just imagine

how challenging this is going to be to

get your

hands wrapped around the biggest

opportunity here is to educate the

public that

iq for example predicts outcomes and

mortality to a certain extent but it’s

not a destiny

it’s not the panacea and

also understand what iq means right we

have

one kind of way of thinking about iq but

there’s also ethical iq and that’s an

approach to dealing with issues

and developing studies on

intelligence that may not have to do

with the traditional way that somebody

takes an itq test from a pattern

perspective

meaning my company works by building

a bunch of artificial intelligence

models that are culture fair and really

what culture fair is an equivalent to

wisdom

right you can see somebody older you say

wow they’re really smart

it may not be in the traditional sense

that you think but that wisdom

provides insights and by the way that

wisdom provides insights

that transcend language right a lot of

times people fail iq tests because of

the language background

so when you’re thinking about a world of

122 year olds there’s enormous

promise associated with the change in

these lifespans

now think about this statistic if you if

in the united states we were able to

cure cancer

cure cancer the average life expectancy

would increase by two and a half years

yet if we were to able to focus on

things like heart disease stroke and

diabetes these are things that we can

control to a certain extent life

expectancy would increase

by another 14 years so you’re thinking

about 16 years

of those 16 years that your life

expectancy can increase

only two of those years are related to

cancer

obviously when we’re talking about ways

to help yourself there are things

associated with caloric

in restrictions watching how much you

eat

so for example there’s a recent study

that said if uh

a rodent were

reduced its number of calories it could

basically

equate to the equivalent of a 90 year

old

resembling today’s 50 year old

if you think about it in those terms if

they were to apply this to humans you’re

talking about an

enormous expanse and increase in the

number of years

that you can live

one of the challenges in health care is

not to try to cure

an individual disease but think about a

far greater impact

overall on public health by focusing on

aging research and applying common sense

approaches tied to data

now when we uh we believe strongly that

ai

and research are joined hip and thigh

and what that basically means is

the ability to calculate with math

things like what is your real age

there are two principal parts of

artificial intelligence the first is

machine learning and that’s basically

where you take data

you analyze the data you learn from the

data

you apply that learning and the machine

gets smarter it’s basically the

equivalent

of an exponential number of data

scientists in the room crunching numbers

and the second is a thing called natural

language processing a little less known

but it’s basically taking the signals

that human beings

emit video

audio the way they speak how they speak

and taking that information

turning that into an insight about that

individual and applying it to their

wellness

to their meaningfulness in life to their

capacity to do better in the world so

meetingbot works on all of this

as you know i mentioned earlier genomics

if you think about genomics

think about the fact that the cost to

sequence

the genome was 100 million

dollars seven to 18 years ago

and it now costs a thousand dollars

so that’s the intersection of math with

uh processing power

when you consider

that the most advanced ai models

correctly identified the risk of

76 of patients dying prematurely

by using ai versus prior methods of only

44

that kind of gives you a sense for what

the math is all about and how powerful

that math is

so with meeting mod there are a lot of

companies out there that

do what we do right they do it slightly

different we use

artificial intelligence to understand

people’s words

build models to help them understand

themselves better

and apply that to their careers things

related to

making sure the data is transparent

making sure it’s open we want to share

that data with everybody

but one thing i’d like to have you think

about when you go to a doctor how often

does the organization collect your words

dynamically if we were able to take the

words that you use

process those words it will give us

enormous insights

about who you are what you’re feeling

and we can apply that math

via natural language processing to

understanding

exactly what’s affecting you and apply

that longitudinally

so in summary right follow the money

this is going to be an enormous change

that’s going to create significant

challenges for the world

uh ai is changing everything this notion

of immortality is something you’re going

to hear about

and bias is going to become a social

crisis

unless we acknowledge the importance

of being fair and equitable to everyone

so i leave you with a quote live as if

you were to die tomorrow

learn as if you were to live forever and

this is the first time in history where

there’s

two things that intersect together and

are not desperate thoughts

give the gift of life thank you very

[Music]

[Applause]

much

[音乐]

我打算

在我出生时永远活下去或死去我出生

时就有遗传挑战,所以

我的家人有一种叫做多囊肾的东西

,基本上这

是一个最终

充满液体的肾脏 囊肿,最终

你的肾衰竭,你要么进行

透析,要么进行肾移植,

我的祖父在 50 年代死于多囊肾,享年 40 岁,

所以

我去了野生康奈尔

医学纽约长老会医院

2008 年

为肾移植做准备

我接受了一种特殊的移植,它

被称为

abo 不相容的移植

,这太棒了,太棒了,

因为他们基本上可以做的

就是给我移植

一个来自不同血型的人的器官

现在通常 回到 10 年前

那是不可能的

对 你会自动接受

移植 你会拒绝

移植 那将是它的结束 但

他们会做什么 他们为我所做的

是他们创造了他们

通过称为血浆置换术的过程从我的血液中去除抗体

,然后他们

只是塞进一堆免疫球蛋白

,然后一旦你完成移植,他们就会给你安排一个协议

,你

继续

当你想到 1954 年发生的第一次

肾移植

,那是第一次

移植,你想到

了 1996 年到 2000 年之后的这段时间

,他们不仅能够进行

移植,

而且能够进行移植,这让我感到惊讶。 到不相容的血型

移植,然后是我在 2008 年的移植,

那是一段令人惊叹的时期

,从我祖父

在 50 多岁时 40 岁去世,到我

在移植手术中 10

年,就像过着美好的生活一样

感觉

很棒 太好了,今天太棒了,

他们几乎不再进行与 abo 兼容的

移植,

原因是他们

在过去 10 年或 11 年

的先进 k 上变得如此先进 idney 表示

不再需要它,所以想想

科学是多么的棒,

这个演讲真的是专门

针对

观众中的很多人的,

我可以想象你们中的很多人都

相对年轻,这

肯定会增加

如果您从如何呈现这些信息的角度来考虑事情,那么您的生命还有两年,

而且很可能你们中的许多

人,尤其是从高中毕业的年轻人

,你们中的许多人将活到

100 岁以上,等等 我

今天想和你谈谈

人工智能和

更长寿命之间的联系

这就是我所说的

终极范式转变以及为什么它

是终极范式转变

如果你想一想对大多数人来说更

强大的东西

我刚从大学毕业时,

我在纽约长老会医院工作

,比国王早两三年去那里工作,没有什么比活得更久

了 伊朗人留下来了,他

花了

数百万美元

来翻修地板,

这样他就可以接受癌症治疗

了,我突然意识到,

对他来说没有什么比这更重要了

能够

延长他的生命和健康的寿命,

但到 2030 年它将成为一个价值

1

万亿美元的产业。

所以你想象一下,这是多么强大,是

什么驱使这

件事让你能够生活在

第一个是基因组科学 我们

都知道

与人类基因组相关的挑战

我们都知道已经发生的进步

第二个是大数据,大

数据围绕着你收集的关于个人的数据

第三个是你所有的设备 可以

从可穿戴的角度使用今天

来动态跟踪这些信息

我要和你谈谈的是它

如何与人工智能相交

,它们是一个大c 如果您想到字母表,

现在就专注于提供这些信息

的公司,你们中的许多人都知道

谷歌

诺华 illumina 的母公司,这些

公司将

这项技术带到了

健康

[音乐]

和 所以现在有很多

关于人工智能如何

成为科学家已经开发出来的突破类型的文章,

可以

让你的预期寿命平均增加 20 年,

有很多与此相关的研究,研究

的有趣之

处在于 活到 100

岁以后会在更晚的时间和

更短的时间内生病,

所以我们应该做些什么来创造技术来

了解

如何解决这个问题

模仿

他们

在一生中做出更好反应的能力的药物

,然后将其传递给我们其他

人第二个是

排序数据 现在机器学习是这样

一种事情,你把数据

放进机器机器

学习你训练它

你看它你做所谓

的反向传播你看

数据的结果你看到

你去了哪里 错了,

但他们今天的建议是

使用机器学习与科学家坐在

一起,携手合作进行

创新

和复制,只要机器

在运行

第三是查看

那些寿命更长的人在 DNA 方面的相似之处 然后

显然设计了一个

如何识别和

阻断分子途径的计划

科学家们已经建立了许多人工智能

模型来分析数据

现在数据的挑战之一是

数据分析

是我们都知道的 hipaa 法规

,但他们 实际上已经建立了

一些非常先进的技术,能够

从他们出生的时候就评估一个人,然后

评估他们年轻时死去的风险是什么

,例如大学 诺丁汉的 y

在该领域做了大量工作

,他们意识到,通过

这种数学,

医学知识

到 2020 年每 73 天将翻一番

,而

在过去的每三年半中

,第二 正如我之前提到的,

这种基因组测序

的价格已经大幅下降,以至于让

你有

机会做一些基因工作的能力正在

更快地

接近我们从人工智能的角度发现

的这个概念是智商的概念,为什么它是一个

挑战

他们已经进行了许多研究,

他们已经能够将

智商与

长寿联系起来,所以这现在是

即将出现的事情的重要组成部分,他们

称之为认知

流行病学

链接

智商和它

的智商证明只有 100 点 15 点

[音乐]

他们做了这个测试,他们所做的是你可以做

的最大的反向传播

形式 agine

他们在 1932 年开始对学生进行智商测试

,他们基本上是在等他们

变老,一旦他们意识到自己

变老了,他们就会看着

智商有 15 分优势的人

,他们意识到那是 21

那个人现在活下来的机会更大

真正的问题是这

意味着什么 这意味着基本上

人们足够聪明地认识到

,如果他们锻炼、不吸烟

和系安全带

,那么 15 分与 寿命更长,

还是基于

我之前提到

的遗传

倾向

,流行病学研究领域的这些

人真的想出了这些答案

,所以你知道有很多关于

人的研究,就像

常识一样 你们

是住在希腊或

意大利的一群人,平均年龄

比正常人高很多,

这是基于地中海饮食,我们

有点明白,

你知道我们和 理解,如果你吸烟,

你有更高的死亡风险,并且

你想确保你不

吸烟,这些都是非常

简单的事情,但关于智商的事情以及

我的公司关注的部分

智商,它提高了 巨大的道德

标准,并且

真的会产生重大的社会

不公正我的意思是,如果你认为这

是绝对的范式转变,

如果你认为世界上最富有的

人想要活得

更久,

什么样的道德挑战

会带来正确的结果? 一开始会非常

昂贵 我的意思是,如果你考虑

全民医疗保健,无论

你处于政治等式的

哪一边 会为此付出

代价的

,所以这些差异将

继续是一个巨大的挑战,所以我

将其称为

不平等和死亡率,其中一些

困境确实非常具有挑战性 问题是

对的,例如,

如果我们可以

,我们是否致力于无限期地延长生命?我们是什么?我们如何致力于拯救生命?

每个人都有平等的

机会

吗?有大量的

问题出现了,

呃,失去了一个孩子,

一位老人的逝世也是同样的

事情,所以想象

一下这将是多么具有挑战性,

让你的

双手紧紧

抓住这里最大的机会是教育

公众,

例如智商

在一定程度上预测结果和死亡率,但这

不是 命运

它不是灵丹妙药,

也明白智商意味着什么我们

一种思考智商的方式,但

也有道德智商,这是一种

处理问题

和开展

智力研究的方法,可能

与传统方式无关 有人

从模式的角度进行 itq 测试,这

意味着我的公司通过构建

一堆人工智能

模型来工作,这些模型是文化 公平和真正

什么文化公平等同于

智慧

对你可以看到一个年长的人你说

哇他们真的很聪明

这可能不是你认为的传统意义上的

智慧,但智慧

提供了洞察力,顺便说一下,

智慧提供了

洞察力 超越语言 很多

时候人们因为语言背景而未能通过智商测试,

所以当你想到一个 122 岁的世界时,

与这些寿命的变化相关的巨大希望

现在想想这个统计数据,如果你

在美国 声明我们能够

治愈癌症

治愈癌症平均预期寿命

将增加两年半

但是如果我们能够专注

于心脏病中风和糖尿病等疾病,

这些都是我们可以

在一定程度上控制

预期寿命的事情

将再增加 14 年,因此您正在

考虑这 16 年

中的 16 年,您的

预期寿命可以增加,其中

只有两年是相关的

当我们谈论

帮助自己的方法时,

很明显会患上

癌症 相当于一个 90

岁的人,

类似于今天的 50 岁,

如果你从这些方面考虑,如果

他们将其应用于人类,你正在

谈论的是一个

巨大的扩张和增加

你可以活

的年数 医疗保健面临的挑战

不是试图

治愈单个疾病,而是考虑

通过关注

老龄化研究和应用

与数据相关的常识方法,对整体公共卫生产生更大的影响,

因为我们现在坚信

人工智能

和研究是 连接臀部和大腿

,这基本上意味着

用数学计算的能力,

比如你的真实年龄

有两个主要

部分 第一个是

机器学习,这基本上

是你获取数据的地方,

你分析你从应用该学习的数据中学到的

数据

,机器

变得更聪明,它基本上

相当于房间里的数据科学家的指数级数字

,第二个 是一种称为自然

语言处理的东西,鲜为人知,

但它基本上是

接收人类

以说话方式发出视频音频的信号,

并将这些信息

转化为对个人的洞察力

并将其应用于他们的

健康

生活中的意义对他们

在世界上做得更好的能力,所以

如你所知,我之前提到过基因组学,

如果你考虑基因组学,想想

基因组测序的成本是 1 亿

美元,需要 7 到 18 年。

以前它现在要花费一千美元

所以这是数学与

呃处理能力的

交集 你

认为最先进的人工智能模型

正确地识别了

76 名患者

使用人工智能过早死亡的风险,而之前的方法只有

44 种

,这让你

了解数学的全部内容以及

数学的强大

会议模式有很多

公司

做我们做对的事情他们做的略有

不同我们使用

人工智能来理解

人们的话

建立模型来帮助他们

更好地了解自己

并将其应用于他们的职业生涯

确保数据相关的事情 是

透明的 你使用

处理这些词它会给我们

关于你是谁你的感受的巨大见解

,我们可以

通过自然语言处理应用该数学来

理解

准确了解影响您的因素并

纵向应用,

因此总而言之,正确跟随金钱,

这将是一个巨大的变化

,将为世界带来重大

挑战

除非我们承认

对每个人都公平和平等的重要性,否则偏见将成为一场社会危机,

所以我给你留下一句名言,活得好像

你明天就会死去

学习,好像你要永远活着,

这是第一个 历史上

两件事相交而

不是绝望的时间

给予生命的礼物非常感谢

[音乐]

[鼓掌]

非常